| Home |
| Search |
| Today's Posts |
|
#1
Posted to rec.audio.tubes
|
|||
|
|||
Boosted Triode
This seems like an interesting idea to get more power out of a triode.
http://www.edn.com/article/CA302240.html Cheers Ian |
|
#2
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
flipper wrote:
On Tue, 12 May 2009 10:42:36 +0100, Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Yeah, I saw that one a while back but I'm not sure it's worth it because that takes a fully floating supply for each tube and it might be simpler to just do it the old fashioned way, double the output tubes. I presume you mean double as in parallel rather than push pull? Given the cost of the extra PSU then there's probably not much in it cost wise. Cheers ian |
|
#3
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Ian Bell wrote:
This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Those load lines aren't what the author says they are and he doesn't say how he measured power. Is it a joke? Maybe it works, but in reality not for long. In simulation you can't melt the screen. Unless he, or I, have got the battery symbol the wrong way round, it appears that the screen voltage is set at 100V greater than the anode. The ensuing screen meltdown could be avoided by increasing the anode voltage by the same amount, i.e. by simply increasing the HT and discarding the extra screen supplies. Perhaps that takes the valve out of it's SOA? Not so far as does setting Vs Va, surely? In the introduction he likens the anode to a JFET drain. That analogy is only true at higher voltages, or at low voltages only if the screen voltage is very low. Otherwise the screen becomes the drain. It looks to me like the screen would be passing most of the current in that region where he thinks he's found more power. The whole article is a garbled half-story, AFAICS on the face of it. If you want more triode power, use more valves, or bigger ones. Or use in pentode mode with a distributed load or with global feedback. It's not like there's a shortage of options. It's not even a credible retrofit...get more power out of your existing amp...because it requires a much greater drive signal and a bias voltage to suit. Ian |
|
#4
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
On May 12, 12:11 pm, flipper wrote:
On Tue, 12 May 2009 10:42:36 +0100, Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Yeah, I saw that one a while back but I'm not sure it's worth it because that takes a fully floating supply for each tube and it might be simpler to just do it the old fashioned way, double the output tubes. I'd say it is altogether a pretty grim idea, even if it works. Just for a start, the power supply is always the element making the nastiest class of residual noise on any amp's output. And let us not forget that the power supply, wherever it is fed into the tube, is in the signal circuit, something that these crazed power-hunters have a hard time understanding. So now, to get a little extra power, the amp requires no fewer than three power supplies. Yech! No wonder tha creator of the circuit does not publish noise figures. A whole lotta NFB loops might help but the recombinant spectra will be very nasty indeed, and the micro-view of all this on the waterfall of the sort of high-sensitivity speaker normally used with SE amps will be confused indeed. At the very least matters should be so arranged that the output rolls off very quickly under 50 or 60Hz, depending on where in the world you are. (Be even better to roll this amp off at 120Hz and bi-amp a woofer under it, but that is a kludge even further away from a quality sound.) It is this sort of overweening design that makes for "edgy" amps, the sound too brittle to listen to for long. There are stranger considerations in audio amps than are dreamt of in thy engineering, Horatio. In Scandinavia, in Europe, and in Australia those floating supplies are also ipso facto illegal unless they're double shielded and certified, either being a practical impossibility for the DIYer. People with their brains in gear, when they want louder music, first get more sensitive speakers, then make their SE amps with more powerful tubes, then use parallel tubes SE (cheaper anyway than this grim idea), then go to triodes in PP (operated strictly in Class A1, of course), then PP triode-linked pentodes in Class A1, and so on. Like I said, I'm talking about people with their brains in gear: people with their brains in gear don't perpetrate complicated audio electronics just because they can. Andre Jute Visit Jute on Amps at http://www.audio-talk.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review |
|
#5
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Ian Iveson wrote:
Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Those load lines aren't what the author says they are and he doesn't say how he measured power. Is it a joke? Maybe it works, but in reality not for long. In simulation you can't melt the screen. Unless he, or I, have got the battery symbol the wrong way round, it appears that the screen voltage is set at 100V greater than the anode. I believe that is the case - the screen is 100V above the plate. The ensuing screen meltdown could be avoided by increasing the anode voltage by the same amount, i.e. by simply increasing the HT and discarding the extra screen supplies. Perhaps that takes the valve out of it's SOA? Not so far as does setting Vs Va, surely? In the introduction he likens the anode to a JFET drain. That analogy is only true at higher voltages, or at low voltages only if the screen voltage is very low. Otherwise the screen becomes the drain. It looks to me like the screen would be passing most of the current in that region where he thinks he's found more power. The whole article is a garbled half-story, AFAICS on the face of it. If you want more triode power, use more valves, or bigger ones. Or use in pentode mode with a distributed load or with global feedback. It's not like there's a shortage of options. There's one actual design I have found in an audioxpress magazine advertised on the hi-fi collective site but no other example I can find. Cheers ian It's not even a credible retrofit...get more power out of your existing amp...because it requires a much greater drive signal and a bias voltage to suit. Ian |
|
#6
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
On 12 May, 18:26, Ian Bell wrote:
Ian Iveson wrote: Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Those load lines aren't what the author says they are and he doesn't say how he measured power. Is it a joke? Maybe it works, but in reality not for long. In simulation you can't melt the screen. Unless he, or I, have got the battery symbol the wrong way round, it appears that the screen voltage is set at 100V greater than the anode. I believe that is the case - the screen is 100V above the plate. The ensuing screen meltdown could be avoided by increasing the anode voltage by the same amount, i.e. by simply increasing the HT and discarding the extra screen supplies. Perhaps that takes the valve out of it's SOA? Not so far as does setting Vs Va, surely? In the introduction he likens the anode to a JFET drain. That analogy is only true at higher voltages, or at low voltages only if the screen voltage is very low. Otherwise the screen becomes the drain. It looks to me like the screen would be passing most of the current in that region where he thinks he's found more power. The whole article is a garbled half-story, AFAICS on the face of it. If you want more triode power, use more valves, or bigger ones. Or use in pentode mode with a distributed load or with global feedback. It's not like there's a shortage of options. There's one actual design I have found in an audioxpress magazine advertised on the hi-fi collective site but no other example I can find. Cheers ian It's not even a credible retrofit...get more power out of your existing amp...because it requires a much greater drive signal and a bias voltage to suit. Ian- Hide quoted text - - Show quoted text -- Hide quoted text - - Show quoted text -- Hide quoted text - - Show quoted text - Interesting, Surely the screen would just become another anode and melt or sag and arc a bit and melt some more. The floating(but not for long) supply would be modulated in relation to ground with all the troubles that would entail ie the anode? having to drive the extra capacitance of the mains tranny winding, loads of noise etc(plus the relentless letting of smoke). Seems unlikely that it would work. Can you quote the article in audioexpress Ian? Cheers Matt If the battery symbol is the wrong way round its just a lower voltage screen supply, no probs. |
|
#7
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
On 12 May, 16:37, Andre Jute wrote:
On May 12, 12:11 pm, flipper wrote: On Tue, 12 May 2009 10:42:36 +0100, Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Yeah, I saw that one a while back but I'm not sure it's worth it because that takes a fully floating supply for each tube and it might be simpler to just do it the old fashioned way, double the output tubes. I'd say it is altogether a pretty grim idea, even if it works. Just for a start, the power supply is always the element making the nastiest class of residual noise on any amp's output. And let us not forget that the power supply, wherever it is fed into the tube, is in the signal circuit, something that these crazed power-hunters have a hard time understanding. So now, to get a little extra power, the amp requires no fewer than three power supplies. Yech! No wonder tha creator of the circuit does not publish noise figures. A whole lotta NFB loops might help but the recombinant spectra will be very nasty indeed, and the micro-view of all this on the waterfall of the sort of high-sensitivity speaker normally used with SE amps will be confused indeed. At the very least matters should be so arranged that the output rolls off very quickly under 50 or 60Hz, depending on where in the world you are. (Be even better to roll this amp off at 120Hz and bi-amp a woofer under it, but that is a kludge even further away from a quality sound.) It is this sort of overweening design that makes for "edgy" amps, the sound too brittle to listen to for long. There are stranger considerations in audio amps than are dreamt of in thy engineering, Horatio. In Scandinavia, in Europe, and in Australia those floating supplies are also ipso facto illegal unless they're double shielded and certified, either being a practical impossibility for the DIYer. People with their brains in gear, when they want louder music, first get more sensitive speakers, then make their SE amps with more powerful tubes, then use parallel tubes SE (cheaper anyway than this grim idea), then go to triodes in PP (operated strictly in Class A1, of course), then PP triode-linked pentodes in Class A1, and so on. Like I said, I'm talking about people with their brains in gear: people with their brains in gear don't perpetrate complicated audio electronics just because they can. Andre Jute Visit Jute on Amps at *http://www.audio-talk.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review I think you may have your idea of floating slightly wrong. Referenced above ground/earth is a no no. In Scandinavia, in Europe, and in Australia those floating supplies are also ipso facto illegal unless they're double shielded and certified, either being a practical impossibility for the DIYer Over here there are several situations that a floating supply is essential. The step down trannys used on building sites to provide 110V for tools etc. These all have a centre tapped secondary, with the tap at earth, so at any instant during a fault (all tools are also double insulated, or so you are led to believe) any extraneous metal will be 55V above earth. Swimming pools or bath rooms must use fully floating supplys in area one ie. where you get wet. This means that above a bath or shower we have to use low voltage lighting. This is a good idea full stop in a bathroom any way. Of course there are shaver sockets that have their own isolation tranny. The only other one I have seen is fully insulated labs. These have a strange set of rules that are bendable on acceptance of the IEE. Normally though all services entering such a room must be isolated, Ie insulated sections in all pipes, bunded floor with no Rawl bolts etc breaking the bund. The supply for such rooms is from an isolation transformer that has insulation to suit the purpose. The one I worked on was merely a room with test gear in to measure a plating plant, but it did make me wonder about impurities in the water that fed a solitary, floating above earth sink. Hmmmm. Cheers Matt. |
|
#8
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
On May 12, 11:19*pm, bigwig wrote:
On 12 May, 16:37, Andre Jute wrote: On May 12, 12:11 pm, flipper wrote: On Tue, 12 May 2009 10:42:36 +0100, Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Yeah, I saw that one a while back but I'm not sure it's worth it because that takes a fully floating supply for each tube and it might be simpler to just do it the old fashioned way, double the output tubes. I'd say it is altogether a pretty grim idea, even if it works. Just for a start, the power supply is always the element making the nastiest class of residual noise on any amp's output. And let us not forget that the power supply, wherever it is fed into the tube, is in the signal circuit, something that these crazed power-hunters have a hard time understanding. So now, to get a little extra power, the amp requires no fewer than three power supplies. Yech! No wonder tha creator of the circuit does not publish noise figures. A whole lotta NFB loops might help but the recombinant spectra will be very nasty indeed, and the micro-view of all this on the waterfall of the sort of high-sensitivity speaker normally used with SE amps will be confused indeed. At the very least matters should be so arranged that the output rolls off very quickly under 50 or 60Hz, depending on where in the world you are. (Be even better to roll this amp off at 120Hz and bi-amp a woofer under it, but that is a kludge even further away from a quality sound.) It is this sort of overweening design that makes for "edgy" amps, the sound too brittle to listen to for long. There are stranger considerations in audio amps than are dreamt of in thy engineering, Horatio. In Scandinavia, in Europe, and in Australia those floating supplies are also ipso facto illegal unless they're double shielded and certified, either being a practical impossibility for the DIYer. People with their brains in gear, when they want louder music, first get more sensitive speakers, then make their SE amps with more powerful tubes, then use parallel tubes SE (cheaper anyway than this grim idea), then go to triodes in PP (operated strictly in Class A1, of course), then PP triode-linked pentodes in Class A1, and so on. Like I said, I'm talking about people with their brains in gear: people with their brains in gear don't perpetrate complicated audio electronics just because they can. Andre Jute Visit Jute on Amps at *http://www.audio-talk.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review I think you may have your idea of floating slightly wrong. Referenced above ground/earth is a no no. In Scandinavia, in Europe, and in Australia those floating supplies are also ipso facto illegal unless they're double shielded and certified, either being a practical impossibility for the DIYer Over here there are several situations that a floating supply is essential. The step down trannys used on building sites to provide 110V for tools etc. These all have a centre tapped secondary, with the tap at earth, so at any instant during a fault (all tools are also double insulated, or so you are led to believe) any extraneous metal will be 55V above earth. Swimming pools or bath rooms must use fully floating supplys in area one ie. where you get wet. This means that above a bath or shower we have to use low voltage lighting. This is a good idea full stop in a bathroom any way. Of course there are shaver sockets that have their own isolation tranny. The only other one I have seen is fully insulated labs. These have a strange set of rules that are bendable on acceptance of the IEE. Normally though all services entering such a room must be isolated, Ie insulated sections in all pipes, bunded floor with no Rawl bolts etc breaking the bund. The supply for such rooms is from an isolation transformer that has insulation to suit the purpose. The one I worked on was merely a room with test gear in to measure a plating plant, but it did make me wonder about impurities in the water that fed a solitary, floating above earth sink. Hmmmm. Cheers Matt. Take a remedial English comprehension course, feller. You haven't contradicted what I said in the slightest. -- Andre Jute |
|
#9
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
On Tue, 12 May 2009 14:52:17 -0700 (PDT), bigwig
wrote: On 12 May, 18:26, Ian Bell wrote: Ian Iveson wrote: Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html ........snip! I believe that is the case - the screen is 100V above the plate. .......snip! Interesting, Surely the screen would just become another anode and melt or sag and arc a bit and melt some more. The floating(but not for long) supply would be modulated in relation to ground with all the troubles that would entail ie the anode? having to drive the extra capacitance of the mains tranny winding, loads of noise etc(plus the relentless letting of smoke). Seems unlikely that it would work. Can you quote the article in audioexpress Ian? Cheers Matt If the battery symbol is the wrong way round its just a lower voltage screen supply, no probs. Instead of using a separate supply, if you have a voltage (B+ boost) a few hundred volts greater than B+ (can do with doubler or other circuit tricks), run a resistor to the screen from B+boost, and put a hefty capacitor between screen and plate. Select the value of resistance that puts about 100v DC between screen and plate. The screen current is normally 1/5 - 1/20 the plate current. In order to raise the screen voltage above the plate by 100V, you are going to need more screen current, and the screen current will be a higher proportion than before. The screen current is roughly proportional to the plate current (if away from the "knee" of plate characteristic), so during signal, if the tube runs fairly linearly, the 100V should stay on the screen (where the screen current increases and drops the voltage is balanced by the screen current decrease). If you are a purist, or suspect a lot of unlinearity, you could place a 100V zener across the plate-screen, but you'd need to draw more current. That extra current ends up in the plate circuit, probably not a good idea. The circuit that I proposed will end up being part of the plate load, and will draw off some off the output power. If your B+boost is quite high, the resistor mentioned above would be quite large, and would not sap significant audio power. The advantage using this circuit is that you don't need to worry about isolation or extra power supplies. The down side is that you need some regulation of the 100V. I'm assuming that a decent capacitor (10 - 50 uF ??) should hold down the screen-plate voltage. That capacitor should also reduce the effects of hum on the B+boost supply since it effectively clamps the screen voltage (sensitive to variations) to the plate, except at DC and very low frequencies. Is this a silly idea? Paul G |
|
#10
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Ian Bell wrote: flipper wrote: On Tue, 12 May 2009 10:42:36 +0100, Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Yeah, I saw that one a while back but I'm not sure it's worth it because that takes a fully floating supply for each tube and it might be simpler to just do it the old fashioned way, double the output tubes. I presume you mean double as in parallel rather than push pull? Given the cost of the extra PSU then there's probably not much in it cost wise. Cheers ian The cost of the floating 100V PSU does not have to be high because the screen current involved is low. Therefore a small 7VA mains tranny with say 240V:12V may be used running backwards off the 6.3V heater voltages. This gives 126Vac which then can be rectified and shunt regged to make a floating 100V supply. But the page at http://www.edn.com/article/CA302240.html does not show detailed test results from a breadboard circuit. Does it work at all? It looks like all being too much free cake at too low a price. I've never tried Boosted Triode, which does seem a very nice idea *if* it works as claimed. But I have tried UL with a zener diode feed from the UL tap to the screen, and with a nice big 470uF bypass cap. The simple zener and 470uF work as a floating voltage supply. I concluded the the UL character didn't get better, but got worse with regard for available Va swing, more like triode, and I didn't investigate any further. Boosted triode could be done the same way with a zener and cap from anode to screen. Patrick Turner. |
|
#11
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Ian Iveson wrote: Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Those load lines aren't what the author says they are and he doesn't say how he measured power. Is it a joke? Maybe it works, but in reality not for long. In simulation you can't melt the screen. Unless he, or I, have got the battery symbol the wrong way round, it appears that the screen voltage is set at 100V greater than the anode. Nope, he has the battery the right way. I assumed when I looked at figure3 that the battery made the screen a lower Vdc than the anode, which then means you have to reduce G1 bias to maintian the same Ia. The triode then becomes the opposite of boosted when the grid line for 0V bias moves to the right. Looking closer, the raising of Eg2 by 100Vdc above Ea means that Ia is increased for the same Ea, but if Eg1 is then increased from his -32V to -44V, then the G1 can swing much further without running into grid current. For Boosted Triode, the normal method used for over 50 years has been so simply drive the triodes in class A2, or more usually class AB2 using a CF stage direct coupled to the output grids. This will much increase the class AB1 from a pair of 6L6 from say 16W to 32W. I know because I have tried this, and all without a pair of floating power supplies and without turning screen grid wires red. Some tube like the EL34 don't like being driven into class A2 or AB2. The ensuing screen meltdown could be avoided by increasing the anode voltage by the same amount, i.e. by simply increasing the HT and discarding the extra screen supplies. Perhaps that takes the valve out of it's SOA? Not so far as does setting Vs Va, surely? Nope, your'e guessing. In the introduction he likens the anode to a JFET drain. That analogy is only true at higher voltages, or at low voltages only if the screen voltage is very low. Otherwise the screen becomes the drain. It looks to me like the screen would be passing most of the current in that region where he thinks he's found more power. The whole article is a garbled half-story, AFAICS on the face of it. If you want more triode power, use more valves, or bigger ones. Or use in pentode mode with a distributed load or with global feedback. It's not like there's a shortage of options. It's not even a credible retrofit...get more power out of your existing amp...because it requires a much greater drive signal and a bias voltage to suit. There are easier and better ways to extract linear power from a multigrid. My schematic for a 32W amp using ONE 13E1 is a classic example. See the schematic at http://www.turneraudio.com.au/monobl...-13ei-cfb.html Its a case of local cathode feedback and the screen voltage is well below the anode voltage which complies best with the data on the tube for best PO and minimum wasted screen current. Patrick Turner. Ian |
|
#12
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Andre Jute wrote: On May 12, 12:11 pm, flipper wrote: On Tue, 12 May 2009 10:42:36 +0100, Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Yeah, I saw that one a while back but I'm not sure it's worth it because that takes a fully floating supply for each tube and it might be simpler to just do it the old fashioned way, double the output tubes. I'd say it is altogether a pretty grim idea, even if it works. Just for a start, the power supply is always the element making the nastiest class of residual noise on any amp's output. And let us not forget that the power supply, wherever it is fed into the tube, is in the signal circuit, something that these crazed power-hunters have a hard time understanding. Yes Andre, but if the tranny for the floating supply has large leakage inductance and the capacitance between adjacent primary coils is very low the noise injection is utterly negligible. This can be achieved best with a C-core tranny with low voltage primary running at 6.3V stepping up the voltage magnetically on the sec to say 120V, and the two windings on opposite sides of the single O core and with a an electrostatic screen wound around the primary, which is taken to 0V. I reckon an extra tube would be easier. But not if you were using a 13E1. Too expensive. But that tube doesn't like a high Eg2 which is higher than Ea. So now, to get a little extra power, the amp requires no fewer than three power supplies. Yech! No wonder tha creator of the circuit does not publish noise figures. A whole lotta NFB loops might help but the recombinant spectra will be very nasty indeed, and the micro-view of all this on the waterfall of the sort of high-sensitivity speaker normally used with SE amps will be confused indeed. At the very least matters should be so arranged that the output rolls off very quickly under 50 or 60Hz, depending on where in the world you are. (Be even better to roll this amp off at 120Hz and bi-amp a woofer under it, but that is a kludge even further away from a quality sound.) It is this sort of overweening design that makes for "edgy" amps, the sound too brittle to listen to for long. There are stranger considerations in audio amps than are dreamt of in thy engineering, Horatio. In Scandinavia, in Europe, and in Australia those floating supplies are also ipso facto illegal unless they're double shielded and certified, either being a practical impossibility for the DIYer. People with their brains in gear, when they want louder music, first get more sensitive speakers, then make their SE amps with more powerful tubes, then use parallel tubes SE (cheaper anyway than this grim idea), then go to triodes in PP (operated strictly in Class A1, of course), then PP triode-linked pentodes in Class A1, and so on. Like I said, I'm talking about people with their brains in gear: people with their brains in gear don't perpetrate complicated audio electronics just because they can. I just use CFB windings on the OPT. Quad did it very successfully about 60 years ago in Quad-II. SE versions of the Quad-II idea work well to make a KT66 ( or 6L6 ) give the same power as in pure beam tetrode mode and with a similar effective Ra as the KT66 in triode. If one drew the effective Ra lines for the KT66 with 10% CFB and a fixed Eg2 supply, I bet they'd look very similar if not more linear than the Boosted Triode Ra lines. So I would suggest THD/IMD with SE CFB would be less than real triode at all PO levels up to clipping. The external loop FB around the local OPT and acting between grid and cathode and between screen and cathode are linear delivery mechanisms for the FB. In the triode connection or real triodes such as a 300B, the internal NFB is delivered via a non linear network because gm of the control grid and screen are not the same for changing levels of voltages. Patrick Turner. Andre Jute Visit Jute on Amps at http://www.audio-talk.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review |
|
#13
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Patrick Turner wrote:
This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Those load lines aren't what the author says they are and he doesn't say how he measured power. Is it a joke? Maybe it works, but in reality not for long. In simulation you can't melt the screen. Unless he, or I, have got the battery symbol the wrong way round, it appears that the screen voltage is set at 100V greater than the anode. Nope, he has the battery the right way. I assumed when I looked at figure3 that the battery made the screen a lower Vdc than the anode, which then means you have to reduce G1 bias to maintian the same Ia. The triode then becomes the opposite of boosted when the grid line for 0V bias moves to the right. Looking closer, the raising of Eg2 by 100Vdc above Ea means that Ia is increased for the same Ea, but if Eg1 is then increased from his -32V to -44V, then the G1 can swing much further without running into grid current. Obviously, and equally true were HT simply raised by the same 100V. For Boosted Triode, the normal method used for over 50 years has been so simply drive the triodes in class A2, or more usually class AB2 using a CF stage direct coupled to the output grids. This will much increase the class AB1 from a pair of 6L6 from say 16W to 32W. I know because I have tried this, and all without a pair of floating power supplies and without turning screen grid wires red. Different thing altogether. Some tube like the EL34 don't like being driven into class A2 or AB2. The ensuing screen meltdown could be avoided by increasing the anode voltage by the same amount, i.e. by simply increasing the HT and discarding the extra screen supplies. Perhaps that takes the valve out of it's SOA? Not so far as does setting Vs Va, surely? Nope, your'e guessing. Nope to what? What do you think I'm guessing? In the introduction he likens the anode to a JFET drain. That analogy is only true at higher voltages, or at low voltages only if the screen voltage is very low. Otherwise the screen becomes the drain. It looks to me like the screen would be passing most of the current in that region where he thinks he's found more power. The whole article is a garbled half-story, AFAICS on the face of it. If you want more triode power, use more valves, or bigger ones. Or use in pentode mode with a distributed load or with global feedback. It's not like there's a shortage of options. It's not even a credible retrofit...get more power out of your existing amp...because it requires a much greater drive signal and a bias voltage to suit. There are easier and better ways to extract linear power from a multigrid. My schematic for a 32W amp using ONE 13E1 is a classic example. See the schematic at http://www.turneraudio.com.au/monobl...-13ei-cfb.html By "classic example" do you mean you copied it? From where? Why not link to the original, or at least credit it? Ian |
|
#14
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
bigwig wrote:
On 12 May, 18:26, Ian Bell wrote: Ian Iveson wrote: Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Those load lines aren't what the author says they are and he doesn't say how he measured power. Is it a joke? Maybe it works, but in reality not for long. In simulation you can't melt the screen. Unless he, or I, have got the battery symbol the wrong way round, it appears that the screen voltage is set at 100V greater than the anode. I believe that is the case - the screen is 100V above the plate. The ensuing screen meltdown could be avoided by increasing the anode voltage by the same amount, i.e. by simply increasing the HT and discarding the extra screen supplies. Perhaps that takes the valve out of it's SOA? Not so far as does setting Vs Va, surely? In the introduction he likens the anode to a JFET drain. That analogy is only true at higher voltages, or at low voltages only if the screen voltage is very low. Otherwise the screen becomes the drain. It looks to me like the screen would be passing most of the current in that region where he thinks he's found more power. The whole article is a garbled half-story, AFAICS on the face of it. If you want more triode power, use more valves, or bigger ones. Or use in pentode mode with a distributed load or with global feedback. It's not like there's a shortage of options. There's one actual design I have found in an audioxpress magazine advertised on the hi-fi collective site but no other example I can find. Cheers ian It's not even a credible retrofit...get more power out of your existing amp...because it requires a much greater drive signal and a bias voltage to suit. Ian- Hide quoted text - - Show quoted text -- Hide quoted text - - Show quoted text -- Hide quoted text - - Show quoted text - Interesting, Surely the screen would just become another anode and melt or sag and arc a bit and melt some more. The floating(but not for long) supply would be modulated in relation to ground with all the troubles that would entail ie the anode? having to drive the extra capacitance of the mains tranny winding, loads of noise etc(plus the relentless letting of smoke). Seems unlikely that it would work. Can you quote the article in audioexpress Ian? If memory serves it was December 2003. Someone an another group says he has that issue and has promised to scan and post the article. I'll make it available here when I get it. Cheers Ian Cheers Matt If the battery symbol is the wrong way round its just a lower voltage screen supply, no probs. |
|
#15
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
On May 13, 11:44*am, Patrick Turner wrote:
Andre Jute wrote: On May 12, 12:11 pm, flipper wrote: On Tue, 12 May 2009 10:42:36 +0100, Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Yeah, I saw that one a while back but I'm not sure it's worth it because that takes a fully floating supply for each tube and it might be simpler to just do it the old fashioned way, double the output tubes. I'd say it is altogether a pretty grim idea, even if it works. Just for a start, the power supply is always the element making the nastiest class of residual noise on any amp's output. And let us not forget that the power supply, wherever it is fed into the tube, is in the signal circuit, something that these crazed power-hunters have a hard time understanding. Yes Andre, but if the tranny for the floating supply has large leakage inductance and the capacitance between adjacent primary coils is very low the noise injection is utterly negligible. This can be achieved best with a C-core tranny with low voltage primary running at 6.3V stepping up the voltage magnetically on the sec to say 120V, and the two windings on opposite sides of the single O core and with a an electrostatic screen wound around the primary, which is taken to 0V. Lovelly. Now we don't only have two extra supplies, they have to be C- cores -- why not just make themmy favourite Swedish cut C-cores and get the best? And to be certain the circuit is noiseless, let's drive the 6V side with a bank of those farmhouse 6V batteries... And let's not worry about the double insulation on the 130V secondary because we can do with a holiday paid for by the taxpayers in the graybar hotel chain. This entire circuit is a bodge for a non-existent problem, and then the bodge starts creating a problem of its own (that's the nature of a bodge) and other bodges are stacked on top of it. The ooh-aah bird comes to mind. I reckon an extra tube would be easier. Exactly. But not if you were using a 13E1. Too expensive. But that tube doesn't like a high Eg2 which is higher than Ea. What's the difference between unobtanium tubes, self-flagellation and sado-masochism? There is always a simple, straightforward solution. In the end the simple, straighforward solution always costs less money and frustration than the complicated bodge. Andre Jute Visit Jute on Amps at http://www.audio-talk.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review So now, to get a little extra power, the amp requires no fewer than three power supplies. Yech! No wonder tha creator of the circuit does not publish noise figures. A whole lotta NFB loops might help but the recombinant spectra will be very nasty indeed, and the micro-view of all this on the waterfall of the sort of high-sensitivity speaker normally used with SE amps will be confused indeed. At the very least matters should be so arranged that the output rolls off very quickly under 50 or 60Hz, depending on where in the world you are. (Be even better to roll this amp off at 120Hz and bi-amp a woofer under it, but that is a kludge even further away from a quality sound.) It is this sort of overweening design that makes for "edgy" amps, the sound too brittle to listen to for long. There are stranger considerations in audio amps than are dreamt of in thy engineering, Horatio. In Scandinavia, in Europe, and in Australia those floating supplies are also ipso facto illegal unless they're double shielded and certified, either being a practical impossibility for the DIYer. People with their brains in gear, when they want louder music, first get more sensitive speakers, then make their SE amps with more powerful tubes, then use parallel tubes SE (cheaper anyway than this grim idea), then go to triodes in PP (operated strictly in Class A1, of course), then PP triode-linked pentodes in Class A1, and so on. Like I said, I'm talking about people with their brains in gear: people with their brains in gear don't perpetrate complicated audio electronics just because they can. I just use CFB windings on the OPT. Quad did it very successfully about 60 years ago in Quad-II. SE versions of the Quad-II idea work well to make a KT66 ( or 6L6 ) give the same power as in pure beam tetrode mode and with a similar effective Ra as the KT66 in triode. If one drew the effective Ra lines for the KT66 with 10% CFB and a fixed Eg2 supply, I bet they'd look very similar if not more linear than the Boosted Triode Ra lines. So I would suggest THD/IMD with SE CFB would be less than real triode at all PO levels up to clipping. The external loop FB around the local OPT and acting between grid and cathode and between screen and cathode are linear delivery mechanisms for the FB. In the triode connection or real triodes such as a 300B, the internal NFB is delivered via a non linear network because gm of the control grid and screen are not the same for changing levels of voltages. Patrick Turner. Andre Jute Visit Jute on Amps at *http://www.audio-talk.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review |
|
#16
Posted to rec.audio.tubes
|
|||
|
|||
|
Bloated Triode, was Boosted Triode
In all those technical points, I forgot to make the moral point:
And the worst thing about these bodges is that we sit here making an intellectual sport of bodges upon bodges while the world goes to hell in a handbasket. Slainte! --AJ On May 14, 4:19*am, Andre Jute wrote: On May 13, 11:44*am, Patrick Turner wrote: Andre Jute wrote: On May 12, 12:11 pm, flipper wrote: On Tue, 12 May 2009 10:42:36 +0100, Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Yeah, I saw that one a while back but I'm not sure it's worth it because that takes a fully floating supply for each tube and it might be simpler to just do it the old fashioned way, double the output tubes. I'd say it is altogether a pretty grim idea, even if it works. Just for a start, the power supply is always the element making the nastiest class of residual noise on any amp's output. And let us not forget that the power supply, wherever it is fed into the tube, is in the signal circuit, something that these crazed power-hunters have a hard time understanding. Yes Andre, but if the tranny for the floating supply has large leakage inductance and the capacitance between adjacent primary coils is very low the noise injection is utterly negligible. This can be achieved best with a C-core tranny with low voltage primary running at 6.3V stepping up the voltage magnetically on the sec to say 120V, and the two windings on opposite sides of the single O core and with a an electrostatic screen wound around the primary, which is taken to 0V. Lovelly. Now we don't only have two extra supplies, they have to be C- cores -- why not just make themmy favourite Swedish cut C-cores and get the best? And to be certain the circuit is noiseless, let's drive the 6V side with a bank of those farmhouse 6V batteries... And let's not worry about the double insulation on the 130V secondary because we can do with a holiday paid for by the taxpayers in the graybar hotel chain. This entire circuit is a bodge for a non-existent problem, and then the bodge starts creating a problem of its own (that's the nature of a bodge) and other bodges are stacked on top of it. The ooh-aah bird comes to mind. I reckon an extra tube would be easier. Exactly. But not if you were using a 13E1. Too expensive. But that tube doesn't like a high Eg2 which is higher than Ea. What's the difference between unobtanium tubes, self-flagellation and sado-masochism? There is always a simple, straightforward solution. In the end the simple, straighforward solution always costs less money and frustration than the complicated bodge. Andre Jute Visit Jute on Amps at *http://www.audio-talk.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review So now, to get a little extra power, the amp requires no fewer than three power supplies. Yech! No wonder tha creator of the circuit does not publish noise figures. A whole lotta NFB loops might help but the recombinant spectra will be very nasty indeed, and the micro-view of all this on the waterfall of the sort of high-sensitivity speaker normally used with SE amps will be confused indeed. At the very least matters should be so arranged that the output rolls off very quickly under 50 or 60Hz, depending on where in the world you are. (Be even better to roll this amp off at 120Hz and bi-amp a woofer under it, but that is a kludge even further away from a quality sound.) It is this sort of overweening design that makes for "edgy" amps, the sound too brittle to listen to for long. There are stranger considerations in audio amps than are dreamt of in thy engineering, Horatio. In Scandinavia, in Europe, and in Australia those floating supplies are also ipso facto illegal unless they're double shielded and certified, either being a practical impossibility for the DIYer. People with their brains in gear, when they want louder music, first get more sensitive speakers, then make their SE amps with more powerful tubes, then use parallel tubes SE (cheaper anyway than this grim idea), then go to triodes in PP (operated strictly in Class A1, of course), then PP triode-linked pentodes in Class A1, and so on. Like I said, I'm talking about people with their brains in gear: people with their brains in gear don't perpetrate complicated audio electronics just because they can. I just use CFB windings on the OPT. Quad did it very successfully about 60 years ago in Quad-II. SE versions of the Quad-II idea work well to make a KT66 ( or 6L6 ) give the same power as in pure beam tetrode mode and with a similar effective Ra as the KT66 in triode. If one drew the effective Ra lines for the KT66 with 10% CFB and a fixed Eg2 supply, I bet they'd look very similar if not more linear than the Boosted Triode Ra lines. So I would suggest THD/IMD with SE CFB would be less than real triode at all PO levels up to clipping. The external loop FB around the local OPT and acting between grid and cathode and between screen and cathode are linear delivery mechanisms for the FB. In the triode connection or real triodes such as a 300B, the internal NFB is delivered via a non linear network because gm of the control grid and screen are not the same for changing levels of voltages. Patrick Turner. Andre Jute Visit Jute on Amps at *http://www.audio-talk.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review |
|
#17
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Ian Iveson wrote: Patrick Turner wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Those load lines aren't what the author says they are and he doesn't say how he measured power. Is it a joke? Maybe it works, but in reality not for long. In simulation you can't melt the screen. Unless he, or I, have got the battery symbol the wrong way round, it appears that the screen voltage is set at 100V greater than the anode. Nope, he has the battery the right way. I assumed when I looked at figure3 that the battery made the screen a lower Vdc than the anode, which then means you have to reduce G1 bias to maintian the same Ia. The triode then becomes the opposite of boosted when the grid line for 0V bias moves to the right. Looking closer, the raising of Eg2 by 100Vdc above Ea means that Ia is increased for the same Ea, but if Eg1 is then increased from his -32V to -44V, then the G1 can swing much further without running into grid current. Obviously, and equally true were HT simply raised by the same 100V. Nope. The B+ must be left alone and the screen voltage raised 100V above Ea. For Boosted Triode, the normal method used for over 50 years has been so simply drive the triodes in class A2, or more usually class AB2 using a CF stage direct coupled to the output grids. This will much increase the class AB1 from a pair of 6L6 from say 16W to 32W. I know because I have tried this, and all without a pair of floating power supplies and without turning screen grid wires red. Different thing altogether. Some tube like the EL34 don't like being driven into class A2 or AB2. The ensuing screen meltdown could be avoided by increasing the anode voltage by the same amount, i.e. by simply increasing the HT and discarding the extra screen supplies. Perhaps that takes the valve out of it's SOA? Not so far as does setting Vs Va, surely? Nope, your'e guessing. Nope to what? What do you think I'm guessing? Never mind, you have no ability to criticise yourself. In the introduction he likens the anode to a JFET drain. That analogy is only true at higher voltages, or at low voltages only if the screen voltage is very low. Otherwise the screen becomes the drain. It looks to me like the screen would be passing most of the current in that region where he thinks he's found more power. The whole article is a garbled half-story, AFAICS on the face of it. If you want more triode power, use more valves, or bigger ones. Or use in pentode mode with a distributed load or with global feedback. It's not like there's a shortage of options. It's not even a credible retrofit...get more power out of your existing amp...because it requires a much greater drive signal and a bias voltage to suit. There are easier and better ways to extract linear power from a multigrid. My schematic for a 32W amp using ONE 13E1 is a classic example. See the schematic at http://www.turneraudio.com.au/monobl...-13ei-cfb.html By "classic example" do you mean you copied it? From where? Why not link to the original, or at least credit it? AFAIK, my addaption of CFB in the CFB32W amp with 13E1 is unique, and classic. Probably there were others who used CFB OPT windings well before Walker moved in to exploit the classic idea. So the original inventor of CFB may not be ever be indentified. He's probably dead. Patrick Turner. Ian |
|
#18
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Andre Jute wrote: On May 13, 11:44 am, Patrick Turner wrote: Andre Jute wrote: On May 12, 12:11 pm, flipper wrote: On Tue, 12 May 2009 10:42:36 +0100, Ian Bell wrote: This seems like an interesting idea to get more power out of a triode. http://www.edn.com/article/CA302240.html Cheers Ian Yeah, I saw that one a while back but I'm not sure it's worth it because that takes a fully floating supply for each tube and it might be simpler to just do it the old fashioned way, double the output tubes. I'd say it is altogether a pretty grim idea, even if it works. Just for a start, the power supply is always the element making the nastiest class of residual noise on any amp's output. And let us not forget that the power supply, wherever it is fed into the tube, is in the signal circuit, something that these crazed power-hunters have a hard time understanding. Yes Andre, but if the tranny for the floating supply has large leakage inductance and the capacitance between adjacent primary coils is very low the noise injection is utterly negligible. This can be achieved best with a C-core tranny with low voltage primary running at 6.3V stepping up the voltage magnetically on the sec to say 120V, and the two windings on opposite sides of the single O core and with a an electrostatic screen wound around the primary, which is taken to 0V. Lovelly. Now we don't only have two extra supplies, they have to be C- cores -- why not just make themmy favourite Swedish cut C-cores and get the best? And to be certain the circuit is noiseless, let's drive the 6V side with a bank of those farmhouse 6V batteries... And let's not worry about the double insulation on the 130V secondary because we can do with a holiday paid for by the taxpayers in the graybar hotel chain. This entire circuit is a bodge for a non-existent problem, and then the bodge starts creating a problem of its own (that's the nature of a bodge) and other bodges are stacked on top of it. The ooh-aah bird comes to mind. I reckon an extra tube would be easier. Exactly. But not if you were using a 13E1. Too expensive. But that tube doesn't like a high Eg2 which is higher than Ea. What's the difference between unobtanium tubes, self-flagellation and sado-masochism? Probably not a lot. But some audiophiles like to eat a breakfast of unobsadoflagellate wholemeal or else they feel they are missing out on something. There is always a simple, straightforward solution. In the end the simple, straighforward solution always costs less money and frustration than the complicated bodge. Of course. Patrick Turner. |
|
#19
Posted to rec.audio.tubes
|
|||
|
|||
|
Bloated Triode, was Boosted Triode
Andre Jute wrote: In all those technical points, I forgot to make the moral point: And the worst thing about these bodges is that we sit here making an intellectual sport of bodges upon bodges while the world goes to hell in a handbasket. Ah, but sometimes its important to know the most elegant way to lose a war, or ruin an economy, or build an bad aeroplane. Despite taking part in these bodgying discussions about Triodes With Flatulance, I am not drawn at all to trying the idea out. But say you had 4 suitable beam or pentode tubes such as KT88, etc, and you wanted to run in triode with Ea at only 300V and Ia at 70mA, the Pda would be a comfortable 21 watts, and in normal triode you'd maybe get 5W per tube, or 20W from the 4. Because the KT88 screen Pdiss ratings are what they are, you could probably have Eg2 at 400V, and therefore get a much wider anode Vswing over a higher value load, therefore reducing THD and raising PO before clipping to maybe 30W, and all without causing pain in the screens. There would need to be only ONE 100V floating PS, or if not a floating PS, then you could have a choke feed from a grounded 400V supply to the screens and a cap between anode to screens to convey the anode signal to the screen to get true triode performance. The screen Pdiss is much less in normal triode strapped operation than in beam tet operation where the Eg2 is fixed. Methinks the one PSU HT winding of 312Vrms could be used to generate +350V from a CRC filtered supply for the screen feed, then you could have a choke input supply from the same HT winding to give a lower B+ for the anode supply, 250V in fact, so all you'd be up for are a couple of extra moderate size filter caps, a resistor for the CRC filter, and a choke of 20H capable of about 30mAdc, and a cap between anode and screen, probably a 250V rated 40uF motor start. Where would be the noise injection you winged about now? Now with CFB or plain beam tube format used in SE class A amps, the Ea can be anywhere between 250V and 400V. Best Ea I have found for most big beams and pents is about 350V only, but say you make it only 250V, and if Pda was 20W, then Ia would be 80mA. Ig2 is somewhat proportional to Ia, so maybe Ig2 = 8mA even if Eg2 was at +350V. The load used for the BT config will be the same as for the same tube in beam or pentode, and in this case RL = 0.9 x Ea / Ia, or 2.8k per tube in this case, giving a load of 700 ohms for the 4 x KT88 in BT. The peak Ea swing without G1 current would then become 0.9 x 250, and PO = 36W, with anode power efficiency = 45%, and from a total of 80W of anode diss. Screen diss would be 11W total, so PSU must make a total of 91W. Normal healthy triode efficiency for KT88 is about where Ea = Eg2 = +450V, and Ia + Ig2 = 44mA. Eficiency is 33% max, so to get 36W you'd need 6 x KT88 to comfortably dissipate 18W each, or 109W total. The load will be about 1,400 ohms for the 36W and the OPT a bit more difficult to wind than the one wanted for RL = 700 ohms, although the normal triode total Ia = 264mA versus 352mA for the BT. I'm presently building a pair of mono amps for 6 x KT88/6550/6L6GC,6CA7 but I am using SE CFB and a fixed voltage supply to the screens so Eg2 will be 250V to 300V. Ia total = 400mA, and Ea = 360V, and total Pda = 144W and I should get up to 64W. There are obviously more ways than one to compose a symphony. If Lizst could compose somewhat awkward piano music which propelled or attracted rational men and women to actually play it, then I see no reason why anyone might not dabble with BT, under certain circumstances. Patrick Turner. Slainte! --AJ |
|
#20
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Patrick Turner wrote:
Looking closer, the raising of Eg2 by 100Vdc above Ea means that Ia is increased for the same Ea, but if Eg1 is then increased from his -32V to -44V, then the G1 can swing much further without running into grid current. Obviously, and equally true were HT simply raised by the same 100V. Nope. Nope what? Nope its not obvious, or nope it would not be equally true if HT were raised by the same 100V? If the latter, then that is the crux of the biscuit, I feel. Are you suggesting that, if screen and anode are both increased by 100V, the appropriate bias voltage would be less than 44V? That would lead to the interesting conclusion that if you apply 500V to both screen and anode and a certain current results, then as you reduce Va, current will increase. Or are you just quibbling with "equally"? My view is that changing Va wouldn't have much effect because screen mu is so dominant. That's the main reason why I wondered if the design is a joke: it would be much simpler to increase Va and Vs by the same amount, than to increase Vs alone, obviously. The B+ must be left alone and the screen voltage raised 100V above Ea. For Boosted Triode, the normal method used for over 50 years has been so simply drive the triodes in class A2, or more usually class AB2 using a CF stage direct coupled to the output grids. This will much increase the class AB1 from a pair of 6L6 from say 16W to 32W. I know because I have tried this, and all without a pair of floating power supplies and without turning screen grid wires red. Different thing altogether. Some tube like the EL34 don't like being driven into class A2 or AB2. The ensuing screen meltdown could be avoided by increasing the anode voltage by the same amount, i.e. by simply increasing the HT and discarding the extra screen supplies. Perhaps that takes the valve out of it's SOA? Not so far as does setting Vs Va, surely? Nope, your'e guessing. Nope to what? What do you think I'm guessing? Never mind, you have no ability to criticise yourself. I have displayed that ability infinitely more times than you have, cheeky, but if that's what you want and you tell me what you want me to criticise myself for, I'll have a shot at it.. What did you think I was guessing? I asked two questions, and was hoping for at least one answer. If you were to answer the second question with "No, because the valve never gets near enough to, or inboard of, its knee.", or words to that effect, then I would be quite happy. Happier still if you can direct us to a datasheet that demonstrates that fact. I could do that myself but I don't want to steal your thunder. In the introduction he likens the anode to a JFET drain. That analogy is only true at higher voltages, or at low voltages only if the screen voltage is very low. Otherwise the screen becomes the drain. It looks to me like the screen would be passing most of the current in that region where he thinks he's found more power. The whole article is a garbled half-story, AFAICS on the face of it. If you want more triode power, use more valves, or bigger ones. Or use in pentode mode with a distributed load or with global feedback. It's not like there's a shortage of options. It's not even a credible retrofit...get more power out of your existing amp...because it requires a much greater drive signal and a bias voltage to suit. There are easier and better ways to extract linear power from a multigrid. My schematic for a 32W amp using ONE 13E1 is a classic example. See the schematic at http://www.turneraudio.com.au/monobl...-13ei-cfb.html By "classic example" do you mean you copied it? From where? Why not link to the original, or at least credit it? AFAIK, my addaption of CFB in the CFB32W amp with 13E1 is unique, and classic. How can it be unique and classic at the same time? Probably there were others who used CFB OPT windings well before Walker "Distributed load" is a better description IMO. moved in to exploit the classic idea. I can see the meanings here of "belonging to a class", "regognised as definitive", "from the classic era"; so "classic" seems apposite. So the original inventor of CFB may not be ever be indentified. He's probably dead. You mean "distributed load", presumably. There are umpteen common ways of applying cathode feedback. Ian |
|
#21
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
bigwig wrote:
...Interesting, Surely the screen would just become another anode and melt or sag and arc a bit and melt some more. ... But check the second application example, for "line out" he http://tdsl.duncanamps.com/show.php?des=6L6GT ...The floating(but not for long) supply would be modulated in relation to ground with all the troubles that would entail ie the anode? having to drive the extra capacitance of the mains tranny winding, loads of noise etc(plus the relentless letting of smoke). Seems unlikely that it would work. Can you quote the article in audioexpress Ian? Cheers Matt If the battery symbol is the wrong way round its just a lower voltage screen supply, no probs. But then you have to wonder where the extra power comes from! Like I did for a quite some time, until I looked up what that pointy thing is in the transistor symbol, and wondered if he'd got the PS the wrong way up. Ian |
|
#22
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Paul G wrote:
Instead of using a separate supply, if you have a voltage (B+ boost) a few hundred volts greater than B+ (can do with doubler or other circuit tricks), run a resistor to the screen from B+boost, and put a hefty capacitor between screen and plate. Select the value of resistance that puts about 100v DC between screen and plate. The screen current is normally 1/5 - 1/20 the plate current. In order to raise the screen voltage above the plate by 100V, you are going to need more screen current, and the screen current will be a higher proportion than before. The screen current is roughly proportional to the plate current (if away from the "knee" of plate characteristic), ... A moot point. How near to the knee does the original link's loadline go in "boosted triode" mode? Otherwise your idea seems reasonable to me. ...so during signal, if the tube runs fairly linearly, the 100V should stay on the screen (where the screen current increases and drops the voltage is balanced by the screen current decrease). If you are a purist, or suspect a lot of unlinearity, you could place a 100V zener across the plate-screen, but you'd need to draw more current. That extra current ends up in the plate circuit, probably not a good idea. The circuit that I proposed will end up being part of the plate load, and will draw off some off the output power. If your B+boost is quite high, the resistor mentioned above would be quite large, and would not sap significant audio power. The advantage using this circuit is that you don't need to worry about isolation or extra power supplies. The down side is that you need some regulation of the 100V. I'm assuming that a decent capacitor (10 - 50 uF ??) should hold down the screen-plate voltage. That capacitor should also reduce the effects of hum on the B+boost supply since it effectively clamps the screen voltage (sensitive to variations) to the plate, except at DC and very low frequencies. Is this a silly idea? I don't think it adds any silliness to the original idea :-) Another possibility would be a screen winding on the OPT of same turns as for the anode, instead of the resistor and cap? But why not simply increase the HT by 100V? I really don't see the point of all this jiggerypokery. Let's say the circuit was derived the other way round, and the chap began with a triode with Va 500V, and he proposed his circuit as a way of reducing Va. What for? Would it be more linear? Would the increase in HT efficiency be more than that lost by the extra supplies? How long would be payback time be, considering the cost of components and extra construction? Ian |
|
#23
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
flipper wrote:
Instead of using a separate supply, if you have a voltage (B+ boost) a few hundred volts greater than B+ (can do with doubler or other circuit tricks), run a resistor to the screen from B+boost, and put a hefty capacitor between screen and plate. Select the value of resistance that puts about 100v DC between screen and plate. The screen current is normally 1/5 - 1/20 the plate current. In order to raise the screen voltage above the plate by 100V, you are going to need more screen current, and the screen current will be a higher proportion than before. The screen current is roughly proportional to the plate current (if away from the "knee" of plate characteristic), ... A moot point. How near to the knee does the original link's loadline go in "boosted triode" mode? Otherwise your idea seems reasonable to me. ...so during signal, if the tube runs fairly linearly, the 100V should stay on the screen (where the screen current increases and drops the voltage is balanced by the screen current decrease). If you are a purist, or suspect a lot of unlinearity, you could place a 100V zener across the plate-screen, but you'd need to draw more current. That extra current ends up in the plate circuit, probably not a good idea. The circuit that I proposed will end up being part of the plate load, and will draw off some off the output power. If your B+boost is quite high, the resistor mentioned above would be quite large, and would not sap significant audio power. The advantage using this circuit is that you don't need to worry about isolation or extra power supplies. The down side is that you need some regulation of the 100V. I'm assuming that a decent capacitor (10 - 50 uF ??) should hold down the screen-plate voltage. That capacitor should also reduce the effects of hum on the B+boost supply since it effectively clamps the screen voltage (sensitive to variations) to the plate, except at DC and very low frequencies. Is this a silly idea? I don't think it adds any silliness to the original idea :-) Another possibility would be a screen winding on the OPT of same turns as for the anode, instead of the resistor and cap? But why not simply increase the HT by 100V? I really don't see the point of all this jiggerypokery. Because you'd red plate. His example is a class A amplifier and you'd need the 75mA idle at the higher B+. The plate can't handle it. Good, thanks. That's my first question dealt with, assuming you are right about the anode's max power. The chap needs 6550 perhaps. It's been a while since they've been mentioned AFAIR. Hooray for the 6550. Or KT88 mebs. Now what I'm wondering is where all that energy goes. The electrons, to put it crudely, arrive at the anode having lost 100Vs-worth of energy since they passed the screen. What happens to it? Let's say the circuit was derived the other way round, and the chap began with a triode with Va 500V, and he proposed his circuit as a way of reducing Va. What for? Would it be more linear? Would the increase in HT efficiency be more than that lost by the extra supplies? How long would be payback time be, considering the cost of components and extra construction? The problem with your question is the chap can't begin with it the other way around. It would red plate. I wasn't suggesting he did it for real...just for the sake of logical process. The increase in efficiency is worthwhile if it relieves the anode without detriment elsewhere, and if for some reason only 6L6 are available. So, who'se only got 6L6, and wants more class A triode power? Thanks, Ian |
|
#24
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
er...actually, hasn't it got a red anode at 400V, 75mA?
What's the max dissipation of the 6L6 you have in mind? They vary quite a lot I would guess. The data I posted elsewhere gives max as 19W, but some of the application examples appear to exceed that. Especially the first "line out" example: http://tdsl.duncanamps.com/show.php?des=6L6GT Ian |
|
#25
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Ian Iveson wrote:
bigwig wrote: ...Interesting, Surely the screen would just become another anode and melt or sag and arc a bit and melt some more. ... But check the second application example, for "line out" he http://tdsl.duncanamps.com/show.php?des=6L6GT I checked all the links on this page and could not find any reference to 'line out'. Can you point to the right place please? Cheers ian |
|
#26
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Ian Bell wrote:
Surely the screen would just become another anode and melt or sag and arc a bit and melt some more. ... But check the second application example, for "line out" he http://tdsl.duncanamps.com/show.php?des=6L6GT I checked all the links on this page and could not find any reference to 'line out'. Can you point to the right place please? Actually it's "line output", and it's in the "class" column of the "application data". Second row. May as well quote from it: Va = 60; Vs = 250; Vg = 0; Ia = 180: Is = 18 First row is also interesting, as it's at the other extreme: Va = 700; Vs = 350; Ia = 70; Is = 6 Notice that the ratio Is/Ia is only a bit greater in the first example, so it is still some way from the knee. I was saying not long ago how good the 6L6 family is for distributed load, because the knee is so far to the left, but I'm surprised it's *so* far. What does "line output" mean? Why does it qualify to be a class? Perhaps it's intermittent with a short duty cycle? How come two such different setups are both in the "line output" class? Ian |
|
#27
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
flipper wrote:
snip But why not simply increase the HT by 100V? I really don't see the point of all this jiggerypokery. Because you'd red plate. His example is a class A amplifier and you'd need the 75mA idle at the higher B+. The plate can't handle it. Good, thanks. That's my first question dealt with, assuming you are right about the anode's max power. I see from this and your second post you're now off on whether 75mA is the 'actual max power'... Off? Not me, I know power is measured in Watts. It doesn't matter if his example using 75mA is precisely the 'max power', the principle remains. What principle? I can't see that you've introduced one, so how can it remain? Was it here before we arrived? Tell me what and where it is and I'll do my best to address it. The chap needs 6550 perhaps. It's been a while since they've been mentioned AFAIR. Hooray for the 6550. Or KT88 mebs. One can always propose getting a 'bigger tube' or, as I did, to use two. The same principle would apply to those as well. As did we both, but a bigger one would be better IMO. But there's that principle again, dammit. Now what I'm wondering is where all that energy goes. The electrons, to put it crudely, arrive at the anode having lost 100Vs-worth of energy since they passed the screen. What happens to it? On which side of the screen are you asking? The 'extra' acceleration from the 'extra' 100 volts between the grid and screen or the 'loss' of acceleration from the 'extra' 100 volts going from the screen to plate? The question is perfectly clear as asked, and I can't see what your quibble is. Let's say the circuit was derived the other way round, and the chap began with a triode with Va 500V, and he proposed his circuit as a way of reducing Va. What for? Would it be more linear? Would the increase in HT efficiency be more than that lost by the extra supplies? How long would be payback time be, considering the cost of components and extra construction? The problem with your question is the chap can't begin with it the other way around. It would red plate. I wasn't suggesting he did it for real...just for the sake of logical process. Same for my reply. Whether the example physically exists or not the logic has to fit reality to be of any use. Whatever, sigh... The increase in efficiency is worthwhile if it relieves the anode without detriment elsewhere, and if for some reason only 6L6 are available. If one were 'stuck' and really needed the power it might make sense but, IMO, 67%, only 2.2 dB, more power isn't worth the expense and complexity for an audio amp. I mean, compared to the alternatives. Seems to be a consensus, then. Ian So, who'se only got 6L6, and wants more class A triode power? Thanks, Ian |
|
#28
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
flipper wrote:
er...actually, hasn't it got a red anode at 400V, 75mA? Don't know because he didn't say B+ in his examples and what he did give indicates it wasn't any of the 'typical' values. He shows 400V HT in all three examples. What's the max dissipation of the 6L6 you have in mind? They vary quite a lot I would guess. 'I' didn't have anything in mind. His example comparison was with 75mA in all three, pentode, triode, and boosted triode so that's why it appeared in my post. So why did you tell me that it would "red plate" if both screen and anode were raised, rather than just the screen? I could just about work that out if he had said that his unboosted triode was set to give maximum class A output power available short of red-anode at any Va, but I don't think he did. Perhaps that what he meant, considering 400V, 75mA is 30W, which is what you say is the max anode dissipation. But even then, his 75mA includes the screen, so his anode dissipation will be less. I was remarking on the design posted, not a general principle. 75mA with 250V plate would just barely sneak in under 19 Watts but since he was doing simulations there's no telling what he decided ot do. The data I posted elsewhere gives max as 19W, but some of the application examples appear to exceed that. Especially the first "line out" example: http://tdsl.duncanamps.com/show.php?des=6L6GT EL34, which he also references, is 25 watt plate and the 6L6GC is 30 Watt plate. Well, some are design center and others design maximum. I wonder if Duncan's 19W is a mistake? Ian |
|
#29
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
On Thu, 14 May 2009 17:31:48 -0500, flipper wrote:
.....snip!.... Now what I'm wondering is where all that energy goes. The electrons, to put it crudely, arrive at the anode having lost 100Vs-worth of energy since they passed the screen. What happens to it? On which side of the screen are you asking? The 'extra' acceleration from the 'extra' 100 volts between the grid and screen or the 'loss' of acceleration from the 'extra' 100 volts going from the screen to plate? There are some issues when the screen is actually more positive than the plate, you'd think that the electrons might decide to turn around since they are in an electric field that is now opposite to the direction from where they came. It appears that they have enough kinetic energy to "roll uphill" to the plate. I checked some pentode plate characteristic curves where the screen is kept at a fixed voltage, and it seems that even when the plate is considerably less than the screen, the tube still works like you'd expect. That's for a beam pentode..... other geometries are quite different. Early tetrodes showed some pretty weird behaviour when this happened. Sometimes the plate resistance would go negative (tube would most likely oscillate). There are a number of factors to consider, electron velocity, space charge, etc. Check Reich's "Theory and Applications of Electron Tubes", McGraw Hill,1944, starting around p56. This book and many other great old texts is available from: http://www.pmillett.com/technical_books_online.htm From looking at the curves of several popular power tubes, it seems that you need a few hundred volts more positive on the screen before anything significant happens. The screen current does increase under the boost conditions, but not excessively, until more than a few hundred volts difference (7027 pentode). Older style pentodes like the 807 seem to show weird kinks once you get the screen more than 100V greater than the plate. Let's say the circuit was derived the other way round, and the chap began with a triode with Va 500V, and he proposed his circuit as a way of reducing Va. What for? Would it be more linear? Would the increase in HT efficiency be more than that lost by the extra supplies? How long would be payback time be, considering the cost of components and extra construction? The problem with your question is the chap can't begin with it the other way around. It would red plate. I wasn't suggesting he did it for real...just for the sake of logical process. Same for my reply. Whether the example physically exists or not the logic has to fit reality to be of any use. The increase in efficiency is worthwhile if it relieves the anode without detriment elsewhere, and if for some reason only 6L6 are available. If one were 'stuck' and really needed the power it might make sense but, IMO, 67%, only 2.2 dB, more power isn't worth the expense and complexity for an audio amp. I mean, compared to the alternatives. ...... snip!.... To the best of my knowledge, the reason for doing this trick in the first place is to shift the plate characteristic curves of the "effective" triode to the left, or closer to the Y-axis. This ought to increase efficiency and power (once an optimal operating point is chosen). In terms of AC signals, the boosted triode and the plate-screen connected triode are the same as far as the grid-one are concerned. I would expect the plate resistance of both circuits to be pretty similiar (since there is an electric field feedback between plate and grid that causes low plate resistance). If the screen ends up glowing, it would not have any extra current emitted since it's at the highest voltage, so it wouldn't act like a new cathode. I'm not sure otherwise what happens when you run the DC settings of the screen way beyond max recommended values. Paul G |
|
#30
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
flipper wrote:
On Thu, 14 May 2009 23:38:00 +0100, Ian Bell wrote: Ian Iveson wrote: bigwig wrote: ...Interesting, Surely the screen would just become another anode and melt or sag and arc a bit and melt some more. ... But check the second application example, for "line out" he http://tdsl.duncanamps.com/show.php?des=6L6GT I checked all the links on this page and could not find any reference to 'line out'. Can you point to the right place please? He said the second *application*. 6L6-GT Application Data Class Va Vg2 Vg1 Ia Ig2 Rk Line output 700 350 70.0 6.0 100 Line output 60 250 0.0 180.0 18.0 I'm sure he's referring to Vg2 being 190V over plate. Ah, I noticed the 'application' word but thought it referred to some section in one of the data sheets the page referred to - never thought to just look further down the initial page - doh! So precisely what is meant by a 'line out' application??? Is this not a reference to a TV flyback circuit of some sort?? Cheers ian Cheers ian |
|
#31
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Ian Iveson wrote:
Ian Bell wrote: Surely the screen would just become another anode and melt or sag and arc a bit and melt some more. ... But check the second application example, for "line out" he http://tdsl.duncanamps.com/show.php?des=6L6GT I checked all the links on this page and could not find any reference to 'line out'. Can you point to the right place please? Actually it's "line output", and it's in the "class" column of the "application data". Second row. May as well quote from it: Va = 60; Vs = 250; Vg = 0; Ia = 180: Is = 18 First row is also interesting, as it's at the other extreme: Va = 700; Vs = 350; Ia = 70; Is = 6 Notice that the ratio Is/Ia is only a bit greater in the first example, so it is still some way from the knee. I was saying not long ago how good the 6L6 family is for distributed load, because the knee is so far to the left, but I'm surprised it's *so* far. What does "line output" mean? Why does it qualify to be a class? Perhaps it's intermittent with a short duty cycle? How come two such different setups are both in the "line output" class? Isn't 'line out' something to do with TV line scanning drivers?? Cheers Ian Ian |
|
#32
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Here are the first two images of the AudioXpress article.
http://www.ianbell.ukfsn.org/data/807boost1.jpg http://www.ianbell.ukfsn.org/data/807boost2.jpg There's still a page missing so I'll ask for that too. Cheers Ian |
|
#33
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
"flipper" wrote in message ... On Fri, 15 May 2009 00:31:14 +0100, "Ian Iveson" wrote: flipper wrote: snip But why not simply increase the HT by 100V? I really don't see the point of all this jiggerypokery. Because you'd red plate. His example is a class A amplifier and you'd need the 75mA idle at the higher B+. The plate can't handle it. Good, thanks. That's my first question dealt with, assuming you are right about the anode's max power. I see from this and your second post you're now off on whether 75mA is the 'actual max power'... Off? Not me, I know power is measured in Watts. You're off on whether 75mA at 400V are the 'right numbers' for the blooming tube, as if it made any difference to explaining his circuit. Off? Not me, I was correctly remarking on the actual circuit posted, not your concept of some general principle. It doesn't matter if his example using 75mA is precisely the 'max power', the principle remains. What principle? I can't see that you've introduced one, so how can it remain? Was it here before we arrived? Tell me what and where it is and I'll do my best to address it. The principle of boosting the blooming triode mode to get more power out of the thing. If you mean that, as a general rule, more power can be extracted from a pentode or beam tetrode if the screen voltage is increased, than that seemed too obvious to be worth discussing, but go ahead if that's what you want to do. Why some want to give it the name "boosted", as if it is a different mode of operation, I really don't know. Neither can I see what "principle" is involved that is peculiar to the posted circuit, which conforms to the usual operating rules. To me, the real issues here are about the safe operation of a particular valve, the details of a particular circuit, and the principle of what sounds best. Valves are different. Those with knees further to the right will not be so amenable. Be wary of making a principle out of it. The chap needs 6550 perhaps. It's been a while since they've been mentioned AFAIR. Hooray for the 6550. Or KT88 mebs. One can always propose getting a 'bigger tube' or, as I did, to use two. The same principle would apply to those as well. As did we both, but a bigger one would be better IMO. But there's that principle again, dammit. Right. You could also boost the bigger tube or the pair for even more power. So what? On which side of the screen are you asking? The 'extra' acceleration from the 'extra' 100 volts between the grid and screen or the 'loss' of acceleration from the 'extra' 100 volts going from the screen to plate? The question is perfectly clear as asked, and I can't see what your quibble is. The 'quibble' is you only talk about 'one side' of the screen. They pick up 100V of acceleration on the cathode side and lose it again on the other. So why did you ask me which side I was talking about? And why have you switched from the "energy" in my question, to "acceleration"? This was the question: "Now what I'm wondering is where all that energy goes. The electrons, to put it crudely, arrive at the anode having lost 100Vs-worth of energy since they passed the screen. What happens to it?" What you say about acceleration appears obvious to me, but the question about lost energy rather less so, which is why I raised it. Let's say the circuit was derived the other way round, and the chap began with a triode with Va 500V, and he proposed his circuit as a way of reducing Va. What for? Would it be more linear? Would the increase in HT efficiency be more than that lost by the extra supplies? How long would be payback time be, considering the cost of components and extra construction? The problem with your question is the chap can't begin with it the other way around. It would red plate. I wasn't suggesting he did it for real...just for the sake of logical process. Same for my reply. Whether the example physically exists or not the logic has to fit reality to be of any use. Whatever, sigh... I don't know what's so confusing. The purpose is to get more power out of the tube and it doesn't add anything to postulate running the tube at x times what would burn it up as an 'alternative'. That doesn't require someone imagine you meant a 'real chap' or a 'real tube', neither of which I did so there was no need to 'explain' it. You asked why not do X and I explained that X would red plate. Barking, truly. I made it clear from the start that I was talking about this particular valve in this particular circuit. You can ponder principles if you like, but don't expect others to be constrained by your restricted thought processes. Ian |
|
#34
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Ian Iveson wrote: Patrick Turner wrote: Looking closer, the raising of Eg2 by 100Vdc above Ea means that Ia is increased for the same Ea, but if Eg1 is then increased from his -32V to -44V, then the G1 can swing much further without running into grid current. Obviously, and equally true were HT simply raised by the same 100V. Nope. Nope what? Nope its not obvious, or nope it would not be equally true if HT were raised by the same 100V? If the latter, then that is the crux of the biscuit, I feel. Are you suggesting that, if screen and anode are both increased by 100V, the appropriate bias voltage would be less than 44V? That would lead to the interesting conclusion that if you apply 500V to both screen and anode and a certain current results, then as you reduce Va, current will increase. Or are you just quibbling with "equally"? My view is that changing Va wouldn't have much effect because screen mu is so dominant. That's the main reason why I wondered if the design is a joke: it would be much simpler to increase Va and Vs by the same amount, than to increase Vs alone, obviously. You still don't get it. Ea has little effect on Ia where Eg2 is kept constant. With normal triode connection with Ea = Eg2 = 350V and say Ia at 80mA, with Ig2 at say 6mA, then you'd have total Pda + Pdg2 = 30W. From this we'd expect to see 7.5W of triode class A PO at 25% triode efficiency. Now, suppose we drop Ea to 250V, but keep Eg2 at 350V, while ac coupling the screen to anode for triode operation. If the Ia was say 75mA, and and Ig2 maybe 10mA, then the total Pda + Pdg2 = 19W + 3.5 = 22.5W, and we could get the anode to make the at least the same swing but probably a bit more so you'd get 9W, same swing and get get 40% efficiency. If we raised Ea and Ia and increased the grid 1 bias so that Pda + Pdg2 = 30W, if the efficiency was 40% we could get 12W. Just how much Eg2 needs to be above Ea is not known for all tubes. But you're the ****ing expert on ****ing simulation, so why have you ****ing not worked it all out for us?????? Patrick Turner. |
|
#35
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Ian Iveson wrote: flipper wrote: Instead of using a separate supply, if you have a voltage (B+ boost) a few hundred volts greater than B+ (can do with doubler or other circuit tricks), run a resistor to the screen from B+boost, and put a hefty capacitor between screen and plate. Select the value of resistance that puts about 100v DC between screen and plate. The screen current is normally 1/5 - 1/20 the plate current. In order to raise the screen voltage above the plate by 100V, you are going to need more screen current, and the screen current will be a higher proportion than before. The screen current is roughly proportional to the plate current (if away from the "knee" of plate characteristic), ... A moot point. How near to the knee does the original link's loadline go in "boosted triode" mode? Otherwise your idea seems reasonable to me. ...so during signal, if the tube runs fairly linearly, the 100V should stay on the screen (where the screen current increases and drops the voltage is balanced by the screen current decrease). If you are a purist, or suspect a lot of unlinearity, you could place a 100V zener across the plate-screen, but you'd need to draw more current. That extra current ends up in the plate circuit, probably not a good idea. The circuit that I proposed will end up being part of the plate load, and will draw off some off the output power. If your B+boost is quite high, the resistor mentioned above would be quite large, and would not sap significant audio power. The advantage using this circuit is that you don't need to worry about isolation or extra power supplies. The down side is that you need some regulation of the 100V. I'm assuming that a decent capacitor (10 - 50 uF ??) should hold down the screen-plate voltage. That capacitor should also reduce the effects of hum on the B+boost supply since it effectively clamps the screen voltage (sensitive to variations) to the plate, except at DC and very low frequencies. Is this a silly idea? I don't think it adds any silliness to the original idea :-) Another possibility would be a screen winding on the OPT of same turns as for the anode, instead of the resistor and cap? But why not simply increase the HT by 100V? I really don't see the point of all this jiggerypokery. Because you'd red plate. His example is a class A amplifier and you'd need the 75mA idle at the higher B+. The plate can't handle it. Good, thanks. That's my first question dealt with, assuming you are right about the anode's max power. The chap needs 6550 perhaps. It's been a while since they've been mentioned AFAIR. Hooray for the 6550. Or KT88 mebs. Now what I'm wondering is where all that energy goes. The electrons, to put it crudely, arrive at the anode having lost 100Vs-worth of energy since they passed the screen. What happens to it? Maybe that's not the right way to consider it. If Ea is always kept 100V below Eg2, then the ratio of Ig2 : Ia become lower, ie, a greater % of PS feed is used in the screen circuit. The idea of BT is to lower Ea, and get a wider anode voltage swing, like you get in beam or pentode operation. The screen need not be run into excessive Pdg2 region. The energy liberated by the tube as heat is simply the sum of enrgy input at anode and screen. And BTW, unlike beam or pentode operation with a fixed Eg2, the screen contributes to the power production. Let's say the circuit was derived the other way round, and the chap began with a triode with Va 500V, and he proposed his circuit as a way of reducing Va. What for? Would it be more linear? Would the increase in HT efficiency be more than that lost by the extra supplies? How long would be payback time be, considering the cost of components and extra construction? The problem with your question is the chap can't begin with it the other way around. It would red plate. I wasn't suggesting he did it for real...just for the sake of logical process. The increase in efficiency is worthwhile if it relieves the anode without detriment elsewhere, and if for some reason only 6L6 are available. So, who'se only got 6L6, and wants more class A triode power? Well, lotsa people have 6L6 laying around, but like 807, maybe they are not so hot in BT. The better tube would probably be EL34, 5881, or 6L6GC where the Eg2 rating is a bit higher, ie, the Pdg2 rating may not be challenged so easily. Patrick Turner. Thanks, Ian |
|
#36
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Patrick Turner wrote:
Obviously, and equally true were HT simply raised by the same 100V. Nope. Nope what? Nope its not obvious, or nope it would not be equally true if HT were raised by the same 100V? If the latter, then that is the crux of the biscuit, I feel. Are you suggesting that, if screen and anode are both increased by 100V, the appropriate bias voltage would be less than 44V? That would lead to the interesting conclusion that if you apply 500V to both screen and anode and a certain current results, then as you reduce Va, current will increase. Or are you just quibbling with "equally"? My view is that changing Va wouldn't have much effect because screen mu is so dominant. That's the main reason why I wondered if the design is a joke: it would be much simpler to increase Va and Vs by the same amount, than to increase Vs alone, obviously. You still don't get it. How would you know? You can't read, or entertain a structured thought process. Ea has little effect on Ia where Eg2 is kept constant. Va has little effect on Ia whether Vs is kept constant or not, above the knee. With normal triode connection with Ea = Eg2 = 350V and say Ia at 80mA, with Ig2 at say 6mA, then you'd have total Pda + Pdg2 = 30W. From this we'd expect to see 7.5W of triode class A PO at 25% triode efficiency. Now, suppose we drop Ea to 250V, but keep Eg2 at 350V, while ac coupling the screen to anode for triode operation. If the Ia was say 75mA, and and Ig2 maybe 10mA, then the total Pda + Pdg2 = 19W + 3.5 = 22.5W, and we could get the anode to make the at least the same swing but probably a bit more so you'd get 9W, same swing and get get 40% efficiency. Obviously, and efficiency wasn't the issue until I raised it. I was questioning the legitimacy of the particular circuit, which was introduced as a way of increasing power. If we raised Ea and Ia and increased the grid 1 bias so that Pda + Pdg2 = 30W, if the efficiency was 40% we could get 12W. Just how much Eg2 needs to be above Ea is not known for all tubes. But you're the ****ing expert on ****ing simulation, so why have you ****ing not worked it all out for us?????? Maybe I would if you asked nicely, maybe not, although not through simulation because I wouldn't trust my 6L6 screen model at low Va, and it won't tell me where the energy goes anyway. Anyway, I like to encourage the group to progress in a spirit of social endeavour. Telling everyone the answers is what you think you're here for. I don't need to do that to make a living. I just raise issues, mostly, and cultivate conversation, just like as if it were an amateur newsgroup. Here I am, bright eyes and bushy tail, eager to leave no stone unturned, while you strut, puffed chest and stern brow, keen to impress your stupid-rich patrons and sell more of your antediluvian hardware. It could have been a symbiotic relationship, but you turned out to be too much of a ****. Ian |
|
#37
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Ian Iveson wrote: Patrick Turner wrote: Obviously, and equally true were HT simply raised by the same 100V. Nope. Nope what? Nope its not obvious, or nope it would not be equally true if HT were raised by the same 100V? If the latter, then that is the crux of the biscuit, I feel. Are you suggesting that, if screen and anode are both increased by 100V, the appropriate bias voltage would be less than 44V? That would lead to the interesting conclusion that if you apply 500V to both screen and anode and a certain current results, then as you reduce Va, current will increase. Or are you just quibbling with "equally"? My view is that changing Va wouldn't have much effect because screen mu is so dominant. That's the main reason why I wondered if the design is a joke: it would be much simpler to increase Va and Vs by the same amount, than to increase Vs alone, obviously. You still don't get it. How would you know? You can't read, or entertain a structured thought process. Ea has little effect on Ia where Eg2 is kept constant. Va has little effect on Ia whether Vs is kept constant or not, above the knee. With normal triode connection with Ea = Eg2 = 350V and say Ia at 80mA, with Ig2 at say 6mA, then you'd have total Pda + Pdg2 = 30W. From this we'd expect to see 7.5W of triode class A PO at 25% triode efficiency. Now, suppose we drop Ea to 250V, but keep Eg2 at 350V, while ac coupling the screen to anode for triode operation. If the Ia was say 75mA, and and Ig2 maybe 10mA, then the total Pda + Pdg2 = 19W + 3.5 = 22.5W, and we could get the anode to make the at least the same swing but probably a bit more so you'd get 9W, same swing and get get 40% efficiency. Obviously, and efficiency wasn't the issue until I raised it. I was questioning the legitimacy of the particular circuit, which was introduced as a way of increasing power. If we raised Ea and Ia and increased the grid 1 bias so that Pda + Pdg2 = 30W, if the efficiency was 40% we could get 12W. Just how much Eg2 needs to be above Ea is not known for all tubes. But you're the ****ing expert on ****ing simulation, so why have you ****ing not worked it all out for us?????? Maybe I would if you asked nicely, maybe not, although not through simulation because I wouldn't trust my 6L6 screen model at low Va, and it won't tell me where the energy goes anyway. Anyway, I like to encourage the group to progress in a spirit of social endeavour. Telling everyone the answers is what you think you're here for. I don't need to do that to make a living. I just raise issues, mostly, and cultivate conversation, just like as if it were an amateur newsgroup. Here I am, bright eyes and bushy tail, eager to leave no stone unturned, while you strut, puffed chest and stern brow, keen to impress your stupid-rich patrons and sell more of your antediluvian hardware. It could have been a symbiotic relationship, but you turned out to be too much of a ****. Ian Have a nice sunday Ian. When your'e finished thinking I am a "****" which means "****" in australian english, perhaps you'd be so kind to simulate something useful for us all. I apologise to any females reading the group if I have offended any, but its unlikely any females are present. I have not seen any post here by any F for years, so perhaps when I do swear, its merely a man to man thing. Patrick Turner. |
|
#38
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
Patrick Turner wrote:
snip Have a nice sunday Ian. When your'e finished thinking I am a "****" which means "****" in australian english, perhaps you'd be so kind to simulate something useful for us all. I apologise to any females reading the group if I have offended any, but its unlikely any females are present. I have not seen any post here by any F for years, so perhaps when I do swear, its merely a man to man thing. Patrick Turner. That's a bit sexist isn't it Patrick? I thought women were supposed to be our equals now so they should expect to sworn at just like us blokes are ;-) Cheers Ian |
|
#39
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
flipper wrote:
Obviously, and equally true were HT simply raised by the same 100V. Nope. Nope what? Nope its not obvious, or nope it would not be equally true if HT were raised by the same 100V? If the latter, then that is the crux of the biscuit, I feel. Are you suggesting that, if screen and anode are both increased by 100V, the appropriate bias voltage would be less than 44V? That would lead to the interesting conclusion that if you apply 500V to both screen and anode and a certain current results, then as you reduce Va, current will increase. Or are you just quibbling with "equally"? My view is that changing Va wouldn't have much effect because screen mu is so dominant. That's the main reason why I wondered if the design is a joke: it would be much simpler to increase Va and Vs by the same amount, than to increase Vs alone, obviously. You still don't get it. How would you know? He knows because you keep making the same faulty suggestion "it would be much simpler to increase Va and Vs by the same amount." But that is true. What I questioned originally was whether that would take it out of its SOA. You should make an effort to read properly rather than pursue your own pet principles regardless. And just coz you think you know doesn't answer my question of how Patrick would know. You have more formal structure to your thinking, presumably as a result of some education. Patrick is comparatively clueless. Go ahead, keep increasing them until you hit max PDA. Then you can get more Po if you increase Vs in 'boosted triode'. You can't read, or entertain a structured thought process. Ea has little effect on Ia where Eg2 is kept constant. Va has little effect on Ia whether Vs is kept constant or not, above the knee. Not true in triode operation and neither is there a blooming knee in triode operation. So "blooming" what? "Eg2" isn't constant in a triode either. So what mode do you suppose Patrick was talking about? With normal triode connection with Ea = Eg2 = 350V and say Ia at 80mA, with Ig2 at say 6mA, then you'd have total Pda + Pdg2 = 30W. From this we'd expect to see 7.5W of triode class A PO at 25% triode efficiency. Now, suppose we drop Ea to 250V, but keep Eg2 at 350V, while ac coupling the screen to anode for triode operation. If the Ia was say 75mA, and and Ig2 maybe 10mA, then the total Pda + Pdg2 = 19W + 3.5 = 22.5W, and we could get the anode to make the at least the same swing but probably a bit more so you'd get 9W, same swing and get get 40% efficiency. Obviously, If it's so obvious then why do you persist in claiming the same thing can be done by raising B+. If I've said "the same thing can be done by raising" HT, I didn't intend to. Where did I say that? Where have I persisted? and efficiency wasn't the issue until I raised it. You never raised efficiency at all and pondering where the 'lost' 100Vs goes has nothing to do with it. Porky, pure and simple. You replied to the very paragraph in which I raised it. I was questioning the legitimacy of the particular circuit, which was introduced as a way of increasing power. The principle works. Sigh...I was questioning the legitimacy of the particular circuit. Once mo ***everyone knows the "principle" works*** ***the "principle" has never been disputed*** Let it go, you're chasing your own red herring, and you've got Patrick following you down the path of idiocy. If we raised Ea and Ia and increased the grid 1 bias so that Pda + Pdg2 = 30W, if the efficiency was 40% we could get 12W. Just how much Eg2 needs to be above Ea is not known for all tubes. But you're the ****ing expert on ****ing simulation, so why have you ****ing not worked it all out for us?????? Maybe I would if you asked nicely, maybe not, although not through simulation because I wouldn't trust my 6L6 screen model at low Va, Don't try modeling a TV horizontal amp then. You neither. Your model is no better than mine, but just in case it is, I think you should post a copy, thanks. and it won't tell me where the energy goes anyway. You got no means to measure I and A? Measuring tells me it's gone, but doesn't say where to. Anyway, I like to encourage the group to progress in a spirit of social endeavour. Telling everyone the answers is what you think you're here for. I don't need to do that to make a living. I just raise issues, mostly, and cultivate conversation, just like as if it were an amateur newsgroup. Here I am, bright eyes and bushy tail, eager to leave no stone unturned, while you strut, puffed chest and stern brow, keen to impress your stupid-rich patrons and sell more of your antediluvian hardware. It could have been a symbiotic relationship, but you turned out to be too much of a ****. Pat can sometimes wander all over the place and be stubborn enough to make a mule look down right flexible but you give him a good run on both counts and you meandering over everything except the circuit's principle is an example of it. But the principle was established at the outset, silly. Your first post wasn't about the principle. The only reason you've been banging on about it is because you want to think I don't understand it, for reasons perhaps you haven't yet grasped, despite my posts that make the contrary clear. I am generally on topic, raising pertinent issues as I have done here. You may argue that the dialectic, for example, is not pertinent, but I know otherwise and hope that some may see that sometime, or honestly ask me why. I don't wish to discuss Patrick's record. I don't read much of his stuff because it is cringingly embarassing, numbingly tedious, or bullying, in various combinations, and I don't go to his site because I assume it's more of the same. He doesn't need you to defend him, either. He's insensitive to all input. In any tube operated triode mode your max Po is limited by the plate's PDA but you can get more Po if you use 'boosted triode' mode (within screen limits). That's all there is and the rest is simply 'showing it'. All there is for you, maybe, but there is much of interest to discuss for those with curious and open minds. In particular, there is the absolutely key question of how it sounds. That inherently means your efficiency is better, is 'why' it works, and why Pat mentioned it. It was me who first introduced the issue of efficiency, and who first pointed out that efficiency is improved. I raised the issue, typically, in the form of a question. Your supposedly 'just as easy' suggestion to raise both Va and Vs won't work because you increase PDA. Raising Vs alone does not. I thought I said it was *easier*. And so it is. I also questioned whether it would take the valve out of its SOA. I also questioned at the same time whether the "boost" (horrible and useless term) would also take it out of its SOA. Now everyone else knows the answers to my questions, IFAICS, and won't forget. Much better than the wet blanket of a definitive pronouncement, I always think. The penalty of having smart-arses yapping at my ankles is a small price to pay for world domination. Where the 'lost' electron energy goes might amuse you but it's irrelevant to the circuit as it's analyzed the same as any other circuit. Not irrelevant to the circuit, but only to your narrow mind. Not off topic, either. I thought it might be a question that interested someone, particularly because the answer explains why the anode doesn't get so hot, just in case someone was wondering, or should have been. But maybe the connection between heat and energy is lost on you? Whether the author's example is 'practical' might be a question for builders but it's irrelevant to the principle and you weren't going to build one anyway. Nor does it seem anyone else is as all comments, so far, have been on the order of "not worth it." Seems a shame to decide it's not worth it without considering its contribution to the sound. Two other possible ways of achieving the same effect have been suggested. Are they not relevant either, according to your mind? Are they also not worth it, in your mind? Maybe it's just too small. I can help, if you let me. Ian |
|
#40
Posted to rec.audio.tubes
|
|||
|
|||
|
Boosted Triode
"flipper" wrote in message ... On Fri, 15 May 2009 11:22:02 +0100, "Ian Iveson" wrote: "flipper" wrote in message . .. On Fri, 15 May 2009 00:31:14 +0100, "Ian Iveson" wrote: flipper wrote: snip But why not simply increase the HT by 100V? I really don't see the point of all this jiggerypokery. Because you'd red plate. His example is a class A amplifier and you'd need the 75mA idle at the higher B+. The plate can't handle it. Good, thanks. That's my first question dealt with, assuming you are right about the anode's max power. I see from this and your second post you're now off on whether 75mA is the 'actual max power'... Off? Not me, I know power is measured in Watts. You're off on whether 75mA at 400V are the 'right numbers' for the blooming tube, as if it made any difference to explaining his circuit. Off? Not me, I was correctly remarking on the actual circuit posted, not your concept of some general principle. So you were asking why not increase B+ another 100V on a circuit you already consider to be red plating. Clever. Indeed. Now you see it... It doesn't matter if his example using 75mA is precisely the 'max power', the principle remains. What principle? I can't see that you've introduced one, so how can it remain? Was it here before we arrived? Tell me what and where it is and I'll do my best to address it. The principle of boosting the blooming triode mode to get more power out of the thing. If you mean that, as a general rule, As a general technique. That *is* his point. Who's point? How do you know? What point anyway? What are you on about? more power can be extracted from a pentode or beam tetrode if the screen voltage is increased, It's being operated triode mode, not pentode or tetrode mode. Getting desperately silly now. It's still an example of a pentode or beam tetrode. If you went to a shop and asked for a triode, they would be quite correct in saying they didn't have one for sale, if all they had in stock were 6L6. than that seemed too obvious to be worth discussing, but go ahead if that's what you want to do. If it was so 'obvious' to you then why were you asking about increasing B+? To raise the issue. What's obvious to me is not necessarily obvious to others. It was not immediately obvious that raising the HT would result in a red anode, anyway, until I went to look at, and post, 6L6 data. I could have looked at the datasheet first, of course, but I'd rather raise the issue. It all worked out quite nicely, allowing you display your knowledge, which seems to be what you like to do. Is this getting through to you yet? I'm getting a bit bored now, and would rather talk about the circuit. I'm not keen on discussing me. I'm a private and quite secretive kind of chap when it comes to newsgroups. Why some want to give it the name "boosted", as if it is a different mode of operation, I really don't know. Neither can I see what "principle" is involved that is peculiar to the posted circuit, which conforms to the usual operating rules. Because it's triode mode and not, as you suggest, the 'typical' pentode operation where you simply stick a fixed V on the screen. Suggested what? Where? Pentode operation often depends on *not* having a fixed V on the screen, incidentally. There's another opportunity for you to learn something. My amps modulate the screen in order to operate in pentode mode, when I want them to, which isn't very often because my knees are too far to the right and I risk early screen and grid current, and the extra power isn't worth having. That's why I was considering changing to 6L6, as I have been saying for some time. To me, the real issues here are about the safe operation of a particular valve, the details of a particular circuit, and the principle of what sounds best. Fine. So I'll amend the answer to it's a lousy idea to suggest increasing B+ another 100V on a circuit you think is already red plate. I was asking why not. That's not the same thing as suggesting you do it. I really don't know why you go off into these nonsensical babbles when you weren't going to build one anyway. I can't be wrong and nonsensically babbling at the same time, clot. There is a great deal of the world you can't make sense of. Don't blame me, I'm trying to help you. Valves are different. Those with knees further to the right will not be so amenable. Be wary of making a principle out of it. There ain't no blooming 'knee' when a triode. So what? and there is, anyway, dafty. Look at the original article's characteristic curves. The chap needs 6550 perhaps. It's been a while since they've been mentioned AFAIR. Hooray for the 6550. Or KT88 mebs. One can always propose getting a 'bigger tube' or, as I did, to use two. The same principle would apply to those as well. As did we both, but a bigger one would be better IMO. But there's that principle again, dammit. Right. You could also boost the bigger tube or the pair for even more power. So what? Exactly what it says. The principle applies whether a 6L6, a 'big tube' or duals, or pairs, or quads, or... So what? On which side of the screen are you asking? The 'extra' acceleration from the 'extra' 100 volts between the grid and screen or the 'loss' of acceleration from the 'extra' 100 volts going from the screen to plate? The question is perfectly clear as asked, and I can't see what your quibble is. The 'quibble' is you only talk about 'one side' of the screen. They pick up 100V of acceleration on the cathode side and lose it again on the other. So why did you ask me which side I was talking about? It was to give you a clue that what you 'wonder' about being 'lost' is also gained. So you asked a question when you already knew the answer. Isn't that what you are telling me is nonsensical babbling? And why have you switched from the "energy" in my question, to "acceleration"? What makes you think "100V" is energy? I don't. I asked where the "100Vs-worth" of energy goes. Since the electron-volt (eV, I think) is a unit of energy, I hoped you would know what I meant. You replied in terms of acceleration. I'm telling you that doesn't answer my question about energy, sigh... Acceleration is where the energy comes from. Er...maybe, maybe not. Could be a very misleading statement, and I can't see where it would be a useful one without qualification. If you think about it, acceleration doesn't come from, or go to, anywhere, does it? But energy must always go somewhere, and come from somewhere, mustn't it? That's why there is a principle (something you're fond of, yes?) of conservation of energy, but not a principle (I'll say it as often as I can if it pleases you) of conservation of acceleration. There you are, more education for your efforts. Keep 'em coming! This was the question: "Now what I'm wondering is where all that energy goes. The electrons, to put it crudely, arrive at the anode having lost 100Vs-worth of energy since they passed the screen. What happens to it?" Same place it goes in an audio amp when the screen is fixed at 300V and the anode swings to 200V, or to 100V, or to 50V. Or, in other words, the entire negative half of every sine wave. But where does it go? What you say about acceleration appears obvious to me, but the question about lost energy rather less so, which is why I raised it. Never bothered you before. How would you know? The energy gained/lost due to the electrostatic field comes from the electrostatic field. Hmm...getting warm. Where does it go to? Let's say the circuit was derived the other way round, and the chap began with a triode with Va 500V, and he proposed his circuit as a way of reducing Va. What for? Would it be more linear? Would the increase in HT efficiency be more than that lost by the extra supplies? How long would be payback time be, considering the cost of components and extra construction? The problem with your question is the chap can't begin with it the other way around. It would red plate. I wasn't suggesting he did it for real...just for the sake of logical process. Same for my reply. Whether the example physically exists or not the logic has to fit reality to be of any use. Whatever, sigh... I don't know what's so confusing. The purpose is to get more power out of the tube and it doesn't add anything to postulate running the tube at x times what would burn it up as an 'alternative'. That doesn't require someone imagine you meant a 'real chap' or a 'real tube', neither of which I did so there was no need to 'explain' it. You asked why not do X and I explained that X would red plate. Barking, truly. I made it clear from the start that I was talking about this particular valve in this particular circuit. You can ponder principles if you like, but don't expect others to be constrained by your restricted thought processes. If you now want claim you were daft enough to propose increasing B+ in a circuit that you already considered to be red plate, not to mention one you had no intention of building, then so be it. I never proposed it. I asked a question. As you have already remarked, it was a clever question, amongst others. But enough of my cleverness, what about the sound? Ian |
| Reply |
|
| Thread Tools | |
| Display Modes | |
|
|
Similar Threads
|
||||
| Thread | Forum | |||
| 814 like as triode? | Vacuum Tubes | |||
| Triode Feedback | Vacuum Tubes | |||
| Power Boosted FM Transmitter for iPod MP3 Player | Car Audio | |||
| Using power triode or power pentode wired as a triode as a split-load phase splitter tube? | Vacuum Tubes | |||
| Using power triode/pentode wired as triode as a split load phase splitter tube? | Vacuum Tubes | |||
![Ian Bell[_2_] is offline](images/SultanThemeVB3R/statusicon/user_offline.gif)
Linear Mode
