I'm looking at both CCS vs Interstage transformers.
Many folks like the CCS clean spotless sound.
However, some folks like the more articulate "lush" sound.
Since I have not build nor heard either one yet, I'm looking for
opinion of both world.

I'll be using either 645 or 7788 or EC8010 with 1:1 interstage
transformers.
I was told that 417a do not have enough gain for 1:1 interstage
transformer.

TIA

Anonymous wrote:

What do you prefer in your 300B SE?
rec.audio.tubes

Audiquattro wrote

Im looking at both CCS vs Interstage transformers.
Many folks like the CCS clean spotless sound.
However, some folks like the more articulate "lush" sound.
Since I have not build nor heard either one yet, Im looking for
opinion of both world.

Ill be using either 645 or 7788 or EC8010 with 1:1 interstage
transformers.
I was told that 417a do not have enough gain for 1:1 interstage
transformer.

TIA

Okay! You got the right idea here, Master Chang.

You dont tell us what your speakers are and how much power you need fro=
m your 300B to drive them. We are left to assume that you will hog the ma=
ximum power out of the 300B. Thats not necessarily smart. In fact you sho=
uld design backwards from the speakers, not forwards from the driver stag=
e. How much power you need, and the load you will put on the 300B, higher=
for lower distortion, in effect determine the driver stage and other arr=
angements.

Lets take an example. Say you want the gennie 8W that 300B are easily c=
apable of (10 and 12W 300B operate on ponycar watts, SAE lies, nothing to=
do with truthful DIN standards), you must then probably choose 3K or 3K5=
primary load, which fixes your tranny and many of your other operating c=
onditions. You will probably want to operate the 300B around 385 plate vo=
lts (that is after the drop for autobias, and after the drop over the tra=
nsformer copper drop, not total B+) and pretty near 80mA and 80V on the g=
rid, or something nearby. If this is your first 300B, thats a bit higher =
than most here would advise, around 80 per cent of max dissipation. Howev=
er, at this point 417A are not out of contention at all even with constan=
t current load (what you call CCS) or 1:1 IT. You can still, if you like =
its warm sound, very complementary to 300B, in fact superbly matched, cho=
ose 417A. It has a mu of 43, so you need under 2V from your source to dri=
ve a two-stage 417A-300B amp to full output. Most
CD players put out 2Vrms so you dont even need a pre-amp, you can build=
a linestage 417A-300B and Bobs your uncle. Less is more. If you use viny=
l, any pre-amp that cant put out 2V is pretty useless

Andre is quite right about all of this.

The 417 isn't the most linear triode in the universe.
But in this case, it probably doesn't matter, because even when optimally=

set up the DC load resistor and CR coupled to the 300B, the 2H generated =
by the
non linear triode driver will tend to cancel the 300B 2H, due to the rela=
tive phases
of the 2H distortions considered.
In fact, Andre's correct assertion that going for pure grunt from a 300B =
is
not all that wonderful an idea, and a higher ohmic value RL for the 300B
is going to sound better, and along with the better sound is lower tube d=
istortions
at the level at which the listening is done.
The driver still has to make a fair few volts so its 2H will still be pre=
sent,
but the thd cancelation is more complete if the thd of the output tube is=
lowish.



I dont know these other driver tubes you mention and cant be bothered t=
o look them up. The tubes I have used for driving 300B include 300B, 417A=
, EL34, mu stage SV83 over various 9-pin tubes with a Svet 6N1P as the st=
andout, 6SN7, 6SL7 choke coupled (this is with the choke in the grid leak=
position, not loading the plate directly), most of these also in SRPP, m=
any HF choke or tube or silicon constant current loads, various IT scheme=
s. Those are the ones I liked. Like most on RAT, I kinda forget to mentio=
n the failed experiments! What most of these schemes have in common is th=
at they put quite a bit of current on the driver plate. I used to use sle=
w rate current as a short cut to working out how much current power tubes=
need on the plate of the driver before them but that sent some of the ra=
ilroad minds that used to come to RAT into paroxysms of rage. (Difficult =
to decide if their rage was engineering-based as they went into paroxysms=
of rage when I said the sun shone in Ireland.)

The sun shines in Ireland, really, we know it does, between showers of ra=
in,
which maybe they need to cool the passions, keep them out of the fields,
and keep the population high.

But I have never used a choke as a grid leak, instead of a resistor.
But the 1:1 transformer has its merits if the tranny is a good one.
It may need some resistance across the secondary to damp the
peak in the response at HF that is sometimes seen with transformers.
The R value has to be experimented with, and should never be
too low a value, lest the driver tube be loaded adversly.

An alternative is to use a choke to supply the DC to the driver anode,
and have normal CR coupling to the 300B grid.

Since there is no rather low value of R which one would normally use to
]get DC to the driver anode, the distortion it causes is reduced, and the=
main
load seen by the driver is the C coupled grid bias R on the 300B, and its=
value is
usually much higher than the DC supply R, so in effect
the driver load produces much less 2H and in effect the load line
when drawn across the tube curves appears flatter, and triodes enjoy
flat loadlines to operate at their lowest possible distortions.

I have had good reports from those who also retain the normal
DC supply R to the driver anode,
and place a nice choke in series with the R, and thus at extremes of the =
audio
band the shunt L and self capacitance of the choke is more isolated from =
the
signal path.
The choke still raises the impedance seen by the driver tube.

I think lowish driver tube current in can produce good sound,
but at least i will agree that lots of current in a driver, more than is
really needed to overcome miller capacitance etc, is a good idea, rather =
than a bad one.
Instead of using say a 6SN7 as a driver, why not use a 6V6 in triode,
with about 20mA of idle current ?
Such a scheme calls out for a choke or transformer load,
and I believe they sound very OK.
I routinely use EL84 in triode to drive output stages in PP amps, with
L+R chokes to deliver the DC, and it works! The dynamics in the sound
is better than if I were to try to use a 6CG7, which is identical electro=
nically to the
6SN7, ( but watch out which one you pick, because driver tubes in amps
with zero FB do make a big difference to the sound. )
One EL84 as a triode is almost the same as 5 halves of 6CG7/6SN7 all para=
lleled.
But the gain will typically be only 19, so the 417 or more gainly triode
will allow you to reduce the tube stages, wheras using a driver gain
of 19 means an input tube is needed, but it can be a low U type...



All
the same it works. As a rule of thumb, good-sounding drivers for 300B p=
ermit 8mA minimum on the plate of the driver, preferably 10mA. (Steve Ben=
ch also told me, publicly on RAT I seem to remember, that 8mA minimum is =
a good rule for a driver for a 300B.) You can see why those of us with go=
od ears think 12A?7 sound like **** with 300B, and with most attractive p=
ower tubes. A 417A sounds real good from somewhere around 18mA, which by =
itself (extending the bandwidth of the 300B) may account for its synergy =
with the 300B.

So, even if youre hogging out the 300B for some reason, good bad or ind=
ifferent but at any rate unknown to us, the 417A suits your topology just=
fine.

I like the hivolt, hicurrent HVHC sound, mainly because it is intrinsic=
ally virtually undistorted for a substantial middle portion of its transf=
er curve and even at full power, given only a high primary load (I like t=
he Lundahl LL1623SE 90mA output transformer on its 5K6 setting), you can =
get distortion figures that are vanishingly small. But even that fades in=
to insignificance when you discover that HVHC 300B with a 5 or 6K load of=
fers this tiny distortion in a package that is all second harmonic with p=
ractically zero third and higher harmonics. Once you get into this class =
of amp what matters isnt how low the distortion is but what its harmonic =
distribution is.

In fact, with horn speakers, the power needed is so low that thd
isn't very audible, but only the magic of triodes is....

One fellow I know uses SET EL84 in triode, only 1.5 watts,
and it still sounds OK on his Lowthers, ( although he has serious respons=
e flatness bothers
with these diy horns he built, but that's another story.)



However, a little clear thinking will often persuade you that hogging t=
he 300B out to high dissipation is not necessary. Im currently building a=
300B with my last pair of brand new matched WE300B. This will probably b=
e my last amp and it will be a straightforward RC coupled stereo linestag=
e integrated. It will put out only 3.8W. Other provisional operating poin=
ts are 325V plate voltage, Eg-65V, Ia 65mA, RL 5K6. This 3W8 is more than=
enough to drive my custom Lowther Fidelio type horns. (In fact I am as w=
e speak driving these horns with a 417A potato amp putting out about a th=
ird of a watt. See
http://users.rcn.com/jbyrns/HWAFhorn.jpg
http://users.rcn.com/jbyrns/MZamp.jpg
http://www.geocities.com/mixtel99/index.html
) The other purpose of this 3W8 300B amp is to be booster amp for a 75W=
PSE SV572-xx amp, with the 300B delivering very high drive voltages to t=
he grids of SV572-xx or 845.

Now lets investigate this rather relaxed regime from the 417A voltage g=
ain viewpoint you state as your main consideration. At about 175V, grid b=
ias 2.4V (two AAA rechargables), and 20mA or so, 2Vrms out of my Quad 67 =
CD player causes the 417A to swing 63V peak to the grid of the 300B.

If you use 417A, take a few tips. Never forget it is primarily a radio =
tube. If has a miraculous sound but if you treat it as just another tube =
it will soon turn into a pain in the ass. For a start, check out Steve Be=
nchs netsite in detail. Make sure you understand all his remarks and sche=
matics for the 417A. Dont use 417A without gridstoppers to all grids. I s=
older mine between the pin and the centre pin of the socket and then brin=
g the signal in to the centre pin. I dont see the point of a plate stoppe=
r, or maybe my airwaves are just cleaner than the guys who need plate sto=
ppers. At the input end use a low value pot, say twice the load your sour=
ce expects, max. My Quad CD expects 10K, so I use 20K DACT stepped attenu=
ators (they are made on Elna switches for medical instruments in the firs=
t instance and there are none better) before 417A with the grounded arm a=
cting as the grid leak. If you use a separate grid leak, keep it small, l=
ike 10-20K. I like battery bias on 417A:
convenient, simple, another way of keeping radio crud out of my amp. 41=
7A draw grid current well to the negative side of zero grid volts. I alwa=
ys allow 400mV but it has been said by some others whose experience I res=
pect that one should allow 1V. If I need to tell you to keep all leads to=
417A short you shouldnt be playing with them, so I just add it for the n=
ewbies who have never heard of 417A before.

This is still not my favourite application of 417A. My alltime best app=
of this super tube was direct coupling it to medium-voltage SV572-xx wit=
h the plate voltage to the 417A derived through a silicon CCS from the ca=
thode of the SV572. (Best of all, I didnt even have to scratch my head fo=
r this scheme. It was sent to me as a sketch to illustrate a point in pri=
vate discussion by a great Australian, Anthony Mills. I promptly built i=
t, loved it, and only afterwards worked out how it operated.) Id explain =
it to you but the problem is that 300B, with a 450V limit (probably dicey=
for some Chinese tubes anyway) dont have the DC headroom to drop the pla=
te voltage for 417A even if you are willing to operate the 300B on fixed =
bias (which I dont recommend in any event - - whyever choose a tube whose=
greatest glory is its natural sound and then give it an artificial bass =
boost with fixed bias?).

Is there a difference between constant current loads and IT? Mmm. What =
I think is that a lowered output impedance at the driver improves your so=
und, whether it be real CCS by audio spectrum choke or silicon simulation=
, SRPP (which is not a constant current load despite claims to the contra=
ry), mu-follower or stage, choke coupling, whatever you like. (I think ca=
thode drive often sounds dull but, hey, whatever turns you on.)

If there is a truly meaningful sonic difference between these plate loa=
d schemes for the driver, you will need an ultra-high definition chain to=
hear it. I have 75W SE amps and can just barely hear the difference on Q=
uad ESL63 electrostats, by common consent the most ruthlessly revealing s=
peakers ever made. (Except for their older sister, the ESL57. I have thos=
e as well. Anyone who thinks Im going to put 75W into them is teenage ano=
rak who should go pick his nose elsewhere. ESL57, being of a different co=
nstruction, are anyhow not quite so useful for differentiating even small=
SE amps.) Horns, even good Lowthers like mine, simply dont have the reso=
lution for this kind of microthin hairsplitting. Perhaps the best horns i=
n the world, the Tannoy Westminster Royal, would be more useful. I doubt =
it. Part of the charm of horns of all types is that they eschew that wret=
ched American and Japanese audiophool fascination with hearing the saxoph=
onist burbling his spittle in his instrument.
Frankly, I think you should agonize over the quality of your speakers b=
efore you start splitting hairs over the difference between driver-loadin=
g schemes. I cant think of any speakers you can drive with SE 300B that w=
ill allow you to hear the difference well enough to get a good fix on it.=


Mind you, there is a practical consideration. A CCS can be a few bits o=
f silicon and a scrap of matrix-board. A good IT or audio choke costs rea=
l money. Try the CCS first and if you like it a lot, which you will, it i=
s likely that you will like either choke loading or the IT even better. Y=
ou need buy only the IT because you can use the inductance of its primary=
as a choke load with the secondaries unconnected. Mo iron is betta iron!=


I hav tried transistor CCS to the anode of the input tube ahead of the dr=
iver,
and found the sound improved.
I also like CCS applied to preamp anode circuits....
The sound seems clearer, but airy, and alive, but chokes in such
situations where the tube only has to make a few vrms seems to be over th=
e top,
but perhaps it sounds better.


I hope this helps, even if it isnt quite what you were hoping to hear. =
I have quite a bit of 417A and 300B experience and professionally good ea=
rs yet I am constantly surprised at claims people make for amps built wit=
h these tubes.

What of the 6C45pi?

What about a 6EJ7 frame grid pentode strapped in triode?
Lowish Ra, U about 60....

I am sure there are a few possibilities.

Patrick Turner.






Andre Jute

-=3D-
This message was posted via two or more anonymous remailing services.

Interstage transformers are extremely rar these days, and i suppose it is
VERY difficult (and $$$$$) to get one which does not fail in terms of
bandwidth and phase rotations.
Unless You're talking of a very high budget project I'd avoid using "iron".
A good alternative is the old GEC scheme where stray inductance is shunted
with a suitable capacitor: a LC coupling, improved with this simple "fix",
should be an effective solution.
BTW my 2A3 amp uses a plain LC coupling (with an EC86 driver) and I love its
sound, but it has serious bandwidth limits: I wouldn't suggest it as a
reference.

Ciao

Fabio


"Audiquattro" ha scritto nel messaggio
om...
I'm looking at both CCS vs Interstage transformers.
Many folks like the CCS clean spotless sound.
However, some folks like the more articulate "lush" sound.
Since I have not build nor heard either one yet, I'm looking for
opinion of both world.

I'll be using either 645 or 7788 or EC8010 with 1:1 interstage
transformers.
I was told that 417a do not have enough gain for 1:1 interstage
transformer.

TIA

Hi Pat,

6C45pi is also on the tube list. However, I want to experiment with
other alternatives as well. I have not look at 6EJ7 yet but
7788/E180F will be strapped in the triode.

thank you for your experiment with CCS. Which CCS did you use for
your amp?
- -- - - -- - - - - - - ---
Hi Andre, I'm designing from the speakers. I'm just curious what
other folks have tried in their experiment. I'm using the
Electra-Print 3K OPT and also got the RIT5 sr ITs.

My speakers are ranging from 94+ dbs (Jensens and Altecs) to 100+ dbs
(JBL horns). It's a good experiment for all the different setups I
can play with various output (from 3W to 8W) and I do have preamp
already. I also consider using the SV572 or the 300B+ tubes with the
417A down the road.

thank you for the long post regarding to the things you have tried.

Fabio Berutti wrote:

Interstage transformers are extremely rar these days, and i suppose it is
VERY difficult (and $$$$$) to get one which does not fail in terms of
bandwidth and phase rotations.

IST are not that rare at all. Sure, they cost, but so what?
Does somebody who owns a yact complain about occasionally having to spend money
in his pride and joy moored out in the bay?


Unless You're talking of a very high budget project I'd avoid using "iron".
A good alternative is the old GEC scheme where stray inductance is shunted
with a suitable capacitor: a LC coupling, improved with this simple "fix",
should be an effective solution.

The leakage inductance of an IST can be somewhat high,
and indeed there is the shunt C also involved, and they form a resonant circuit,

or rather an underdamped second order low pass filter, hence a peak in the
response
at over 5 kHz.

The "parasitic L and C components in the model of the IST are a fact of life
which cannot be avoided.

The traditional solution is to damp the LC critically with some R across the
sec,
but it loads the primary also, so that the triode driver ends up
being loaded by R at the mid-frequencies, something we might be trying to avoid,

to make the triode produce its lowest Dn.
So inded a snubber R+C network might be able to be connected across the sec, so
that at HF the
load is enough to damp the response critically, and yet at the mid frequency
the RC load has become a high and negligible impedance, thus allowing the triode

to get on with low Dn signal amplification.

Transformer Dn is the other possible problem, but it is reduced by using
low Ra triodes, and staying well away from pentodes.
The lower the Ra, the lower the Dn caused by the iron, which usually
inserts 3H into the signal path, even at low levels where it is worst as a % of
the total signal,
and right in the range of where we want to do our listening.
Having 0.1% of 3H inserted into the signal at a watt of output isn't uncommon,
and it seems silly to try to make the tube caused thd lower than 0.05%, and all
even order
but them be blighted by a similar amount of odd order crap.

So IST should be very carefully used indeed.

Patrick Turner.



BTW my 2A3 amp uses a plain LC coupling (with an EC86 driver) and I love its
sound, but it has serious bandwidth limits: I wouldn't suggest it as a
reference.

Ciao

Fabio

"Audiquattro" ha scritto nel messaggio
om...
I'm looking at both CCS vs Interstage transformers.
Many folks like the CCS clean spotless sound.
However, some folks like the more articulate "lush" sound.
Since I have not build nor heard either one yet, I'm looking for
opinion of both world.

I'll be using either 645 or 7788 or EC8010 with 1:1 interstage
transformers.
I was told that 417a do not have enough gain for 1:1 interstage
transformer.

TIA

Anonymous wrote:

rec.audio.tubes
What do you prefer in your 300B SE?

What Patrick Turner intends to say on interstage transformers in his confused post below is this:

I am far from sure that my post was confused, or that I intended to say all of the following,
so allow me to expand on what you say I intended, lest anyone have any doubts....



The inductance of a transformer primary is a variable quantity.

True, and therefore the impedance changes during a single voltage cycle.


It varies from transformer to transformer and in a particular transformer with signal voltage.

Yes.

Using this inductance in a resonant circuit increases distortion.

I am sure I would never have intended to say this.
I did say that leakage inductance and stray capacitance get together to form a resonant circuit,
but I don't recall having said Lp resonates with anything.


Primary loading causes tube distortion and secondary loading increases transformer distortion. A resistor shunted across the secondary does smooth out high frequency response and a resistor shunted across the primary extends bass response. However, the primary inductance and the shunt resistor form a series resonant circuit which causes the tube to see a low impedance at the resonant frequency which at high outputs will cause serious distortion. In consequence good design avoids transformer shunts.

Nowhere have I ever said "that the primary inductance and the shunt resistor form a series resonant circuit"
It is impossible for any inductance and resistance to act as a resonant circuit.

Transformer shunts by resistance are usually only used to tame the peaks in the HF response
if there are any, and my optional recommendation included the use of an R+C
series network across the secondary, to provide critical damping of the inevitable
second order roll off of HF caused by leakage L and stray C.
The C does not conduct at mid frequencies, but does at above say 5 kHz,
and such a nework may tame the HF peaks.

RC damping may not always be needed, and perhaps not at all in the case of a very well
wound transformer.

So you have it from the horses mouth what was intended, rather than from your imagination.



Better to buy a good IT in the first instance, as Fabio suggests. He is wrong about good ones being ra the Lundahl catalogue lists a wide variety and so do other winders by appointment to the big broadcasters.

Some of the very best amplifiers I have heard were all transformer coupled,
but good ISTs are rare compared to OPTs.
If folks are willing to pay for them, there are plenty to buy.
Some of the worst amplifiers I have heard were SETs with all transformer coupling,
but done without the slightest attention to the response which turned out
to be attrocious, because the owner doesn't like measurements, and thinks he can
go on the sound quality only.

I recently listened to some really attrocious sound from some
Lowthers in home made horn enclosures, powered by 1.5 watt SET.

I measured the response, and the audio was about flat from over 10k to 5k, but
then rolled off at 3 dB/octave, and the only way to tame this terrible bright bassless
sound is by adding a network at the amp input.
There may still be something wrong with the woodwork of his speakers,
but I wasn't asked to go in with a saw to find out.
The same amps gave a flat response into some Paradigm speakers bought
from a local hi-fi store, and positioned alongside the Lowthers in the same room.
So had I not measured, the owner, who is completely technically untrained,
could have come to a wrong conclusion
about SET amplifiers; best the ears be distrusted at all times, until at least the basic
engineering like a fairly flat response is obtained.

So whatever is built should be rigourously tested, to make sure there is at least a flat response
signal being sent to the speakers, and that the acoustic response is about flat.

Patrick Turner.




Andre Jute
A little, a very little thought will suffice John Maynard Keynes

Patrick Turner sent this:

Fabio Berutti wrote:

Interstage transformers are extremely rar these days, and i suppose it is
VERY difficult (and $$$$$) to get one which does not fail in terms of
bandwidth and phase rotations.

IST are not that rare at all. Sure, they cost, but so what?
Does somebody who owns a yact complain about occasionally having to spend money
in his pride and joy moored out in the bay?


Unless Youre talking of a very high budget project Id avoid using "iron".
A good alternative is the old GEC scheme where stray inductance is shunted
with a suitable capacitor: a LC coupling, improved with this simple "fix",
should be an effective solution.

The leakage inductance of an IST can be somewhat high,
and indeed there is the shunt C also involved, and they form a resonant circuit,

or rather an underdamped second order low pass filter, hence a peak in the
response
at over 5 kHz.

The "parasitic L and C components in the model of the IST are a fact of life
which cannot be avoided.

The traditional solution is to damp the LC critically with some R across the
sec,
but it loads the primary also, so that the triode driver ends up
being loaded by R at the mid-frequencies, something we might be trying to avoid,

to make the triode produce its lowest Dn.
So inded a snubber R+C network might be able to be connected across the sec, so
that at HF the
load is enough to damp the response critically, and yet at the mid frequency
the RC load has become a high and negligible impedance, thus allowing the triode

to get on with low Dn signal amplification.

Transformer Dn is the other possible problem, but it is reduced by using
low Ra triodes, and staying well away from pentodes.
The lower the Ra, the lower the Dn caused by the iron, which usually
inserts 3H into the signal path, even at low levels where it is worst as a of
the total signal,
and right in the range of where we want to do our listening.
Having 0.1 of 3H inserted into the signal at a watt of output isnt uncommon,
and it seems silly to try to make the tube caused thd lower than 0.05, and all
even order
but them be blighted by a similar amount of odd order crap.

So IST should be very carefully used indeed.

Patrick Turner.


BTW my 2A3 amp uses a plain LC coupling (with an EC86 driver) and I love its
sound, but it has serious bandwidth limits: I wouldnt suggest it as a
reference.

Ciao

Fabio

"Audiquattro" ha scritto nel messaggio
om...
Im looking at both CCS vs Interstage transformers.
Many folks like the CCS clean spotless sound.
However, some folks like the more articulate "lush" sound.
Since I have not build nor heard either one yet, Im looking for
opinion of both world.

Ill be using either 645 or 7788 or EC8010 with 1:1 interstage
transformers.
I was told that 417a do not have enough gain for 1:1 interstage
transformer.

TIA

-=-
This message was posted via two or more anonymous remailing services.

Hi ,

It looks like the CCS is a bit forgotten in this postings , but
a 7788 or E180F will make a nice CCS also .
You might like to read this :
http://www7.taosnet.com/f10/mustage.html

Ronald .


"Audiquattro" schreef in bericht
om...
Hi Pat,

6C45pi is also on the tube list. However, I want to experiment with
other alternatives as well. I have not look at 6EJ7 yet but
7788/E180F will be strapped in the triode.

thank you for your experiment with CCS. Which CCS did you use for
your amp?
- -- - - -- - - - - - - ---
Hi Andre, I'm designing from the speakers. I'm just curious what
other folks have tried in their experiment. I'm using the
Electra-Print 3K OPT and also got the RIT5 sr ITs.

My speakers are ranging from 94+ dbs (Jensens and Altecs) to 100+ dbs
(JBL horns). It's a good experiment for all the different setups I
can play with various output (from 3W to 8W) and I do have preamp
already. I also consider using the SV572 or the 300B+ tubes with the
417A down the road.

thank you for the long post regarding to the things you have tried.

George Orwell wrote:

Just checking to see if youre awake, Turner, or in auto-fulminate mode again.

What occasioned the spew below about SET and horns? The last time you told that story it was a Fostek driver in the horn, not a Lowther, and I demonstrated conclusively that you dont know **** about horns, so where do you come from sawing open speakers? Ever hear of asking for the plan instead of a saw, or hitting the books so you actually know what youre talking about?

Andre Jute

Who is fulminating now?

The owner of the horns I mentioned could have made mistakes with the woodwork.
Looking at the box plans may not have indicated where he may have gone wrong.
I brought horns into picture in support of the idea one has to be careful
to come to sweeping conclusions based on "how it sounds"
without having got the basic engineering right.

Patrick Turner.



nonymous wrote:

rec.audio.tubes
What do you prefer in your 300B SE?

What Patrick Turner intends to say on interstage transformers in his confused post below is this:

I am far from sure that my post was confused, or that I intended to say all of the following,
so allow me to expand on what you say I intended, lest anyone have any doubts....



The inductance of a transformer primary is a variable quantity.

True, and therefore the impedance changes during a single voltage cycle.

It varies from transformer to transformer and in a particular transformer with signal voltage.

Yes.

Using this inductance in a resonant circuit increases distortion.

I am sure I would never have intended to say this.
I did say that leakage inductance and stray capacitance get together to form a resonant circuit,
but I dont recall having said Lp resonates with anything.

Primary loading causes tube distortion and secondary loading increases transformer distortion. A resistor shunted across the secondary does smooth out high frequency response and a resistor shunted across the primary extends bass response. However, the primary inductance and the shunt resistor form a series resonant circuit which causes the tube to see a low impedance at the resonant frequency which at high outputs will cause serious distortion. In consequence good design avoids transformer shunts.

Nowhere have I ever said "that the primary inductance and the shunt resistor form a series resonant circuit"
It is impossible for any inductance and resistance to act as a resonant circuit.

Transformer shunts by resistance are usually only used to tame the peaks in the HF response
if there are any, and my optional recommendation included the use of an R+C
series network across the secondary, to provide critical damping of the inevitable
second order roll off of HF caused by leakage L and stray C.
The C does not conduct at mid frequencies, but does at above say 5 kHz,
and such a nework may tame the HF peaks.

RC damping may not always be needed, and perhaps not at all in the case of a very well
wound transformer.

So you have it from the horses mouth what was intended, rather than from your imagination.



Better to buy a good IT in the first instance, as Fabio suggests. He is wrong about good ones being ra the Lundahl catalogue lists a wide variety and so do other winders by appointment to the big broadcasters.

Some of the very best amplifiers I have heard were all transformer coupled,
but good ISTs are rare compared to OPTs.
If folks are willing to pay for them, there are plenty to buy.
Some of the worst amplifiers I have heard were SETs with all transformer coupling,
but done without the slightest attention to the response which turned out
to be attrocious, because the owner doesnt like measurements, and thinks he can
go on the sound quality only.

I recently listened to some really attrocious sound from some
Lowthers in home made horn enclosures, powered by 1.5 watt SET.

I measured the response, and the audio was about flat from over 10k to 5k, but
then rolled off at 3 dB/octave, and the only way to tame this terrible bright bassless
sound is by adding a network at the amp input.
There may still be something wrong with the woodwork of his speakers,
but I wasnt asked to go in with a saw to find out.
The same amps gave a flat response into some Paradigm speakers bought
from a local hi-fi store, and positioned alongside the Lowthers in the same room.
So had I not measured, the owner, who is completely technically untrained,
could have come to a wrong conclusion
about SET amplifiers best the ears be distrusted at all times, until at least the basic
engineering like a fairly flat response is obtained.

So whatever is built should be rigourously tested, to make sure there is at least a flat response
signal being sent to the speakers, and that the acoustic response is about flat.

Patrick Turner.



Andre Jute
A little, a very little thought will suffice John Maynard Keynes

Patrick Turner sent this:

Fabio Berutti wrote:

Interstage transformers are extremely rar these days, and i suppose it is
VERY difficult (and $$$$$) to get one which does not fail in terms of
bandwidth and phase rotations.

IST are not that rare at all. Sure, they cost, but so what?
Does somebody who owns a yact complain about occasionally having to spend money
in his pride and joy moored out in the bay?


Unless Youre talking of a very high budget project Id avoid using "iron".
A good alternative is the old GEC scheme where stray inductance is shunted
with a suitable capacitor: a LC coupling, improved with this simple "fix",
should be an effective solution.

The leakage inductance of an IST can be somewhat high,
and indeed there is the shunt C also involved, and they form a resonant circuit,

or rather an underdamped second order low pass filter, hence a peak in the
response
at over 5 kHz.

The "parasitic L and C components in the model of the IST are a fact of life
which cannot be avoided.

The traditional solution is to damp the LC critically with some R across the
sec,
but it loads the primary also, so that the triode driver ends up
being loaded by R at the mid-frequencies, something we might be trying to avoid,

to make the triode produce its lowest Dn.
So inded a snubber R+C network might be able to be connected across the sec, so
that at HF the
load is enough to damp the response critically, and yet at the mid frequency
the RC load has become a high and negligible impedance, thus allowing the triode

to get on with low Dn signal amplification.

Transformer Dn is the other possible problem, but it is reduced by using
low Ra triodes, and staying well away from pentodes.
The lower the Ra, the lower the Dn caused by the iron, which usually
inserts 3H into the signal path, even at low levels where it is worst as a of
the total signal,
and right in the range of where we want to do our listening.
Having 0.1 of 3H inserted into the signal at a watt of output isnt uncommon,
and it seems silly to try to make the tube caused thd lower than 0.05, and all
even order
but them be blighted by a similar amount of odd order crap.

So IST should be very carefully used indeed.

Patrick Turner.


BTW my 2A3 amp uses a plain LC coupling (with an EC86 driver) and I love its
sound, but it has serious bandwidth limits: I wouldnt suggest it as a
reference.

Ciao

Fabio

"Audiquattro" ha scritto nel messaggio
om...
Im looking at both CCS vs Interstage transformers.
Many folks like the CCS clean spotless sound.
However, some folks like the more articulate "lush" sound.
Since I have not build nor heard either one yet, Im looking for
opinion of both world.

Ill be using either 645 or 7788 or EC8010 with 1:1 interstage
transformers.
I was told that 417a do not have enough gain for 1:1 interstage
transformer.

TIA

-=-
This message was posted via two or more anonymous remailing services.

Ronald wrote:

Hi ,

It looks like the CCS is a bit forgotten in this postings , but
a 7788 or E180F will make a nice CCS also .
You might like to read this :
http://www7.taosnet.com/f10/mustage.html

Allan Kimmel provides us all with many thoughts
of interest re the series use of tubes ranging between
SRPP and mu follower.

I am rather fond of the "bootstrapped follower" gain stage
using any of the range of twin triodes, or or other pair of independant
triode tubes.
The top tube is can be biased at say +200v, to allow a wide undistorted
voltage swing,
if that is wanted to drive something like a 300B.
A pair of 6SN7 or 6CG7 could be used, with each triode section paralleled,
so one tube is used for the top triode, and used with a heater voltage
which
is biased at about +250v, and the bottom tube has its heater voltage biased

at about +30v, which my books say gives the lowest heater noise.
DC on the heaters is a great idea, and the same rules of biasing the
heaters apply.

The top triode has fixed bias applied via a 1M resistor to say a fixed
+250v.
The bottom tube has a 10k resistor from its anode to the top tube's
cathode.
If Ia totalled 8 mA, the voltage across the 10k is 80v, and Ea
of the bottom tube anode is about 180v.

The cathode of the bottom tube has an unbiased resistor to get the right
bias current
which will be the same for both tubes, regardless of the grid bias applied
to the top triode.
Because fixed bias is applied to the top tube, the DC drift is minimised.
With 180v across the bottom triode, you would need about 180v across the
top triode,
perhaps 20% more is also OK, so the B+ for the whole set up needs to be
about 450v at least.

The grid drive from the bottom tube anode to top tube grid is vai a normal
0.47 uf.
The pole introduced by this CR coupling is at 0.337 Hz, and the bass
performance will
remain unaffected.

The output is taken from the top tube's cathode, and is CR coupled to the
300B grid.

The voltage gain depends on the load the bottom triode sees at its anode,
and this will comprise the 1M bias R for the top triode in parallel
with the input resistance looking into the 10k.
If the load on the output top triode follower is say 100k, then its
open loop gain is about 16, so the load the bottom triode sees
is 160k, not the 10k, and with the 1M, the load is approximately
140k, and so its gain will be about 18.5, with a slight loss for
the cathode resistor current FB, so its overal gain would be about 16.3.

The overall gain of the top triode is about 16/17 = 0.94,
so the total overall gain = 0.94 x 16.3 = 15.3 approximately.

The distortion will be low because the load the bottom triode sees
is many times the approximate 5k of Ra of the two paralleled halves of the
6SN7.
Using a CCS as a load won't dramatically reduce the thd any further.

The distortion of the top triode follower will be about 1/16 of the normal
plate loaded condition with a 100k + load of the bottom tube,
some what an awkward concept to grasp, because who is the master, who is
the slave?
But some 2H cancellation of distortion does occur, but it is minor,
so expect the low distortion of the follower to not increase the distortion

of the gain tube, which should be able to operate with very low
distortion, and about a 70vrms voltage swing is easily possible,
and thd should be well under 1%, mainly due to the
fact that there is only 0.5 mA rms change in the bottom triode's signal
current.

The top tubes signal current change is greater because it provides the
lion's share of the
load current in the following cap coupled bias resistor.

Ronald .

"Audiquattro" schreef in bericht
om...
Hi Pat,

6C45pi is also on the tube list. However, I want to experiment with
other alternatives as well. I have not look at 6EJ7 yet but
7788/E180F will be strapped in the triode.

EF80 or 6BX6 is a common sharp cut off pentode also very suitable
as a triode.

The most gutsy 9 pin frame grid signal pentode I can recall just now is the

E280F, but where you find a pair these days is beyond me.

The frame grid pentodes have lots of Gm, and
as triodes they usually become medium Ra tubes, rather like 12AT7,
but with much higher U than a 12AU7, 6CG7, 6SN7, etc,
and so that may allow them to be used as drivers and with enough gain
without using an extra input stage.

And in the case of the bootstapped follower, the value of Ra of the tube is

relatively unimportant because the load is connected to the low output
resistance of the top triode's cathode; this Ro is nearly as low
as a real cathode follower.
I doubt there is anything to be gained by arranging the top tube
to act as a pure pentode follower, but if it is, with a screen supply
kept at the same signal level as its cathode, then the screen supply
resistor is seen
by the bottom gain tube, and some of the purity of the high RL value for
the
bottom triode is lost, so I can't recommend the pentode use of the top
tube.
Too complex, and nothing much to benefit....
And I doubt you'd hear much difference.

Patrick Turner.



thank you for your experiment with CCS. Which CCS did you use for
your amp?
- -- - - -- - - - - - - ---
Hi Andre, I'm designing from the speakers. I'm just curious what
other folks have tried in their experiment. I'm using the
Electra-Print 3K OPT and also got the RIT5 sr ITs.

My speakers are ranging from 94+ dbs (Jensens and Altecs) to 100+ dbs
(JBL horns). It's a good experiment for all the different setups I
can play with various output (from 3W to 8W) and I do have preamp
already. I also consider using the SV572 or the 300B+ tubes with the
417A down the road.

thank you for the long post regarding to the things you have tried.

Patrick Turner wrote:

Anonymous wrote:

Patrick Turner wrote:

No doubt there will be a time soon where Ill just have to ignore you

Ive been explaining to you for three years now, you boneheaded bricklayer, that I want you to put me in your killfile. Even the remnant of your own gang has been praying out loud for you to get some sense. Lord Valve, who despises you as heartily as I do, has most kindly taken time out from shopping the gun catalogues to explain to you that you cannot win. Everyone is bored ****less by your futile attempts to dragoon me into your narrow, aridly joyless view of hi-fi. I am glad that at last you are getting the message. Put me in your killfile today. Youll feel so much better when you can no longer see me talking about Class A amps, SET, point source speakers, electrostats, originality, art, joy, and other gear and concepts you do not have the hang of and never will get the hang of because you are the wrong sort of person. RAT will be a more agreeable place without your squealing every time I step on your tail.

I don't have a kill file, I am a realist, and I deal with all the muck as it comes along.
I like to see every single post of the news group.
I have no intention of ever avoiding entirely the incorrect
technical and personal assesments you make repeatedly, and I hope to be still here in 30 years,
so you better get used to it, or just **** off.

Its no good trying to snuggle up behind Lord Valve, and I don't care what he feels about my presence
here. The fact is I agree with LV about many tube issues, but he and I disagree on politics,
but we don't need to conduct a daily war over it.
And we surely don't stoop to denigrating each other's business activities.
Only morons do that.
LV has an extremely small amount to say on audio technical matters.
So take a leaf out of LV's book, and unless you are gonna show
some real style around here, and stick with the audio issues only, then shut the **** up.

Quite a few of my clients and friends use SET amps and some use horns, or try to,
and all the agreable things you take so seriously are matters which I also take seriously.

In your last post you laid the boot into me and my business ( which has never done it any harm )
so my reaction is to warn everyone
about what you are like, since we get a few newcomers here who just don't realise
what an arsole you are when they disagree with you, or challenge the validity of your
opinions.

Andre Jute has not the slightest monopoly on correct audio thinking, wu, or anything
else, he just has his own opinion, and he is just another little human being like each of
the rest of us who is trying to battle his way through life,
only he does it with such a grating style that nearly everyone is mightly
peeved to witness it.

I strongly suggest Andre write a book about tube audio and horn speakers, and thus be useful
to allow folks to obtain great sound from horns.
A book sells for $40, so there's a silver lining to such an endeavour.

He says he knows so much more than I do on the subject, so let him proove it,
lest we all think he's a dunce.

Patrick Turner.