View Single Post
  #51   Report Post  
Posted to rec.audio.tubes
[email protected] wfc5687@gmail.com is offline
external usenet poster
 
Posts: 3
Default Questions for Pat Turner

Hi Pat!
My friend just show me this amplifier recently and this is capable 64 ohm.

http://mcc.berners.ch/power-amplifiers/MC3500.pdf

Is that ok as I come across an used pair with reasonable price?

Regards,
J

On Wednesday, June 19, 2013 2:30:49 PM UTC+8, Patrick Turner wrote:
On Monday, 17 June 2013 13:52:32 UTC+10, Jimmy Leung wrote:

Dear Pat, I read your article about Quad ESL and tube amplifier online. I've discussion recently with my friends about stacking Quad ESL 57 and some even thinking to stack 4 x ESL 57 in a big room and I'm interested in that since I have 2 in stacking now. They suggest wired in series for the stacking....and sent me this link: http://forum.audiogon.com/cgi-bin/fr...openflup&5&4#5 I wonder 3 ESL 57 stacking and series wired still ok since the impedance still 30 ohm range. However, 4 in stacking, no kidding the impedance will rise to 60 ohm and maximum can go up to 120 ohm on the low frequency. Some even says 300B push pull can drive it!! Gee...are they making a joke with me?? Regards, JL




I agree fully with what Phil said about driving one pair of ESL57.

There is an impedance curve for ESL57 aboubt 1/10 down the page at my website http://www.turneraudio.com.au/quad2powerampmods.htm



Indeed 300B tubes can drive any speaker you could ever possibly imagine.

But you will need to know about how many 300B to use, and how to hook them up, and I suggest you have a very steep learning curve ahead of you. You are the one who wonders if your friends are making a joke with you about 300B, and this implies to me that you may have attrociously inadequate understanding of very basic issues regarding impedance, resistance, capacitance, inductance, load matching, amp design, etc, etc, etc. My knowledge was attrocious once, but gradually I learnt how get fabulous sound while avoid silly un-necessary expense on amps and never having PDEs, ie, "smoke producing events".



I repeat what Phil said....

** As a long time owner of ESL57s and an audio service tech, I can offer

the following advice:



**1. The impedance of the speaker is nominally 15 ohms ( 250Hz to 5kHz) -

falling smoothly above 5kHz to around 2ohms at 19kHz. The highest value

reached is 30 ohms at 100Hz. Down at 20 Hz ( no usable sound ) the impedance

is 8 ohms.



**2. Normal matching with a hi-fi valve/tube amp is to use the 15 ohm setting

and rely on NFB to keep response flat to within a dB or so across the audio

band.



**3. Amplifiers with high output impedance (ie low or no NFB) are not

sensible with ESL57s cos the sound quality will be rather dull. The same is

true with other speakers, for example the AR11 ( 12 inch 3-way) is much

worse than the ESL57 when used with such amps.



**4. Series operation has many drawbacks as the impedance gets ridiculously

high and input power will not be shared evenly by non identical speakers.

This could lead to sparking and damage.



**5. The ideal way to drive multiple ESL57s is to use multiple tube

amplifiers that use lots of NFB.



6. A good solid state amp can easily drive two ESL57s in parallel - the

**Quad 306 is perfect.



**7. Stacked ESL57s need to be individually aimed at the listening point, so

NOT mounted so as to make a smooth curve.



**8. All the above is based on lots of real experience and testing - NOT

THEORY.



NOW, I cannot see the slightest benefit of using more than 3 stacked pairs of ESL57, but if you had 4 stacked pairs then in theory you can have two pairs in series, then connect these two pairs in parallel and you get the same impedence as ONE ESL57, and described by Phil. I recall my website has an impedance curve for ESL57.



I had a customer with 3 stacked pairs in parallel and who bought an 8585 amp from me and the details are all at http://www.turneraudio.com.au/8585-a...ober-2006.html



The amp was fitted with 8 x KT90 and could give 100W into 4 ohms, with good load tolerance for where the 3 x parallel ESL57 impedance falls to about 0.6 ohms at about 18kHz, where average audio level is very much lower than in the middle of the LF band where most energy is needed, say around 200Hz..



The idea of having 3 of 4 x ESL57 in series creates the problem of having enough voltage available at the amp to give sufficient drive signal.

Quad-II amps capable of about 20 Watts into 16 ohms means that 18Vrms is needed at speakers. If you had 4 in series, you need 72Vrms across the 4, and virtually no amps are made to produce such a voltage unless they are say 1,000Watt rated for PA use. One might use a speaker *step UP* matching transformer but the toridal type at zeroimpedance.com would suffer saturation at a frequency that would be too high.



If you had all 3 or all 4 ESL in parallel, then the use of toriodal *step DOWN* speaker matching tranny with 2:1 TR ratio might be used although the response at HF may suffer because leakage L of the tranny is effectively in series with the very low Z at ESL HF.



The other thing you MUST realize is that with 3 of 4 stacked pairs, the speakers begin to resemble line arrays of dyneamic drivers mounted in vertical lines. The overall efficiency increases, so the POWER needed to make a given SPL for average listening becomes less. With the right sort of wide band speaker matching tranny 2, 3, or 4 pairs of ESL57 in parallel can easily be driven with say 2 x KT88 arranged to give 50W max. With multiple stacked pairs, your need for more power is only slight, to cater for the rare even when you might want more SPL than can be achieved with Quad-II and one pair of ESL57.



The ESL57 Z varies from say 32 ohms down to 2 ohms, and they are designed for a flat response with a constant voltage level. So thus if the amp makes 4Vrms at 50Hz, 32 ohms, it means Po = 0.5Watts, and the same 4Vrms at 18kHz, 2 ohms, Po = 8 Watts. Where you had 4 pairs ESL57 in parallel the Z is 1/4 of ONE ES57, and this may well cook some amps unless they have variable impedance settings allowing nominal speaker Z = 2 ohms for between 200Hz and 1kHz.



Patrick Turner.