SY said:
If you're going to put the FET on the bottom, might as well put one on top as well. Makes for the same sort of pentode-ish construct...or just use a pentode; that'll do jus' fine. The cascode element could be of *VERY* low capacitance; the lower element controls this, in concert with the d-s voltage likely established by battery biasing the upper one ~10V over it.
Also, according to the Philips triode data, most of the curvature is over with by 10 mA plate/g2 current, and by 20 mA it is quite a straight line v g1 voltage. That is hardly worse than a 12B4. I also drew a vertical load through tis curves at 100V; the spacing is quite even, as long as we stay away from the bottom( but that is also true for any load line steepness).
cheers,
Douglas
I've done it with a jFET (2N5566) on the bottom and a MOSFET (IRF820) on the top. It was indeed quite linear and swings a lot of volts. Tube fans wouldn't be pleased...
One thing I never got to work but seems worth exploring is to reference the gate voltages of the top devices to the junction of the sources of the bottom devices instead of ground. A battery might work in that application.
One thing I never got to work but seems worth exploring is to reference the gate voltages of the top devices to the junction of the sources of the bottom devices instead of ground. A battery might work in that application.
SY said:
hey-Hey!!!,
A voltage divider will serve well here. Bypass the bottom R( with the ref voltage across it ). I've had good results with this method with triodes on the bottom, and there is no reason it would not work just as well with FET's.
AFIK, there is no reason, sonic or otherwise to avoid the FET on top aside from Religious Fanaticism if a FET is used on the bottom...matter of fact, a fine enhancement-mode heaterless pentode amp can be made using the cascode element.
IMO, the gate capacitance variation due to varying d-s voltage is responsible for any issues I have with 'SS sound'. A good pentode design can be directly implemented with these devices; just be careful of D-S voltage ratings...for the power stage, 900 or 1k2V devices( on the 'top' element) would be very useful...there is no recovery from 'red-plating' a MOSFET...
cheers,
Douglas
SY said:
How about a straight BJT Darlington on the bottom and triode on top, allow a small DC offset on the input, quite small. Again some won't be pleased but is really does work. Use 2SC2547s. BTW, have you tried IRF710, a lot less capacitance and slightly less V and I rating.
Come again with that one? Hmm, that's interesting, but definitely fets on the bottom, otherwise a lot of batts in series. If source resistors, then reference it to the top of CCS.
But does it have to be battery? Look at above Vacuum State schematic and the use of LM329 6.9V @ 8mA - connect to CCS and any current thru LM329 allow an extra 8mA in CCS, otherwise your Anodes will go up.
Joe R.
The 329 approach is exactly the one I took, but it never worked right. So, it makes sense to revisit with batteries, which removes the issue of current. For me, it makes the most sense to have the SS stuff on the bottom, which makes the battery requirement simpler. In fact, you don't want the Vgd to be more than 7V or so anyway, else gate current starts to increase.
The IRF820 was the best device I could get in 1985 when I was doing these experiments. No doubt there are better choices now. I needed the extra voltage rating since the circuit was supposed to drive a screen drive output stage. That said, the capacitances are much less of an issue with the top devices- the gates are either at ground or a low impedance source. Bandwidth of the stage was superb- if memory serves, rise time was something under a microsecond.
The IRF820 was the best device I could get in 1985 when I was doing these experiments. No doubt there are better choices now. I needed the extra voltage rating since the circuit was supposed to drive a screen drive output stage. That said, the capacitances are much less of an issue with the top devices- the gates are either at ground or a low impedance source. Bandwidth of the stage was superb- if memory serves, rise time was something under a microsecond.
SY said:
I did an all fet cascode diff design microphone preamp and I seem to recall Allen thought it was quite cool. We have co-developed several SS designs including this:
www.audiorevelation.com/cre/product_info.php?manufacturers_id=41&products_id=191
Joe R.
Bandersnatch said:
Well, it wasn't. Since this was for audio, the main consideration here was linearity. Gain and BW were what they were after finding a decent Loadline.
Fine ideas for RF amps where you don't care about linearity since you have LC tuners and/or BPFs to reject the resulting harmonic distortion products. also fine for BJT based circuits where the only limit to g(m) is how much current you can pull before the die melts, and where you wrap huge amounts of gNFB around the thing to force it into linearity. Not so swell with hollow state implementations where excessive levels of gNFB are either not possible, and are definitely undesireable since it makes for a very solid statey sound.
gallon said:
Hi, Gary, nice to see you posting here!
As for the WE nomenclature, I frequently confuse the two names - "Harmonic Balancer" vs "Harmonic Equalizer". I think it's because I've seen it referred to both ways. If your documents say "Harmonic Balancer", that's good enough for me. The WE circuit is really obscure, after all, and very few people have ever heard about it until I started mentioning it on the Nutshell webpage. It's still almost unknown, and the mid-Thirties source documents are very rare.
As for Point Two, Steve rescued the "Matrix Amplifier" from an old article - Popular Electronics, something like that - and the source circuit was one of those wacky late-Fifties save-a-few-pennies projects to get people into stereo as cheaply as possible. Thus, the L+R signal was full PP, while the L-R channel kind of looked like SE. Zany. I'm sure it would sound "different" - as for better, I dunno. I have mixed feelings about electronics that have completely different harmonic profiles for the sum and difference signals, as the Matrix amps would be certain to have. I like L and R to be completely separate with their own power supplies, not commingled together.
I haven't heard Steve's latest, but he's posted elsewhere that the WEHE is his current favorite amplifier. It certainly had impressive measurements and excellent dynamic capabilities for a zero-feedback PP amplifier. I was digging all through his website, but I couldn't find the schematic for his latest, which is unfortunate, since the performance is truly outstanding, and the ingenious circuit got around the awkward transformer-coupling requirements of the classic WEHE circuit.
It also let you exactly tune the WEHE ratio to the harmonic profile of the output tubes - moderate for clean-as-a-whistle DHT's, and more for triode-connected pentodes or straight pentodes. The ability to independently adjust only odd harmonics is a wonderful thing for any PP amplifier.
If anybody knows Steve, we should encourage him to post his WEHE amplifier on his webpage or maybe, if we're lucky, right here. Same for any old WE Bell Record or American Projectionist source articles. I personally feel this is best possible topology for PP amplifiers.
Lynn Olson said:
I seem to be hosting the copy of Steve Bench's site that everyone is using at the moment, on what is otherwise the most infrequently updated website in the world (greygum.net). I've done this as AOL pulled the plug on his old hosting arrangement, but it is without his knowledge or explicit permission (implicit permission to do this is in the text of his site). If Steve wanted to update the site I can give him an account on the server etc, I just have no way to contact him. Please pass this on if you know him.
Lynn, et al,
Thanks for your comments. Regarding the matrix, I tend to agree with your general assessment of the concept. Steve’s original source for his first matrix amp was an old article he had saved from his tinkering youth days. It definitely has the pulp magazine flavor, for that is exactly what it was. Steve took this tinkerer’s experiment and refined it to a high art. Once Steve gave it the full Bench treatment, it was a worthy modern design. As you know, with Steve it is all about the final sound. This seems to be a common attribute shared by all the great diy designers.
Steve’s 813 matrix was his main listening amp for several years. More recently it has been replaced by his fabulous PP 801 WEHE amp. Perhaps we can get him to stop by in here and comment on each of those.
I’ll take your hint and make available some of the original papers about the harmonic equalizer and matrix.
Original matrix article:
Bourne, A.G., “A Stereo Amplifier,” Practical Wireless, vol. 35,
pp. 673-674, Dec. 1959.
WEHE
Nadell, Aaron, "The W.E. 86 Type Amplifier, Unit of the 'Mirrophonic System'," International Projectionist, vol. 11, pp. 21-24, December 1936.
Nadell was a technical writer for the IP, possibly the leading review voice for theater audio and video technology of the 1930’s. Here he gives a better walk through of the WE86 than Cousins. He refers to the harmonic equalizer by name. The term WEHE was coined just in the last several years by the present diy crowd.
Cousins, V.M., "The W.E. 86 Type Amplifier," International Projectionist, vol. 11, pp. 20-22, August 1936.
Cousins was a member of the technical staff at Bell Laboratories. Was probably a member of the design team for the WE86 and WE91. This paper is a short introduction to the WE86, wherein he doesn’t specifically refer to the harmonic equalizer.
There was also an engineering paper about this, published in the BSTJ by Ketchledge in 1955:
Ketchledge, R.W., "Distortion in Feedback Amplifiers", The Bell System Technical Journal, Vol. XXXIV No. 6, pp. 1265-1285, Nov. 1955.
This paper is written at the BSTJ level, full of arithmetic. Not much useful to the diy’er. A competent engineer should be able to give this a go however. I have the Ketchledge paper but it is not my scan, so I do not have the freedom to share it.
There is a further patent out there, which also bears:
US Patent 1,970,325 by J.G. Kreer, Jr.
Thanks for your comments. Regarding the matrix, I tend to agree with your general assessment of the concept. Steve’s original source for his first matrix amp was an old article he had saved from his tinkering youth days. It definitely has the pulp magazine flavor, for that is exactly what it was. Steve took this tinkerer’s experiment and refined it to a high art. Once Steve gave it the full Bench treatment, it was a worthy modern design. As you know, with Steve it is all about the final sound. This seems to be a common attribute shared by all the great diy designers.
Steve’s 813 matrix was his main listening amp for several years. More recently it has been replaced by his fabulous PP 801 WEHE amp. Perhaps we can get him to stop by in here and comment on each of those.
I’ll take your hint and make available some of the original papers about the harmonic equalizer and matrix.
Original matrix article:
Bourne, A.G., “A Stereo Amplifier,” Practical Wireless, vol. 35,
pp. 673-674, Dec. 1959.
WEHE
Nadell, Aaron, "The W.E. 86 Type Amplifier, Unit of the 'Mirrophonic System'," International Projectionist, vol. 11, pp. 21-24, December 1936.
Nadell was a technical writer for the IP, possibly the leading review voice for theater audio and video technology of the 1930’s. Here he gives a better walk through of the WE86 than Cousins. He refers to the harmonic equalizer by name. The term WEHE was coined just in the last several years by the present diy crowd.
Cousins, V.M., "The W.E. 86 Type Amplifier," International Projectionist, vol. 11, pp. 20-22, August 1936.
Cousins was a member of the technical staff at Bell Laboratories. Was probably a member of the design team for the WE86 and WE91. This paper is a short introduction to the WE86, wherein he doesn’t specifically refer to the harmonic equalizer.
There was also an engineering paper about this, published in the BSTJ by Ketchledge in 1955:
Ketchledge, R.W., "Distortion in Feedback Amplifiers", The Bell System Technical Journal, Vol. XXXIV No. 6, pp. 1265-1285, Nov. 1955.
This paper is written at the BSTJ level, full of arithmetic. Not much useful to the diy’er. A competent engineer should be able to give this a go however. I have the Ketchledge paper but it is not my scan, so I do not have the freedom to share it.
There is a further patent out there, which also bears:
US Patent 1,970,325 by J.G. Kreer, Jr.
http://www.intactaudio.com/forum/viewtopic.php?t=549
Is it the WEHE amp that you guys are talking about?
Is it the WEHE amp that you guys are talking about?
Thanks for all the research, guys!
Xenu, here's the answer to your question: the WEHE was lost for more than 50 years, re-discovered by a handful of people a few years ago, and Steve Bench has given us a beautiful realization of the concept. You will NOT find the WEHE in any commercial amplifier, at any price, thanks to the notorious Not Invented Here bias of the commercial sector.
You want the cutting edge of audio, well, look right here. The DIY sector found the WEHE, not the fancy-pants high-end guys with their $80,000 amplifiers and four-color full-page ads in the magazines.
Xenu, here's the answer to your question: the WEHE was lost for more than 50 years, re-discovered by a handful of people a few years ago, and Steve Bench has given us a beautiful realization of the concept. You will NOT find the WEHE in any commercial amplifier, at any price, thanks to the notorious Not Invented Here bias of the commercial sector.
You want the cutting edge of audio, well, look right here. The DIY sector found the WEHE, not the fancy-pants high-end guys with their $80,000 amplifiers and four-color full-page ads in the magazines.