What do say: MJK, Kenpeter and others? Seems like this has been around here for a long time.....
http://www.tubecad.com/2009/09/blog0171.htm
http://www.tubecad.com/2009/09/blog0171.htm
Broskie did show a balanced, load splitting series push-pull circuit, based on tubes, in May 2000, which operates on the same principle as the "series" anti-triode. He gives the formula correctly for the offset on the center resistor tap based on the cathode resistance (~1/gm) of the slave device. Anyone who's thought about SRPP as much as JRB has, should come around to this eventually.
Current sensing master-slave push-pull is certainly not a new idea. The series anti-triode is toplogically like SRPP and the parallel, cathode sensing anti-triode is topologically like a differential LTP, with similar current loops. The plate sensing versions also have lots of prior art and some patents. One thing different about our work is that we are exploring the sonic aspects of high-gm slave devices as the "anti-triode".
What we have also done here is explore even more topology variations (for example fixed or servo-controlled DC output, developed different control and feedback strategies, and done a lot of practical work on real world implementations.
For anyone wanting to understand the current loop operation of this circuit, JRB gives a good detailed explanation with math:
http://www.tubecad.com/may2000/page12.html
Cheers,
Michael
PS after 30 years working in computer architecture, on "new" ideas, it clear to me that ideas are a dime a dozen. Outside of theoretical physics, the real challenge is in getting stuff to work.
I agree with Michael, the driving idea behind "anti-triode" was preservation of the triode signature. This pointed to approaches using high gm devices, gm boosters, Op Amp loops etc, to preserve the spectra. The modified SRPP or "SRPP+" (first time I have seen that, but a nice article) and modified LTP topologies were convenient ways to implement this.
I wouldn't be surprised if some lab researcher came across the same ideas long before us, maybe even published them in some obscure journal. But long forgotten in the economic practicalities of the audio market. Hawksford Error Correction, for example, was invented and patented long ago by LLewelyn for tube circuits (Hawksford noted the origin), but ever see a tube amp with EC? The current mirror and voltage mirror circuits were invented by some Japanese researcher using tubes, but lost in the noise when neg. feedback was discovered. The Aikido first stage is the classic voltage mirror. If you really want to be impressed with lost tube technology, look thru some early analog computer texts using tubes. Simply amazing what could be done.
Don
I wouldn't be surprised if some lab researcher came across the same ideas long before us, maybe even published them in some obscure journal. But long forgotten in the economic practicalities of the audio market. Hawksford Error Correction, for example, was invented and patented long ago by LLewelyn for tube circuits (Hawksford noted the origin), but ever see a tube amp with EC? The current mirror and voltage mirror circuits were invented by some Japanese researcher using tubes, but lost in the noise when neg. feedback was discovered. The Aikido first stage is the classic voltage mirror. If you really want to be impressed with lost tube technology, look thru some early analog computer texts using tubes. Simply amazing what could be done.
Don
Last edited:
Sent him a Differential White Cathode Follower design for comments a few
months back. I thought a simple twin triode circuit with near perfect CMRR
and PSRR (maybe useful for DAC?) would have been right up his alley?
Never heard anything back... Not a peep.
I never mentioned anything regarding the Anti-Triode. Didn't want him to
think I'm completely nuts. Not yet anyways...
Mind you, this sim run was undertaken with HUGE amounts of power supply
and common mode noise. If it looks a little squiggly, theres good reason.
When all the crud magically cancelled out, I immediately thought "Broskie!".
Maybe I don't understand the guy, thought cancellations was his thing???
I would never have predicted him to take interest in Anti-triode SEPP.
months back. I thought a simple twin triode circuit with near perfect CMRR
and PSRR (maybe useful for DAC?) would have been right up his alley?
Never heard anything back... Not a peep.
I never mentioned anything regarding the Anti-Triode. Didn't want him to
think I'm completely nuts. Not yet anyways...
Mind you, this sim run was undertaken with HUGE amounts of power supply
and common mode noise. If it looks a little squiggly, theres good reason.
When all the crud magically cancelled out, I immediately thought "Broskie!".
Maybe I don't understand the guy, thought cancellations was his thing???
I would never have predicted him to take interest in Anti-triode SEPP.
Attachments
Last edited:
Does anyone have a copy or know where I can access the info?
Yes, I would like to. However before I go lurking in second-hand book shops are there any reference sites available on the 'net?
The info on this site is very interesting (augmented cathode followers, etc.) http://www.jrossmacdonald.com/ but there must be others?
J.R.s papers on "Active-Error Feedback" and "Complementary Distortion" are of some interest also.
I'm not familiar with any Analog Computer sites on the WWW, but there could be a few. I know Dr. Lundsberg at MIT is a collector of analog computing history.
For books, here is one I have, there are others, but I don't recall them off hand:
"Analogue and Hybrid Computers" by Z. Nenadal and B. Mirtes
OH, also check chapter 26 of "Amplifier Handbook" ed. Richard F. Shea "Non-Linear Amplifiers" by William M. Tremholme. Interesting beam deflection tube computing among others.
http://www.essex.ac.uk/csee/research/audio_lab/malcolms_publications.html
Also search on Hawksford on DIYaudio. Try Llewellyn too. Mostly in the SS section. And Llewellyn's patent: 2245598
http://patent2pdf.com/
Malcolm's papers are the way to understanding the modern concept of EC.
I'm not familiar with any Analog Computer sites on the WWW, but there could be a few. I know Dr. Lundsberg at MIT is a collector of analog computing history.
For books, here is one I have, there are others, but I don't recall them off hand:
"Analogue and Hybrid Computers" by Z. Nenadal and B. Mirtes
OH, also check chapter 26 of "Amplifier Handbook" ed. Richard F. Shea "Non-Linear Amplifiers" by William M. Tremholme. Interesting beam deflection tube computing among others.
http://www.essex.ac.uk/csee/research/audio_lab/malcolms_publications.html
Also search on Hawksford on DIYaudio. Try Llewellyn too. Mostly in the SS section. And Llewellyn's patent: 2245598
http://patent2pdf.com/
Malcolm's papers are the way to understanding the modern concept of EC.
Last edited:
He is smart, creative, and full of energy, so he could come to anything we design here independently. I would be glad if somebody invites him here.
He wants to be lonely... It is quite different to what I want: I want to work in a collectivity. Forum is not enough against sensory deprivation.
May be some day forumers get together and organize some great research and development institute, with manufacturing, marketing, sales subsidiaries...
edit, (too late to get into post 11):
Analog computing history: http://web.mit.edu/klund/www/
He also editted a special issue of IEEE Control Systems Magazine on the history of analog computing (the June 2005 issue) ---- likely to have some references.
Analog computing history: http://web.mit.edu/klund/www/
He also editted a special issue of IEEE Control Systems Magazine on the history of analog computing (the June 2005 issue) ---- likely to have some references.
It may be a god reading for G.K., he exploits modern fashionable transistor topologies adding elements from ancient analog computers. Good exercise!
Member
Joined 2009
Paid Member
This forum is already a great research and development institute and many commercial enterprises benefit from it - didn't somebody say there are chinese amplifier manufacturers combing these pages as we write... ?
Why should we give them such an advantage to make poorly understood thingies if we can hire them for manufacturing building better and better things?
What if some Chinese plant's representatives chime in asking for designs?
I have already some prototypes that can be easily re-engineered according to available technologies and volumes of production.
I looked at the patent but the maths is beyond me! Oh well, I'll keep searching out other info. Anyway thanks for the pointer and links, which are much appreciated.
Because then they have to pay royalties. There's more profit margin in stealing a "proven" design, building a prototype to validate that it works (although maybe not enough understanding to know how well), and then cranking out scads of copies while eliminating the middle man. Plus no pesky single client who does quality control before selling them. Reduce quality over time to increase profit, change names a few times, make outrageous claims - basic Ferrengi capitalism in China.
I've e-mailed Broskie before and he's answered my (admittedly) ignorant questions very quickly and in great detail. I've pretty much combed his entire site and found it to be a vast wealth of knowledge. He likes to take a subject and run it down to it's end point, then back up a bit and run down another rabbit trail. It may not be the most organized or quick method, but it is thorough and well written. I understand why he dislikes forums, too much opinion and bickering to get to the heart of the matter. He's much more disciplined and inward focused, very academic in his outlook. And there's very little new under the sun, probably every circuit topology has been tried many times, almost all documented, and then proptly forgotten or ignored except for a few, usually the simplist to understand and most cost effective to implement. We have to re-learn these things constantly, and I'm thankful for the internet that allows us to communicate easily and to study the works of great minds past and present.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.