which modern schematic is the benchmark based on the Williamson design?
It seems some are saying that the original schematic from 1947 is no longer the standard because of the amount of negative feedback and because it was designed for mono audio.
So what is considered to be today's standard schematic (the ultimate benchmark)? I'm guessing that maybe some minimal revisions have been made to the 1947 version so as to preserve it as much as possible while making it applicable to modern stereo systems.....and to take advantage of any improved industry knowledge. I'm guessing the tube amp industry must have learned something in the last 66 years.
It seems some are saying that the original schematic from 1947 is no longer the standard because of the amount of negative feedback and because it was designed for mono audio.
So what is considered to be today's standard schematic (the ultimate benchmark)? I'm guessing that maybe some minimal revisions have been made to the 1947 version so as to preserve it as much as possible while making it applicable to modern stereo systems.....and to take advantage of any improved industry knowledge. I'm guessing the tube amp industry must have learned something in the last 66 years.
I think that it is safe to say that there is no modern benchmark Williamson circuit. Over the years there have been quite a few variations using different tubes and different biasing of the original tubes, but having built many Williamson amps 20 or 30 years ago, my opinion is that there are now better topologies to drive a PP pair of output tubes. I would suggest the Mullard 5-20 circuit with a CCS in the LTP tail and a triode front end. MOSFET followers can be used for high power tubes. You can tune the amount of feedback empirically and of course the amp doesn't care if it is mono or stereo; just use two for stereo.
"Some" say all sorts of silly things. A few modern upgrades (e.g., solid state regulated supply) and it's a great performer, assuming you want an amplifier that amplifies rather than editorializing. Morgan Jones did a pretty complete analysis of the circuit in "Valve Amplifiers," worth a read.
Berning's EA-230 amp of the 1980s was an unusual version that ran astoundingly cool and efficiently. The schematic is worthy of study.
Hi,
Except it isn't any sort of version of the Williamson and
he quickly discounts that topology as being sensible.
Nice design though, but needs specific transformers.
rgds, sreten.
FWIW the Williamson famously hit the 0.1% distortion
mark first, but in retrospect its fairly meaningless.
Done by high levels of very marginal feedback.
Last edited:
The Philips 5-20W (virtually a Mullard copy) aka 1960 uses 30dB global feedback; so using feedback as an argument tool bears no ground to vintage, however sound quality, yes. Even many early 10-14W amps using EL84's had shocking values of global feedback. I repaired one in the 1980's which used 35dB global nfb plus anode to anode feedback loops !
Modern day Williamsons, eg my big 500W ones have over 20dB stability headroom over the nominal 20dB in the global feedback path, before instability sets in at both ends of the audio spectrum; you are right in mentioning that many earlier amps were suspect with just a 6dB stability margin over nominal. In the meantime alot has been improved, esp o/p transformers and phase shifting calculations become much easier to work out.
richy
Eddie, you are in an not so good path here.
First, you want to learn.
Then, you drop that, and look for a software that will accept voice commands "design me a KT88 PPamp"
Then you are looking for the "bestest" schematic.
Before long, you will go looking for the "best" capacitor...
Better spend some more time building a simple Single Ended. It is both rewarding and educational.
My 2 Yuro cents.
First, you want to learn.
Then, you drop that, and look for a software that will accept voice commands "design me a KT88 PPamp"
Then you are looking for the "bestest" schematic.
Before long, you will go looking for the "best" capacitor...
Better spend some more time building a simple Single Ended. It is both rewarding and educational.
My 2 Yuro cents.
A few years ago, I restored a Crown SXA. This dates from about 1960 and is not a million miles away from the Williamson with a UL output stage tacked on. (Hafler and Keroes themselves recommended this mod.)
scopeblog Crown Super SXA
Modern tube design (something of an oxymoron? 🙂 ) tends towards fewer stages and less global NFB. Once you shun global NFB, you find that you have to use a triode output stage to get adequate speaker damping, or use a speaker with more mechanical damping than normal.
scopeblog Crown Super SXA
Modern tube design (something of an oxymoron? 🙂 ) tends towards fewer stages and less global NFB. Once you shun global NFB, you find that you have to use a triode output stage to get adequate speaker damping, or use a speaker with more mechanical damping than normal.
Last edited:
What can be "shocking" about global feedback, if properly applied? Have you ever thought about how much feedback is used at LF in almost any opamp audio circuit or SS power amp?richwalters said:
Blackophobia is shocking.
The "if properly applied" is quite a big if, though. There are a lot of pitfalls especially in valve amps. If you don't believe me, just try driving a Williamson into clipping. 🙂
There is a modern implantation of the Williamson but I can't find it at the moment. It is in one of these threads here somewhere where it has almost become DC coupled. It's available through one of the companies that still make the O/P transformer. If I find it I'll come back.
It is always useful to provide a link when making a suggestion like that. The only article I could find by Mr. Byrith is here. I can't help but notice that it is a resurrection of the venerable Mullard 5-20 not the Williamson.
It seems you have missed the point here - most people are building with tubes to avoid that outcome. Not your choice (and you would maintain a wrong choice) but a valid one all the same.
If you want to argue religion I am sure another thread on the merit of gNFB could be started.
Shoog
People who build with tubes to avoid negative feedback usually don't understand tubes or negative feedback. You would need an amplifier with no triodes, UL, degeneration or cathode followers. Such animals do exist: single UL84 pentode amplifying a ceramic cartridge in some Dansettes? Quite rare though, and not serious audio.
There are many amps which use feedback and zero gNFB. The argument is about application of types.
Shoog
Shoog
Big differences to todays speakers compared to the previous generations of tube amps which used a single cone and had only minimal crossover phase shift with very low feedback compared to chip stuff these days, ie the 6dB/oct rolloff was very tame for global feedback stability compared to the mid 1980's 24dB/oct Splendor systems. This was a severe test in tube amp stability as many found out when electrostatics and even worse when piezo's came on the market.
What one forgets is it becomes "shockingly applied" if an incompatible LS system is connected to a tube amp which uses unforgiving amount of global nfb to ride it's shortcomings PLUS instability ! Perhaps, initially I didn't make myself clearer.
You must remember those early single coners and maybe a capacitive coupled tweeter, many listeners still claim the simple 6dB system sounded better than the high order types ?
So many people don't realise what's going on at the speaker end and the interaction with the global feedback network, esp when there's nasty phase changes, poor circuit damping; and concentrate too much on a perfect amp design without seeing the rest.. I've been through 50 yrs of it, still baffled to how some people claim excellent sound with some of the most appalling waveforms. Oh well.
richy
Attachments
Sorry, Palustris - My bad 😱
I just briefly read Williamson - and thought Mullard......... sigh!
However - this thread is somewhat interesting. I've built quite a lot of different stuff over the years, both hprofessinally and hobby wise , - some HiFi stuff too. I've been pondering on building a tube amp for fun, but I don't want to go the KT88 / 6550 route, due to the inflated prices. I'd also like to set up a set of high sensitivity loudspeakers, so that I don't need 50-60W ++. That's why I've been looking at the Byriht amp for quite some time, but real estate and low market turn around in my area have sort of prevented be from doing something for quite some years...... 2014 wil hopefuly solve that.
Here's the link for the Byrith amp...... 30W Push Pull amplifier designed by Claus Byrith, Royal Academy of Music in Aarhus, Denmark. | Lundahl Transformers
Comments anyone???
Last edited:
Seems the exact same amplifier published by elektor magazine some time in the 90's
An externally hosted image should be here but it was not working when we last tested it.
In a greek electronics forum everybody and their aunt have built one, so I assume something must be goood about it. I remember the idea was not to use the elektor boards because modern capacitors do not fit well.
????????? ???????? el34 pp (elektor)
Hi AuroraB !
On Mr. Claus Birth schematic I don`t understand real function of 200V/1W zener diode , except that is used there to limit EF86 anode voltage around 200VDC when that tube run cold for several seconds at Amp start .
I suggest you to use there 160V/5W( 1N5384B) zener diode altogether with 22K/5W -W.W.- res. instead of 100K series res. to get there relative stable & hum free 160VDC B+ source for supplying that sensitive input gain stage .
Best Regards !
On Mr. Claus Birth schematic I don`t understand real function of 200V/1W zener diode , except that is used there to limit EF86 anode voltage around 200VDC when that tube run cold for several seconds at Amp start .
I suggest you to use there 160V/5W( 1N5384B) zener diode altogether with 22K/5W -W.W.- res. instead of 100K series res. to get there relative stable & hum free 160VDC B+ source for supplying that sensitive input gain stage .
Best Regards !
Last edited:
- Status
- Not open for further replies.