|
Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets. |
|
Thread Tools |
3rd May 2021, 7:25 pm | #1 |
Dekatron
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 13,953
|
Improving the 'Ultra-Linear' push-pull amplifier.
I've never really been a fan of Ultra-Linear push-pull; with many valves it imposes an unnaturally-low and indecent constraint on the HT-supply in order not to exceed the static maximum G2-voltage.
I instinctively think that for best linearity/power-output you need a nice high anode-voltage! It occurred to me that by including a suitably-rated Zener-diode, or a classic "purple glow" 0D3 regulator-valve in series with each screen-grid-to-transformer-tap you could then beneficially wind the overall HT up by at least 150V. OK, I know Zeners and glow-discharge tubes can be noise-generators - but a bit of parallel-capacitance should sort that. [Don't put too much parallel-capacitance across a glow-tube or it can become a relaxation oscillator]. What say you, good people? I like my 807s to have a least 600V on their anodes! |
3rd May 2021, 9:35 pm | #2 |
Guest
Posts: n/a
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
How about capacitively coupling the ultra linear tap to a resistor fed screen grid, DC conditions ensured, AC as per ultra tap?
edit: The resistor feed could be the anode. Last edited by Guest; 3rd May 2021 at 9:38 pm. Reason: had a thought |
3rd May 2021, 9:57 pm | #3 |
Pentode
Join Date: Dec 2009
Location: Chatteris, Cambridgeshire, UK.
Posts: 176
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Certainly worthy of experimentation, it would be interesting to see if the screen DC voltage can successfully be made independant of the AC signal with stable results. I have a feeling there could be trouble with unwanted phase shifts introduced with capacitive coupling.
As you probably well know, Peter Walker had an alternative solution back in 1952 with the Quad II output stage.
__________________
Andrew Last edited by PYE 405; 3rd May 2021 at 10:04 pm. |
3rd May 2021, 10:16 pm | #4 |
Pentode
Join Date: Dec 2009
Location: Chatteris, Cambridgeshire, UK.
Posts: 176
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Another option is to include a resistor in the screen feed, this has certainly been done before to reduce screen voltage/current.
A zener or voltage stabilizer would play havoc with the AC signal.
__________________
Andrew |
3rd May 2021, 10:43 pm | #5 |
Octode
Join Date: Feb 2009
Location: Wallasey, UK.
Posts: 1,308
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
As in the Mullard 5-10 and 5-20 UL circuits?
|
3rd May 2021, 10:48 pm | #6 |
Pentode
Join Date: Dec 2009
Location: Chatteris, Cambridgeshire, UK.
Posts: 176
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Indeed so... the 5-20 had, I believe, a value of 1K connected to the screen's.
__________________
Andrew |
3rd May 2021, 11:36 pm | #7 | |
Dekatron
Join Date: Sep 2007
Location: Oxfordshire, UK.
Posts: 4,310
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Quote:
Cheers, GJ
__________________
http://www.ampregen.com |
|
3rd May 2021, 11:50 pm | #8 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,943
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
The possibility of having the anode and screen at different voltages was covered in the original Hafler and Keroes ultralinear article, and in the associate patent. It was done by having a separate transformer winding for the screen grid.
In fact, it looks as if this were the primary approach to ultralinear, with the use of a transformer tap for the screen grid viewed as being a simplified approach that incurred the need for a common HT voltage. The original Audio Engineering article and the patent are attached. Cheers, |
4th May 2021, 12:03 am | #9 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,943
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Blumlein also covered the possibility of using different anode and screen voltages in his original distributed loading patent, albeit by different means.
Blumlein’s work precede that of Hafler and Keroes, although the latter appear not to have acknowledged it. Cheers, |
4th May 2021, 12:42 am | #10 | |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,943
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Quote:
Although it was noted that a patent had been applied for, so far, I have been unable to establish that a patent was actually issued. The first “hi-fi” use was in the Acoustical QA12 amplifier of 1947: Quad claimed that its circuit allowed the use of different anode and screen voltages without the complication of separate transformer windings. That claim was made in the third page of the attached article. Cheers, |
|
4th May 2021, 3:29 am | #11 |
Banned
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
I agree with the idea of higher plate voltages making a "better" amplifier. In a recent discussion on 807's (another thread) we "all" agreed that the voltage rating of the 807 was officially upgraded to include the origional williamson design, where it was used as a triode strapped tetrode. I have been using 807's and now some of the variants ( CV428/5B/254 ) for many years and have found them to be like the origional 807's. Essentially indestructable.
( Read between the lines there, and do do the math) My own attempt at plagiarizing Joseph Marshalls Goldenears amplifier will be my test!. The amp is relatively complex and expensive enough without throwing extra power supplies or screen windings into the argument. Larger plate voltages also means more expensive filtering components, regardless of what type of filter you require or design. Running 600 volts DC onto the plates would need at least 750 volt rated capacitors, be they electrolytics or PIO's. Better insulation at least for ground shorts means wider thicker insulation for chokes, and for that matter even the HT windings on the power transformer, especially if you use vacuum rectification. At the end of the day it is like everything else ever designed! 1. Today, as we know the number one factor is profit. Expensive doesnt sell well ( mostly) so profit is down. 2> Longevity, or how much warranty do I provide, especially in todays world of "plug'n'play". There are a few other transformer winders on the forum that will attest that making very nice output transformers wont win the lottery, OR let you sleep well!. A really well designed transformer for a specific, exact job can take quite a few hours and the "software" thats blowing around at present is not really best design, more like best practice, taking into account point one. Finding components like onion paper is IDEAL thicknesses is getting very difficult if not impossible. Whatever happened to 7" thou laminations in SiGO steel stampings? 14" thou is available everywhere for cheap, but doesnt make "very good" output transformers. About the best they can achieve is 100KhZ WITH feedback. I am going to run eight CV428/5B/254 in my amplifier and I will post my build and the end result.. I do have a quick question though G6, WHY do you need at least 75 watts to begin with? A single pair of 807's can do that easily. Forget the slight red plating, thats perfectly normal for 807's. Just my take. Regards Joe |
4th May 2021, 8:26 am | #12 | |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Quote:
David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
|
4th May 2021, 8:57 am | #13 | |
Dekatron
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 13,953
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Quote:
I'm weighing-up the merits and demerits of different modulator designs; a triode-connected [grids strapped together] pair of 807s in zero-bias Class-B as recommended by RCA is the other alternative under consideration. I have a nice big Woden multi-ratio transformer which would allow 'ultra-linear' wiring, and that led to my strange idea. Ultra-linear operation would have the advantage of giving greater tolerance-of-error with the impedance-ratios available. |
|
4th May 2021, 9:11 am | #14 |
Heptode
Join Date: Jun 2017
Location: Southport, Merseyside, UK.
Posts: 646
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Bailey and Radford made some interesting improvements at the phase splitter stage. Hi Fi News ca 1961 had details I think. See also http://www.audiomisc.co.uk/HFN/Radfo.../circuits.html
__________________
Time flies like an arrow, fruit flies like a banana |
4th May 2021, 10:04 am | #15 |
Pentode
Join Date: Dec 2009
Location: Chatteris, Cambridgeshire, UK.
Posts: 176
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
I thought the thread starter would encourage interesting input and it certainly has !
I will continue to read with keen interest and just to add more to the mix, how about the Circlotron amplifier? https://www.angelfire.com/electronic...irclotron.html
__________________
Andrew |
4th May 2021, 10:32 am | #16 |
Heptode
Join Date: May 2017
Location: Konongo, Ghana
Posts: 510
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
There are ofcourse output transformers with dedicated seperate windings for ultra-linear operation, like the ones on this page having "SSCR" in their designation: https://www.mennovanderveen.nl/cms/i...ten/specialist
__________________
Robert |
4th May 2021, 10:47 am | #17 |
Banned
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Oh dear Andrew !!! The circlotron is fraught with problems!!. Wind a power transformer for a start!!.
I have made one only. I had to wind it on gate laminations, which have twice the length of normal laminations for the centre and outside legs. What that acheives is twice the winding length for the bobbin. Regulation is of SUPER importance, because if one supply sags by even a small amount the output bottles will melt almost instantly. My attempt was just a puny pair of ( wonderful ) 6V6's. Its a pity that TRUE 8 ohms speakers are not and never will be available. To G6, well, as modulators it might make sense. The modulator generally runs 100% modulation. so distortion with your real class B doesnt matter. I mostly do audio Hi-Fi, especially in my dotage, so class B is out. AB1 is nice because output power versus distortion and efficiency is quite good. NICE ultra linear transformers work well and can achieve good efficiency. As I am not seeking to modulate the other side of the world, I stay with UL designs. Joe |
4th May 2021, 11:05 am | #18 |
Banned
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Be a little careful of Menno's designs. I bought his book when it was released and "played" with a few ideas he presented. Toroidal output transformers are ten times harder to design than EI types. Yes, you get better coupling, but worse interwinding capacitance. Fix the capacitance and the coupling goes out the door. Toroids are very nice things, originally a spin off of double C core transformers. Double C core still require bobbins that you can section the windings fairly easily. Almost impossible with a toroid as one winding is basically stacked on top of the last, ( or next ). I tried winding a few simple ones, but ran into big problems. Yes I am not Menno Van Der Veen, but my EI designs beat the frequency response and efficiency of a toroid on any given day. Mind you they are more than twice the size. I run VERY conservative flux densities with my EI designs which is almost not a problem with toroids, nor is coupling that big a factor with toroids. BUT!!! interwinding capacity is a major drawback because vertical sectioning on a toroid reduces coupling by a huge percentage.
If you can't follow my discussion, please read up on your transformer theory. With respect Joe |
4th May 2021, 12:25 pm | #19 |
Pentode
Join Date: Dec 2009
Location: Chatteris, Cambridgeshire, UK.
Posts: 176
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
Hi Joe, yes indeed the circlotron is a scary design and best avoided. I threw it in just for interest, but UL is the proven option adopted by most HiFi designers of yore.
__________________
Andrew |
4th May 2021, 12:44 pm | #20 | ||
Dekatron
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,061
|
Re: Improving the 'Ultra-Linear' push-pull amplifier.
It depends why you want UL, after all.
Resistive/capacitive coupling to UL taps starts to negate one of the benefits of UL, that is, that the screen current contributes to power output in addition to anode current. I do like the separate windings for the screens - with care, and provided the voltage separation is not too great, it should be possible to wind bifilar with the appropriate portion of anode winding. Quote:
|
||