Dear all,
I know the basics: a SE OPT amplifier needs a gap to avoid core saturation, the PP avoids the (big) gap as DC currents cancel out. Since I joined Diyaudio discussions regularly pop up about PP transformers being used with only one side driven, the other side being loaded by a CCS (a mosfet or a tube), this allows the SE sonic signature using PP OPTs (smaller, maybe available).
I can’t remember having seen an amplifier that allows (gradual) transition between PP and SE, preferably keeping the gain constant. I saw/have some ideas, like feeding a balanced PP amplifier with an input transformer loaded by a potentiometer, with the wiper connected to ground. In the middle position both legs get the same signal, the amp is PP. Turning the pot, each leg becomes a different signal, something between PP and SE. With the wiper at the extreme of the pot, the amp operates in a sort of SE (but take care that this one leg is not clipping because it now gets twice the signal as in PP operation). But say the amp uses triode output tubes, in the SE case one may not be amplifying, it is however still loading the transformer with its low plate impedance. Good or bad?
Please share existing projects, your ideas or schematics. Given the circuit is doable I am willing to build a prototype, I have input transformers (LL1540), OPTs, lots of different driver and output tubes, not afraid of using mosfets (enhancement and depletion), have multiple adjustable PS that can all be used for testing.
The purpose of this exercise is to have an amplifier that can play multiple distortion signatures by turning pots or maybe some switches, so the influence of variables as PS, employed output tubes, OPTs, being warmed up or not are kept to a minimum. Constant gain would be nice as I plan to use this for the midrange in a bi-amplified system, and because differences in gain can also be perceived as differences in SQ.
Many thanks for your ideas!
Erik
I know the basics: a SE OPT amplifier needs a gap to avoid core saturation, the PP avoids the (big) gap as DC currents cancel out. Since I joined Diyaudio discussions regularly pop up about PP transformers being used with only one side driven, the other side being loaded by a CCS (a mosfet or a tube), this allows the SE sonic signature using PP OPTs (smaller, maybe available).
I can’t remember having seen an amplifier that allows (gradual) transition between PP and SE, preferably keeping the gain constant. I saw/have some ideas, like feeding a balanced PP amplifier with an input transformer loaded by a potentiometer, with the wiper connected to ground. In the middle position both legs get the same signal, the amp is PP. Turning the pot, each leg becomes a different signal, something between PP and SE. With the wiper at the extreme of the pot, the amp operates in a sort of SE (but take care that this one leg is not clipping because it now gets twice the signal as in PP operation). But say the amp uses triode output tubes, in the SE case one may not be amplifying, it is however still loading the transformer with its low plate impedance. Good or bad?
Please share existing projects, your ideas or schematics. Given the circuit is doable I am willing to build a prototype, I have input transformers (LL1540), OPTs, lots of different driver and output tubes, not afraid of using mosfets (enhancement and depletion), have multiple adjustable PS that can all be used for testing.
The purpose of this exercise is to have an amplifier that can play multiple distortion signatures by turning pots or maybe some switches, so the influence of variables as PS, employed output tubes, OPTs, being warmed up or not are kept to a minimum. Constant gain would be nice as I plan to use this for the midrange in a bi-amplified system, and because differences in gain can also be perceived as differences in SQ.
Many thanks for your ideas!
Erik
People have build guitar amps that do this. Basically you just build a conventional PP output stage, and add a 'volume' pot to the grid of one of the output tubes, so you can adjust how much drive it receives.
For your application you'll want to run the output in strict class A, otherwise you'll get DC offset (transformer saturation) effects when only one tube is driven.
For your application you'll want to run the output in strict class A, otherwise you'll get DC offset (transformer saturation) effects when only one tube is driven.
the other option is to drive one tube grid with a signal varying between 180 degrees and 0 phase shift. in the zero phase case the 2 tubes operate in paralell. so both tubes always do a job
Thanks MerlinB,People have build guitar amps that do this. Basically you just build a conventional PP output stage, and add a 'volume' pot to the grid of one of the output tubes, so you can adjust how much drive it receives.
For your application you'll want to run the output in strict class A, otherwise you'll get DC offset (transformer saturation) effects when only one tube is driven.
for the ones interested in visualizing this, on the audiodesign link there is one example of the above (the 5th picture).
To basreflex,
You suggest a PP amplifier, where one can control the phase of the signal presented to one tube? If the phase is 180º off, it will be push pull. But if it is 0º off (that is, the same signal is applied to both output tubes in the PP amp), the amp will do nothing (no swing developed across the primary)?
You suggest a PP amplifier, where one can control the phase of the signal presented to one tube? If the phase is 180º off, it will be push pull. But if it is 0º off (that is, the same signal is applied to both output tubes in the PP amp), the amp will do nothing (no swing developed across the primary)?
This may work with a single sine wave but shifting the phase and recombining them in the OPT with a complex music signal consisting of multiple frequencies at multiple constantly varying amplitudes will just create a mess.To basreflex,
You suggest a PP amplifier, where one can control the phase of the signal presented to one tube? If the phase is 180º off, it will be push pull. But if it is 0º off (that is, the same signal is applied to both output tubes in the PP amp), the amp will do nothing (no swing developed across the primary)?
Shifting the phase is indeed a step too complex, I think. Thinking about it, to get good damping factor, in PP I want each leg to have low impedance, in SE the leg with music should have low impedance, the other should have high impedance. So some form of NFB must be applied.
A first idea, partially based on the Softone Model 7 that I was studying yesterday http://softone.a.la9.jp/english/Model7/M7-eng-3.htm
Each CCS should be decoupled with some big cap!
It uses an input transformer, only the secondary is shown. Load it with some 100k (20k drawn), because the 10k resistors will also load it. Center tap to ground. At the secondary I add a 100R pot, also center tap to ground. When both pots are in the middle position, one has PP. Putting both into either extreme (shown right), the grid of the upper tube is grounded (the CCS), while the lower now gets 2x the signal and 2x the feedback voltage, so overall gain will stay equal. Couple the pots?
what do you think?
Each CCS should be decoupled with some big cap!
It uses an input transformer, only the secondary is shown. Load it with some 100k (20k drawn), because the 10k resistors will also load it. Center tap to ground. At the secondary I add a 100R pot, also center tap to ground. When both pots are in the middle position, one has PP. Putting both into either extreme (shown right), the grid of the upper tube is grounded (the CCS), while the lower now gets 2x the signal and 2x the feedback voltage, so overall gain will stay equal. Couple the pots?
what do you think?
Attachments
But one can't switch between them on the fly, can you? At least I found clearly both SE and PP schematics for this amp, also the OPT needs changes to the taps.There is the Edison 60 that uses a gapped output transformer and can be switched between Parallel SE or PP.
This amplifier works in both SEP and PP, working modes configured automatically using some relays according to the listener's wishes.
To make myself clear, I am looking for options around a PP opt with no gap, so in the SE mode it must use the second active device (probably pentode) to pull DC current to avoid saturation of the core, just as shown in the link from Ciufolli.
Thanks for the link sesebe. Do you think the amp makes sense? The PP output stage looks ok, but in SE the tubes are paralleled, and the now double current goes through one side of the OPT only. I can't use it, as my OPTs are PP and I will not subject them to such large standing DC current.
Thanks for the link sesebe. Do you think the amp makes sense? The PP output stage looks ok, but in SE the tubes are paralleled, and the now double current goes through one side of the OPT only. I can't use it, as my OPTs are PP and I will not subject them to such large standing DC current.
I'm not a fan of the sound on tubes, but I know the author and he is very pleased with the way it sounds.
Hello,
The French made a single ended solid state power amp in the eighties where they used a toriodal power transformer as output transformer. One secondary winding had the signal and the other secondary winding had a similar current running so they did not need an airgap.
I have info at home but can be found online too.
Greetings Eduard
The French made a single ended solid state power amp in the eighties where they used a toriodal power transformer as output transformer. One secondary winding had the signal and the other secondary winding had a similar current running so they did not need an airgap.
I have info at home but can be found online too.
Greetings Eduard
indeed my idea does not work, should have thought about it more thoroughly.
Years ago Bottlehead published an article in their magazine about an amp that would run in either parafeed SE or parafeed PP according to whether the CT of the PP output transformer was connected I think to ground. Can't find it right now but will try to find it.
John
John
@ErikdeBest : have a look to Mesa Boogie schematics : Express, Lonestar Special models notably, which offer switchable SE and PP operation, cathode and fixed bias, diodes or rectifier valves options with EL84 tubes. That said, it's not continuous variation, but instant switchable options.
If it was me, I would simply try to add a A100K to A250K pot with suitable coupling caps in the grid of one of the PP power tubes. This would make some kind of continuous transition from PP to SE operation, without imbalancing too much the DC load on the supply and the power transformer. In a class A operation with common cathode bias, it would work quite fine - at least on the paper !
That's what suggested already @Merlinb in his post above.
I must confess that I never tested that "continuous" solution... My Serial Tone Killer (DIY) amp uses 5 / 15 / 30 / 90W operation by varying the number of power tubes (2 / 4) and the plate voltage, bias and class of operation (250V cl.A / 450V cl. AB), but always in Push-Pull configuration, no SE...
T
If it was me, I would simply try to add a A100K to A250K pot with suitable coupling caps in the grid of one of the PP power tubes. This would make some kind of continuous transition from PP to SE operation, without imbalancing too much the DC load on the supply and the power transformer. In a class A operation with common cathode bias, it would work quite fine - at least on the paper !
That's what suggested already @Merlinb in his post above.
I must confess that I never tested that "continuous" solution... My Serial Tone Killer (DIY) amp uses 5 / 15 / 30 / 90W operation by varying the number of power tubes (2 / 4) and the plate voltage, bias and class of operation (250V cl.A / 450V cl. AB), but always in Push-Pull configuration, no SE...
T
Last edited:
ErikdeBest,
You mentioned an amplifier that can do both single ended, and push pull. Lots of examples for that.
But you said: "this allows the SE sonic signature using PP OPTs"
Other than preventing saturation in a single ended amplifier, is there Any other characteristic that an air gap gives to the 'sound' of single ended?
Some push pull amplifiers use an Air Gapped push pull output transformer.
Why would they do that?
For a different sound?
The French built 2 single ended amplifiers with air gaped output transformers, and then drove the inputs with opposite phase, and then switched the output transformer secondaries in opposite connection (in parallel). That properly combines the two phases at the outputs.
I tried that once myself.
I think we are not finished with the science, and we are not finished with the perceived sound of the many different topologies of vacuum tube amplifiers.
That is why it is so important for you and others to keep developing these ideas.
Happy designing, building, and listening!
It is a Mean thing to say that Karl Friedrich Gauss was an Average man, who was Centered on his Gaussian Curve.
You mentioned an amplifier that can do both single ended, and push pull. Lots of examples for that.
But you said: "this allows the SE sonic signature using PP OPTs"
Other than preventing saturation in a single ended amplifier, is there Any other characteristic that an air gap gives to the 'sound' of single ended?
Some push pull amplifiers use an Air Gapped push pull output transformer.
Why would they do that?
For a different sound?
The French built 2 single ended amplifiers with air gaped output transformers, and then drove the inputs with opposite phase, and then switched the output transformer secondaries in opposite connection (in parallel). That properly combines the two phases at the outputs.
I tried that once myself.
I think we are not finished with the science, and we are not finished with the perceived sound of the many different topologies of vacuum tube amplifiers.
That is why it is so important for you and others to keep developing these ideas.
Happy designing, building, and listening!
It is a Mean thing to say that Karl Friedrich Gauss was an Average man, who was Centered on his Gaussian Curve.
- Home
- Amplifiers
- Tubes / Valves
- Amplifier with continuous change between SE and PP operation