• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Single-ended OR push-pull?

Hi, I always thought that SE transformers were more expensive because they have to be constructed to avoid saturation of the core by the standing current of the output Tube. With P-P the currents in each half cancel each other and hence minimal core saturation.
My personal view is that SE is probably the best way to go with class A but if I was going P-P then it would be strictly using Triodes only.
That's a matter of some debate.

From an objective POV, single-ended amps perform far less accurately than push-pull. They impart distinctive signature to the sound (non-flat frequency response, high distortion). If that's what you want, and LOTS of people want that, go for it. The distortions and frequency response abberrations are of the sort that many find euphonic.

If you're more interested in the amp having a low signature and being faithful to the signal fed it, push-pull is a better choice. But you may not "like" the sound.
Each amplifier sounds different, especially when no NFB is used. I've had good results with both SE and PP. There really seems to be equal pros and cons to each if you break it down.

But yes, SE iron is always bigger and about twice the price as an equivalent PP transformer. That is one of the con's of SE.

I think most people fear that since PP is more electrically correct than SE, that it will sound like your plain vanilla solid-state "blameless" amplifier. I have never found that to be true.


2007-03-02 8:04 pm
I had a chance to hear a truly excellent PP amp using separate hige PSU's (80Lbs each ) and transmiting pentodes in the output stage. It was fantastic and the best I've heard -soft and brutal at the same time.
I think (thought ) that I like SE better but it's all aplication . If you use the same care and component quality in PP amp as in SET amps it should produce equally good results.
Good SET amps aren't euphonic nor warm. I'd say they tend to be fast ,unforgiving sometimes .
I built 45 Set and using Lowther DX4 in front horn crossed from 100 Hz up -104db (guess) . It barely makes it in moderate room with a little more demanding music. It sounds sweet but I think not very true.
In summary good PP amp is not going to be cheap (cheaper than SET)
Regards, L
Originally posted by SY From an objective POV, single-ended amps perform far less accurately than push-pull.

True, perhaps, if you listen to sin waves, but I'm not sure this is quite as much of a fact as stated here. From Jack Elliano:

P-P will cancel even order harmonics but will add or emphasize the odd order. To illustrate this, one experiment that still intrigues me is when we took two audio frequencies and mixed them to one waveform which measured as 80% distortion. It then passed through a very good quality commercial P-P amp where the output waveform measured about 50% hmm! But when we then passed it through a SET amp and its output measured 80% just like the original. What Happened to the other 30% within the P-P amp? It appears that there are occasions that some audio signal complexities may also cancel or change structure from the original as well as the P-P tube transfer characteristics on even order harmonics.


2007-03-02 8:04 pm
they (electraprint) sell push -pull transformers too but SET crowd is willing to pay way more for the Iron (and parts ) than PP people .
I heard that in Japan there is no SET against PP dispute and both topologies are equally well regarded. Just those stupid "competetive " westerners always finding the line of conflict .
kacernator said:

I have decided to build Mikaels Abdellahs SE KT88. But SE OPT are very expensive (I can get just Hammond 1628SEA). PP OPT are half the price of SE OPT.

I am going for the sound, but does SE has so much benefits that are worth the expensive OPTs?

My 2 cents. :D Build a PP amp that uses Mullard topology and a 6GK5 as the voltage gain tube and a 12AT7 as the LTP. The 12AT7/ECC81 is extremely well suited to PP amp service. The non-linearity of the 'T7 works to your advantage in a PP amp, as the NET harmonic distortion spectrum is an ear pleasing "waterfall", of 2nd > 3rd, > than 4th. ;)

The high gm/low RP small signal tubes I suggest give you excellent drive capability and resistance against slew limiting.
Most interesting reading the above replies and one quickly gets the idea that in the end it is down to personal preference. If you like what your ears hear then surely that's what it's all about.
There have been lots of truths and mis-truths about the two schemes. One book I read suggested that good Class A could only be obtained using Pentodes or Beam Tetrodes in SE, the reason why most domestic radios used the design. If you tried to increase power using P-P circuitry then you had to use Triodes to get good Class A. The idea of using ultra linear taps on transformers for Pentodes etc was to fool the circuit into thinking Triodes were installed. Other information one picks up suggests that good class A Triode output stages do not need any NFB but that's another bone of contention. The big problem arises I suppose with a question that opened this thread, i.e. a request for the best scheme to adopt. It is a very difficult question to answer as we all have our own ideas, some backed up by hours of private experimentation. For a newbie, I think it best to go with one of the many designs featured on the Forum that have given success. As for folks like I, it is usually down to finding a nice square piece of wood, mounting the hardware and wiring up with lots of places to hook up test gear. I think for many of us we get more fun that way than actually listening to it :D
John007 said:
Don´t waste time with Push pull, Single-ended sounds much better, if you have good sensitive speakers singer is in the room with you.

With all respect, I doubt you have ever heard what a good PP amp can do. Implementation details matter much, much, more than topology.

Also, amps and speakers must be VERY carefully mated. Certain combinations of otherwise highly satisfactory pieces will sound (sadly) like guano.


2008-01-09 10:38 pm
I have just heard SE and PP guitar amps. So can I expect the same for HIFI:

SE - compressed, detailed, thinner, less bass, looser, faster
PP - tighter, more bass, cleaner, slower

From posts above it seems that SE and PP are actually equal and it is just listeners preference. But is it ok to say that PP has one benefit and that is the damping factor? There is no way SE KT88 in triode will have the same bass response with same speaker as with PP KT88 UL?
My tech, who still has my Sansui 1000A, is the former senior technician at Cary Audio. He tells me that there is definately a difference, and SE is better.

Now Cary Audio produces both SET and P-P, but they are definately SET oriented. I will eventually find out which I like better. But I have to say that my Bogen RPF 60 is pretty hard to beat, sound wise. ;)

It still has the original Bogen output 7555 tubes, made by GE for them. Still pumping strong.


2007-03-02 8:04 pm
SET will be easier to build and cheaper too, than really good PP. Especially if you have 102 dB +speakers or bass performance and scale is not that important .Vintage PP amps or anything commercial PP less than say $15K -$20K (20,000 USD) will not give you a true picture on what PP is capable of.
From your post I understand that you're curious about SET so the best way to find out the sonics is to listen to one in sympathetic setup or build one, paying premium for OPT. I've been trough a few of cheap OPts (e-bay)and it's a waste of time and money. Either buy an OPT from reputable winder or look for a deal on a used one.
Back when flea-power SE began to get favour, the tube amps it got compared to were AB pentodes, with lots of loop/global NFB. The SE were hand built, with better power supplies and good parts.

SE sounds different for several reasons. One being its output iron is gapped. That changes the magnetic performance in some important ways. I suspect the lowered remnance is mostly to blame. I apply a gap to PP signal iron when ever I can.

The output stage when run open loop/no NFB it has a variable output impedance. Swing current up, plate Z goes down and output Z goes down. Swing towards cut off and outut Z goes up.

And then there is the shape of the distortion harmonics...there is lots more even stuff with SE( the asymetric distortion products) compared to PP.

I like the PP, Class A sonics. In order to get power, I need plate dissipation. transmitting pentodes, run Ultra-Linear with balanced plate-to-grid FB to a balanced driver stage have worked quite well. High voltage and high current takes high mass, so amp and PS get broken up...:)