Single-Ended vs. Push-Pull
Martin Keene (LIS)
What are the major differences between a single-ended
design, and a push-pull design? What are the pros/cons of
each? I see these terms used alot, but never explained.
Any info greatly appreciated.
BTW, what are the advantages of an "output transformerless"
design? Another mystery to me. Thanx in advance.
Andy Keene
mke...@luna.cas.usf.edu
Joe Lowe
>
>
> What are the major differences between a single-ended
> design, and a push-pull design?
Single Ended amps. have only one tube in the output. Push-Pull amps.
have two (or an even number). S.E. amps also generally use low Mu
Triodes, with very little or no negative feedback. Push pull amps
generally use pentodes and a good dose of feedback to help get rid
of all the crud that pentodes in push-pull generate. Transistor amps
use gobs of feedback to even get them to work at all. Can you say
square waves? I thought you could.
> What are the pros/cons of
> each? I see these terms used alot, but never explained.
> Any info greatly appreciated.
Well, IMHO a good single ended amp sounds very musical, good depth,
very balanced sound. You can sit and listen to it for hours without
fatigue. As I write this I am listening to my homebrew S.E. amp that
uses type 50's from the early 1930's. Once you hear an amp with any
of the classic triodes (300B's, 50's, ect) you will be spoiled
forever. I thought S.E. was just some crazy fad until I heard for
myself. Now I am a true believer.
Now for the downside. If you are a Megadeath fan and have a set of
Radio Shack speakers, forget S.E. My amp puts out about 3 Watts, and
most do less than 10. You need good speakers for any tube amp,
but they are a must with S.E.
I know I will make some people mad with this but, push-pull pentode
amps are boat anchors. Some of them are very pretty, but still boat
anchors. I have owned 'em all, Mac, Scott, Fisher. 20 years ago they
sounded 10 times better than the transistor junk that was on the
market. That was then and this is now.
> BTW, what are the advantages of an "output transformerless"
> design?
NONE. They are an engineering curiosity. If there were such a thing
as a P-N-P tube, they would be the best thing since fresh air. Until
someone figures out how to get electrons to run backwards, just forget
about OTL.
Well, that should be about enough flamebait for one day! :)
Joe
Joe Mizera
>"Martin Keene (LIS)" <mke...@luna.cas.usf.edu> wrote:
>>
>>
>> What are the major differences between a single-ended
>> design, and a push-pull design?
>
>Single Ended amps. have only one tube in the output. Push-Pull amps.
>have two (or an even number). S.E. amps also generally use low Mu
>Triodes, with very little or no negative feedback. Push pull amps
>generally use pentodes and a good dose of feedback to help get rid
>of all the crud that pentodes in push-pull generate. Transistor amps
>use gobs of feedback to even get them to work at all. Can you say
>square waves? I thought you could.
This is true of course. More importantly imho, the entire waveform is amplified
by one device. In push pull, the two (or multiple) tubes each amplify
approximately one half of the waveform. In class A push pull, both tubes
amplify the full waveform, but since the tubes are out of phase in series
(basically) they still share the load. In single ended, the entire waveform is
amplified by one (or more, in parallel) device. Less power is available is
these designs, because one tube can only dissipate so much power. Also, single
ended triode amps are by nature class A. In being so, the output tube is biased
"on" for the entire waveform. This also limits output power.
>> What are the pros/cons of
>> each? I see these terms used alot, but never explained.
>> Any info greatly appreciated.
>
>Well, IMHO a good single ended amp sounds very musical, good depth,
>very balanced sound. You can sit and listen to it for hours without
>fatigue. As I write this I am listening to my homebrew S.E. amp that
>uses type 50's from the early 1930's. Once you hear an amp with any
>of the classic triodes (300B's, 50's, ect) you will be spoiled
>forever. I thought S.E. was just some crazy fad until I heard for
>myself. Now I am a true believer.
>
Now for the downside. If you are a Megadeath fan and have a set of
>Radio Shack speakers, forget S.E. My amp puts out about 3 Watts, and
>most do less than 10. You need good speakers for any tube amp,
>but they are a must with S.E.
>
>I know I will make some people mad with this but, push-pull pentode
>amps are boat anchors. Some of them are very pretty, but still boat
>anchors. I have owned 'em all, Mac, Scott, Fisher. 20 years ago they
>sounded 10 times better than the transistor junk that was on the
>market. That was then and this is now.
>
No, I'm not mad, this is (imho) just not correct. Scott and Fisher might be
lower on the food chain, but don't write off pentodes without hearing Audio
Research!
>> BTW, what are the advantages of an "output transformerless"
>> design?
>
>NONE. They are an engineering curiosity. If there were such a thing
>as a P-N-P tube, they would be the best thing since fresh air. Until
>someone figures out how to get electrons to run backwards, just forget
>about OTL.
>
I can't really argue with this. The transformer not only blocks the DC voltage
on the plates, but it matches the high impedance output of the tubes, to the low
impedance of the speaker. OTL's are VERY sensitive to speaker load. Never
really popular is the US, there are several made in the orient.
>Well, that should be about enough flamebait for one day! :)
>
>Joe
>
>
Joe Mizera
Austin Stereo Service
5448 Burnet Rd. Ste 5
512-451-0408
http://www.realtime.net/~joe/index.html (The tube audio page)
http://www.jump.net/~austinstereo/ (Austin Stereo Service)
Les Bernstein
>"Martin Keene (LIS)" <mke...@luna.cas.usf.edu> wrote:
>>
>>
>> What are the major differences between a single-ended
>> design, and a push-pull design?
>
>Single Ended amps. have only one tube in the output. Push-Pull amps.
>have two (or an even number). S.E. amps also generally use low Mu
>Triodes, with very little or no negative feedback. Push pull amps
>generally use pentodes and a good dose of feedback to help get rid
>of all the crud that pentodes in push-pull generate. Transistor amps
>use gobs of feedback to even get them to work at all. Can you say
>square waves? I thought you could.
>
>
>> each? I see these terms used alot, but never explained.
>> Any info greatly appreciated.
>
>Well, IMHO a good single ended amp sounds very musical, good depth,
>very balanced sound. You can sit and listen to it for hours without
>fatigue. As I write this I am listening to my homebrew S.E. amp that
>uses type 50's from the early 1930's. Once you hear an amp with any
>of the classic triodes (300B's, 50's, ect) you will be spoiled
>forever. I thought S.E. was just some crazy fad until I heard for
>myself. Now I am a true believer.
>
>Now for the downside. If you are a Megadeath fan and have a set of
>Radio Shack speakers, forget S.E. My amp puts out about 3 Watts, and
>most do less than 10. You need good speakers for any tube amp,
>but they are a must with S.E.
>
>I know I will make some people mad with this but, push-pull pentode
>amps are boat anchors. Some of them are very pretty, but still boat
>anchors. I have owned 'em all, Mac, Scott, Fisher. 20 years ago they
>sounded 10 times better than the transistor junk that was on the
>market. That was then and this is now.
>
>> design?
>
>NONE. They are an engineering curiosity. If there were such a thing
>as a P-N-P tube, they would be the best thing since fresh air. Until
>someone figures out how to get electrons to run backwards, just forget
>about OTL.
>
>
>Joe
>
>
Here's more flamebait. Even the RCA Radiotron manuals that are 50 years old will show you why
S.E. amps are generally inferior to well-executed push-pull designs. This is a practical
matter and not a theoretical one. Physics has not changed. By the way, there are some wonderful
push-pull designs sans heavy feedback. Nothing here is really new-- it just gets "re-invented"
and sold for thousands of $$$$$
Les
Bob Bruhns
to the field of amplification.
"Single-ended" refers to an amplifier design which uses a single
amplifying device, or multiple devices *in parallel*, to produce signal
amplification. The first amplifiers were of this design. In this
design, an input signal controls the current flowing through the output
device(s); output is derived from the variation in the current that flows
in the output device. This technique is natural for low level
amplification; it is typical in radio frequency power amplifiers, and
many people favor it for high level audio amplification as well, due to
certain sonic qualities of single ended tube hi-fi amplifiers.
"Push-pull" refers to a later technique, in which two amplifying
devices produce signal amplification somewhat like those old two-person
rail cars where two people pump on a handle to drive the car down the
track. One person's handle goes up while the other's goes down. In a
push-pull amplifier, one device is pulling down, while the other is
releasing. Together, two devices in push-pull can produce much more
audio output than the same two devices in single-ended service. This
fact has made the push-pull design popular. In radio frequency
amplifiers, push-pull operation does not produce more power, but it does
generate less harmonic energy, making output filtering less strenuous.
RF amplifiers generate large harmonic distortion because they do not
have to conduct current over the entire sine-wave cycle. Audio
amplifiers do.
In push-pull audio amplifiers, it is possible (but not necessary) for one
output device to conduct only on one half of the signal cycle, and the other
output device to conduct only on the other half of the cycle. Single-ended
designs require the output device to conduct over the entire audio cycle.
This allows push-pull amplifiers to operate more efficiently, thus
getting more power out of a given number of output devices.
But push-pull places many demands on the output devices and the output
transformer if present. Precise timings are critical if the output
devices switch between each other during the audio cycle, and there are
considerations of linearity and dynamic impedance over the audio cycle.
Push-pull can be very tricky. Single-ended is not so tricky, but its
constant DC places severe requirements on the power supply and the output
transformer.
Transformers in general are not the best devices for the transmission of
audio signals, and they are big, expensive and heavy, so they are used as
little as possible. Tube circuits generally need them because of the
difference between tube impedances and speaker impedances. OTL designs
eliminate the output transformer in order to simplify the circuit and/or
improve the sound, etc.
Bob Bruhns, WA3WDR, bbr...@li.net
Joe Lowe
> Here's more flamebait. Even the RCA Radiotron manuals that are 50
> years old will show you why S.E. amps are generally inferior to
> well-executed push-pull designs.
I just happen to have a copy of the "RCA RADIOTRON MANUAL No. R-10"
copyright 1932, and for the life of me can't seem to find anything
therein that would support your position.
> This is a practical
> matter and not a theoretical one. Physics has not changed.
You got that right! Push-Pull generates crossover distortion. NASTY.
Also, all the even order harmonics are cancelled, leaving only odd
order harmonics + crossover distortion. SHEW!
> By the way, there are some wonderful
> push-pull designs sans heavy feedback.
Sure I have built a few that sound pretty good, just not as good as
my S.E.
> Nothing here is really new-- it just gets "re-invented"
> and sold for thousands of $$$$$
Yup.
Les Bernstein
>>
>> Here's more flamebait. Even the RCA Radiotron manuals that are 50
>> years old will show you why S.E. amps are generally inferior to
>> well-executed push-pull designs.
>
>I just happen to have a copy of the "RCA RADIOTRON MANUAL No. R-10"
>copyright 1932, and for the life of me can't seem to find anything
>therein that would support your position.
Look at the later ones. Say in the 40's and 50's
>
>> This is a practical
>> matter and not a theoretical one. Physics has not changed.
Read the above again and again and again.
>
>You got that right! Push-Pull generates crossover distortion. NASTY.
>Also, all the even order harmonics are cancelled, leaving only odd
>order harmonics + crossover distortion. SHEW!
Yes, the primary distortion is cancelled!!!!! Single ended is fine for low low power.
>
>
>> By the way, there are some wonderful
>> push-pull designs sans heavy feedback.
>
>Sure I have built a few that sound pretty good, just not as good as
>my S.E.
Then go ahead and enjoy. My reaction was provoked because of your general slam on a fine design
principle
>
>
>> Nothing here is really new-- it just gets "re-invented"
>> and sold for thousands of $$$$$
>
>Yup.
Ah, we agree.
Happy New Year.
Les