I think so as well.
The truth is that is much easier to get a poor quality SE transformer than a poor quality PP transformer. Quality PP transformers are simply better in any way.
In addition to the other advantages you mentioned a PP OT that has to manage a certain power is quite smaller and requires a simpler geometry for a target FR. The PP configuration redistribuites stray capacitances and these are basically 1/2 of SE for the same geometry and size. The smaller size also means lower stray capacitances.
Also a PP transformer can be made with a small gap to deliver the required power without saturation in presence of serious DC imbalance while still getting high inductance without using too large cores (because it doesn't have to withstand the full DC imbalance of a SE). Usually the small natural gap most PP OT's have is enough to take care of typical small imbalance of amplifiers done right.
Last edited:
Yes, I already listed the size, weight, cost, and 'performance', etc. advantages of PP OT's (and I will admit the ease of design to manufacture them, etc.).
So, What is "the small natural gap of most PP OT's"?
I am interested because I have not knowingly seen it, except for those manufacturers who put an air gap there on purpose.
I have stacked my own PP Ot's.
I have built SE and PP, and double SE OT's 'push pull' amps.
Yes, I too know of some specific companies that advertise real air gaps on PP OT's.
Some of them even list the maximum recommended imbalance current.
Do you have a list of those who do and those who don't (air gap and max recommended mA)?
How many of you have ever heard a "Push Pull" amplifier that is configured to use 2 SE OT's per channel, instead of a PP OT's?
Why was the Heathkit W-5M Graph showing 20, 50, and 100 Hz performance versus XmA differential current so easily discounted as unimportant by some writing on this thread?
(especially since it was the residual distortion WITH negative feedback).
Yes, I know that low frequency distortion tends not to be as audible or disturbing as mid frequency distortion.
How many of you measured the differential current on your PP amp (today, with last years tubes, at today's line voltage, and todays line distortion (3rd order makes the same RMS Voltage Power Line have a lesser peak voltage, which affects unregulated B+).
Probably all these factors have already been taken care of.
So, What is "the small natural gap of most PP OT's"?
I am interested because I have not knowingly seen it, except for those manufacturers who put an air gap there on purpose.
I have stacked my own PP Ot's.
I have built SE and PP, and double SE OT's 'push pull' amps.
Yes, I too know of some specific companies that advertise real air gaps on PP OT's.
Some of them even list the maximum recommended imbalance current.
Do you have a list of those who do and those who don't (air gap and max recommended mA)?
How many of you have ever heard a "Push Pull" amplifier that is configured to use 2 SE OT's per channel, instead of a PP OT's?
Why was the Heathkit W-5M Graph showing 20, 50, and 100 Hz performance versus XmA differential current so easily discounted as unimportant by some writing on this thread?
(especially since it was the residual distortion WITH negative feedback).
Yes, I know that low frequency distortion tends not to be as audible or disturbing as mid frequency distortion.
How many of you measured the differential current on your PP amp (today, with last years tubes, at today's line voltage, and todays line distortion (3rd order makes the same RMS Voltage Power Line have a lesser peak voltage, which affects unregulated B+).
Probably all these factors have already been taken care of.
Just to clear up a couple of things before I end my discussion on this thread.
What I am talking about might be called pseudo push pull.
Since it uses the complete circuitry of a push pull amp, except that the PP OT is replaced by two SE transformers that are driven out-of-phase with each other, and re-combined in-phase at the 2 secondaries, it is NOT parallel single ended.
(in any Parallel Single Ended amp I know of, Both output tubes are in phase, and both of them drive one OT in-phase). "Of course all generalizations have exceptions."
The reason I originally brought up the W-5M was to raise the importance of using either individual fixed bias adjustments, or to use individual self bias circuits.
I got off track when I tried to use other examples, such as the pseudo push pull (again not parallel SE).
What I am talking about might be called pseudo push pull.
Since it uses the complete circuitry of a push pull amp, except that the PP OT is replaced by two SE transformers that are driven out-of-phase with each other, and re-combined in-phase at the 2 secondaries, it is NOT parallel single ended.
(in any Parallel Single Ended amp I know of, Both output tubes are in phase, and both of them drive one OT in-phase). "Of course all generalizations have exceptions."
The reason I originally brought up the W-5M was to raise the importance of using either individual fixed bias adjustments, or to use individual self bias circuits.
I got off track when I tried to use other examples, such as the pseudo push pull (again not parallel SE).
I measured the change in a 'vintage hi-fi for Williamson' 10k PP inductance OT with varying net primary DC and observed a reasonable drop off in inductance with practical levels of unbalance.
http://dalmura.com.au/projects/Choke%20measurement.pdf
http://dalmura.com.au/projects/Choke%20measurement.pdf
The small natural gap exists because -except uncut toroidal and R-cores (not all) - there are tolerances and the 2 o more parts that form the core are not truly a single core with no gap.
Some manufactures specify a recommended DC bias that doesn't cause premature saturation (i.e. the transformer can handle the specified power just losing few Hz's at rated power) and doesn't destroy inductance.
See Hashimoto Sansui or Tango, for example.
Here you see the Tango FX-40-5 that has a specified max recommended imbalance of 4.5 mA:
http://www.tubebooks.org/file_downloads/tango_fx-40-5.pdf
Here you can see one of the Hashimoto's (in English!😀) with the additional info about inductance in presence of such imbalance:
http://www.tube-amps.net/images/Hashimoto_Specs/HW-100-5_1024.jpg
None because using two SE OT's for PP is just a waste for me. Tube amplifiers are already big and heavy. I would never do that. If I want the SE sound I just use a SE amp. I use both PP and SE amps. I do not belong to any of the two sects.
The never high enough inductance of SE transformers will always generate non negligible distortion. Only with low plate resistance tubes like 300B and similar ones it is possible to achieve a ratio of 8-10 between inductive reactance and equivalent resistance (i.e. plate resistance in parallel with primary load) at 30Hz. At 20Hz it's impossible (not practical). With a PP amp achieving a factor of 20 or more even at 20Hz is easy. More than 10 @20Hz even in presence of imbalance.
The low frequency distortion harmonics WILL flood the entire upper bass low-midrange and likely cause more IMD.
Yes. Actually I have started to use only regulated supply.
Last edited:
Hopefully more enthusiasts would be capable of doing a periodic, quick THD spectrum using common software for sound systems (like REW), as that would cover not just mismatched idle DC current, but also large signal mismatch in output stage valves, and also relate to the total system performance (not just one test parameter).
Pardon me for dragging this out, but I've done the math and you +are+ off by a factor of 2 - differential voltage or not. Take a look at the attached drawing, which should demonstrate what I've been claiming..
I have depicted the W5M opt. circuit using the same values in the schematic for the 30R and 330R cathode circuit parts, with the Rp of the tubes represented by a pair of 8K resistors - one of which changes to 7K (to simulate a DC imbalance) in the second drawing. The change in current directly correlates to the drop across +one+ 30R, not two 30R in series (60R).
Legendre,
You challenged me to check my work. I had to find a very simple way to calculate the voltage between the 2 test points (which is 0.000V when the two tube currents are matched, but only if the 30 Ohm resistors are also matched).
Tube #1, 40 mA, through 30 Ohms #1 = 1.200V
Tube #2, 41.7mA, through 30 Ohms #2 = 1.251V
1.251V - 1.200V = 0.051V.
A difference in the tube currents of 1.7 mA = 0.051V
A 3.4 mA imbalance = 0.102V differential
You challenged me to check my work. I had to find a very simple way to calculate the voltage between the 2 test points (which is 0.000V when the two tube currents are matched, but only if the 30 Ohm resistors are also matched).
Tube #1, 40 mA, through 30 Ohms #1 = 1.200V
Tube #2, 41.7mA, through 30 Ohms #2 = 1.251V
1.251V - 1.200V = 0.051V.
A difference in the tube currents of 1.7 mA = 0.051V
A 3.4 mA imbalance = 0.102V differential
Yep, and there you have it. The current imbalance is reflected in the delta-V across one 30R part - not 60R. Glad we could reach an understanding on this - there's no point in arguing established electrical laws.
Last edited:
Sorry, Legendre was right. This claim doesn't correlate to another eqation in your same posting:
This is what you measure between the two nodes.
Best regards!
Edit: Was too late, Legendre yet has answered. Sorry again!
Which brings me to the question as an outsider .... what is the topic again/now?
Moderators??
As 20to20 mentioned (in vain) some 44 posts ago this discussion has gone abroad. Decidedly excellent representation of various magnetic aspects turned up - wish I had these sources and graphics available way back during my studies.
Nevertheless, nowhere during those physics tutorials did I see results relating to audio practice. This is still an audio forum/topic, and if it is possible (and has been for several decades of tube amplifier design) to design amplifiers with comfortably no audible distortion/artifacts, the value of recent magnetic gymnastics however worthy, remain academic. With much respect to the learned members, basics have been reraised, checkmates reached and subsequently overturned, etc. To repeat: I salute those.
If one can then return to amplifiers, at the appropriate post number I still could not see the nett practical advantage of using two S.E. topologies in preference to classic push-pull, audio wise. Again: The figures are there, abundantly.
Regarding the extra pole inadvertently inserted with separate bypassed cathode resistors (while normal discussion argues against l.f. poles), there has also been no comment as I recall. Where are we with the original subject?
Or am I off topic .... which is what again at this moment?😕
Moderators??
As 20to20 mentioned (in vain) some 44 posts ago this discussion has gone abroad. Decidedly excellent representation of various magnetic aspects turned up - wish I had these sources and graphics available way back during my studies.
Nevertheless, nowhere during those physics tutorials did I see results relating to audio practice. This is still an audio forum/topic, and if it is possible (and has been for several decades of tube amplifier design) to design amplifiers with comfortably no audible distortion/artifacts, the value of recent magnetic gymnastics however worthy, remain academic. With much respect to the learned members, basics have been reraised, checkmates reached and subsequently overturned, etc. To repeat: I salute those.
If one can then return to amplifiers, at the appropriate post number I still could not see the nett practical advantage of using two S.E. topologies in preference to classic push-pull, audio wise. Again: The figures are there, abundantly.
Regarding the extra pole inadvertently inserted with separate bypassed cathode resistors (while normal discussion argues against l.f. poles), there has also been no comment as I recall. Where are we with the original subject?
Or am I off topic .... which is what again at this moment?😕
OT posts moved to: - Status
- Not open for further replies.