• 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.

Push pull transformer SPICE simulation from the spec

@gorgon53:
Ignoring a parameter in a design isn't a good criterion. I have no error and in fact the problen isn't mine.

In a real tranformer, k is easy to measure. Do it, else the simulation will be ever worse than they normally are.
May be reading the original question will help you to understand wich answer is wrong and wich is not, it started with:
Let's take this one and for the moment put aside resistance, capacitance and leakage. The remaining spec....
 
They must not be ignored nor neglected if you want a close to real world simulation. Each parameter is important and cooperates in the ttansformer behavior. There is why to put them in tje equivalent circuit.

You can ignore or neglect one or more of them, but these action will degrade still more the resulting of the simulation, that is per se of little confidence.
 
They must not be ignored nor neglected if you want a close to real world simulation. Each parameter is important and cooperates in the ttansformer behavior. There is why to put them in tje equivalent circuit.

You can ignore or neglect one or more of them, but these action will degrade still more the resulting of the simulation, that is per se of little confidence.
Did you READ what he asked for???
He said:
.....for the moment PUT ASIDE resistance, capacitance and leakage. The remaining spec....

He knows that those factors exist but he either does not have their value or does not care.
Off course they can be ignored if he choose to do so for whatever reason. And his simulation will be garbage in-garbage just like any simulation that does not include everything that has an impact on the outcome (does output loads, such as "crossover", or "speaker", or passive ring a bell?).
The inclusion of the k-factor won't make much difference at midfrequency, output load does.
 
Last edited:
Guys, please, stop. I just wanted to ask am I reading the spec correctly and is those 591H actually an open circuit inductance from plate to plate (not from the center tap)
it is from plate to plate and if it comes from a manufacturers datasheet it is either calculated or measured under conditions that give the highest possible result, don't take it as carved in stone, in a pp transformer it varies wildly to the downside
 
  • Like
Reactions: easphyx
gorgon53,

You are correct, the transformer of the current thread in question is Push Pull (real, or unreal, or without enough specifications).

I mostly tested single ended transformers with that Vector Network Analyzer.
I had lots of access to multiple models of se output transformers.

I also had access to push pull output transformers. I tested them, and came to similar measurements and conclusions.
However, those tests of push pull output transformers were not completely fair. The vector network analyzer was single ended, it was not balanced / differential.
Do you connect the vector analyzer signal to . . . Plate to center tap. Not fair, too many stray capacitances, the secondary has to be connected to the same ground (unbalanced). And, the other plate connection is left "swinging in the breeze".
Or, do you connect the vector analyzer signal to . . . Plate to Plate. Not fair, too many stray capacitances, the secondary has to be connected to the same ground (unbalanced).

I also tested single ended interstage transformers, and push pull interstage transformers with the vector network analyzer.
Again, the push pull results were less certain and problematic, because of the single ended input and output of the VNA.

I try real hard to be fair when I do testing, and when I report testing results.

If you want the highest possible operating inductance from a push pull primary, the first order of business is to exactly balance the quiescent DC current of each plate winding.

Just my opinions.
 
Basically the question is how to read transformer specs

You can't until you have in your hand the OT. This is the main reason because you can't simulate an OT only by a few parameters that normally any brand publish.
After that you can measure it at the best with a simply backward method; easy and give you all the informations about freq. answer, THD vs level and frequency.
Normally with a good sound card and Arta o Rew you can read until 96 kHz the perfromances and this is enough.

Walter
 
gorgon53,

You are correct, the transformer of the current thread in question is Push Pull (real, or unreal, or without enough specifications).

I mostly tested
Do you connect the vector analyzer signal to . . . Plate to center tap. Not fair, too many stray capacitances, the secondary has to be connected to the same ground (unbalanced). And, the other plate connection is left "swinging in the breeze".
 
6A3sUMMER,
Must have given great insigth having such a vectoranalyzer. But i thinck unbalanced testing schould have given useable results for some pp cases, for instance when testing OPTs intended to work in class B,
one could look at class B as one side is working and the other side dangling in the air
 
Last edited:
A set up to test for saturation of a push pull transformer is to build the amplifier, apply a clean sine wave, and decrease the frequency from 100Hz, to 40Hz, to 20Hz, to 10Hz, then 5Hz (do not apply global negative feedback).
First try that with a 5% quiescent current un-balance. Example, 47.5 mA push; 52.5% pull. That is 100mA total, and 5% un-balance.
Look at the shape of the waveform.
Second, try it again, this time with the quiescent currents very well balanced. Look again at the shape of the waveform.

The very last page of the manual for the Heathkit W5M has a graph of the harmonic distortion of various Bass frequencies, versus the amount of unbalanced primary current.
What is noteworthy is the distortion is quite large, even though global negative feedback is connected (global negative feedback pumps in even more current, in an attempt to correct for the saturated laminations). A dog chasing its tail, so to speak.

The classical definition of Class B push pull is zero quiescent current in both the push tube, and the pull tube.

The practical "Class B" push pull amplifier often has a little quiescent current in the push tube, and a little quiescent current in the pull tube.
Since there is quiescent current there, I believe it makes sense to balance those two low quiescent currents.
When a small signal is applied, then one tube cuts off, and alternatively the other tube cuts off (signal with positive and negative alternations).
You might consider this to be a Class AB amplifier which almost immediately goes from both tubes slightly conducting, into AB mode . . . alternating tubes cutoff, with a very small signal applied.

Just my opinions.
 
Last edited:
  • Like
Reactions: gorgon53
"A set up to test for saturation of a push pull transformer is to build the amplifier, apply a clean sine wave, and decrease the frequency from 100Hz, to 40Hz, to 20Hz, to 10Hz, then 5Hz (do not apply global negative feedback)"


The only way to get the real performances mainly in low end (from low level to the max perimissible, p.e. 3% of THD at 20Hz) around the OT only is to drive it with backward method; a good ss amp that drove a secondary woth a nomila Z ( p.e. 8 ohm) with a clean 20-25 volt rms can do the job in the perfect way.
Then with the test of THD vs Frequency and THD vs level it is easy ( with Arta or Rew) to read the real world non virtual.
If the ratio is 25:1 (just an example) with 20 Vrms of driving you will have around 500 Vrms on secondary (a-a)

Walter
 
25 Vrms into 8 Ohms is 78 Watts!

500 Vrms a-a = 707V peak a-a.
Just be sure the transformer is rated for 80 Watts.

A 500V to 25V transformer turns ratio is 20:1.
The transformer impedances: 3200 Ohms to 8 Ohms.

A 25:1 turns ratio is a 5000 Ohm to 8 Ohms.
 
25 Vrms into 8 Ohms is 78 Watts!

500 Vrms a-a = 707V peak a-a.
Just be sure the transformer is rated for 80 Watts.

A 500V to 25V transformer turns ratio is 20:1.
The transformer impedances: 3200 Ohms to 8 Ohms.

A 25:1 turns ratio is a 5000 Ohm to 8 Ohms.

What I wrote is an example just to explain where it is possible to have as test signal
Then everyone can work properly with loads and level
 
"A set up to test for saturation of a push pull transformer is to build the amplifier, apply a clean sine wave, and decrease the frequency from 100Hz, to 40Hz, to 20Hz, to 10Hz, then 5Hz (do not apply global negative feedback)"


The only way to get the real performances mainly in low end (from low level to the max perimissible, p.e. 3% of THD at 20Hz) around the OT only is to drive it with backward method; a good ss amp that drove a secondary woth a nomila Z ( p.e. 8 ohm) with a clean 20-25 volt rms can do the job in the perfect way.
Then with the test of THD vs Frequency and THD vs level it is easy ( with Arta or Rew) to read the real world non virtual.
If the ratio is 25:1 (just an example) with 20 Vrms of driving you will have around 500 Vrms on secondary (a-a)

Walter
Walter,

I was hoping from previous threads here on diy-audio that you would finally understand that the backward method is NOT the right way to measure this. I see you are again trying to convince people, too bad because what you are doing is wrong. As long as output impedance determines harmonic distortion and as long as a series connected resistor (or a high output impedance) in your backward method will determine distortion, you will not find the right answers to this question. All explained and proven in a previous thread here on diy-audio.
 
Hi Mr 80 mV
How are you?
Have you studied?
It seems no
You haven’t proved nothing because you aren’t able


Walter
Are you still in lie mode? What a pity for this forum.
You can look back on the forum that you didn't explain anything and others, including me, did.

Prove Walter, come up with evidence instead of your silly stories.
Walter, start proving that The Radiotron Desingers Handbook is wrong, because what you write all the time is not in this book.
 
Last edited:
This post is the last where I answer to you
You aren’t able to show any test that can be considered esplicative
Only copy and paste as the lasts
While I sent dozen of the measurement and some of them done also in the standard way

Bye Mr. 80 mV
 
This post is the last where I answer to you
You aren’t able to show any test that can be considered esplicative
Only copy and paste as the lasts
While I sent dozen of the measurement and some of them done also in the standard way

Bye Mr. 80 mV
Because you never designed a transformer yourself, never read/understand the design/theory articles about transformer yourself I am sure that you are wrong.
Measuring a lot transformers doesn't mean a thing if you not understand what you are doing. Your backwards methode is not standard and especially not the way YOU do it.
I explained with several scientific artical that you are wrong so don't ly here, it's a shame that you do that all the time.