Do you think this connection of a pair of toroidal transformers as OT is OK ??

[kagliostro]

I've find this schematic



do you think that the connection of the two toroidal transformers this way  can give the hoped result (PP cancellation) ?

Franco

[Heinz]

Hello Franco,

the basic principle of your circuit will work.

However, when using toroiodal transformers you have to be aware of the fact that even a slight mis-balance of the two idle currents will drive the transformer into saturation. This effect is specific to toroids. The sound quality will suffer severly if that happens. I've made experiments with toroids and found them to be very "touchy". If you insist on using toroids your circuit will definitely need a bias adjustment pot and permanent bias monitoring and correction. Even if the tubes are perfectly matched they will drift as they age. For Hi-Fi toroids can offer very good sound quality, especially when using power transformers but for guitar amps with their limited frequency range it just seems to make no sense to me. You will add a severe constraint for no benefit. Using a "regular" power transformer with the same specs (2x115V/2x9V) yields acceptable results. I have built guitar amps with these transformers and they sound decent.

That being said, you still have to consider two things. The power rating of a power transformer applies to 50Hz. Using it at higher frequencies requires a much higher power rating or sound quality will suffer. Oversizing the transformer by a factor of 2 or 3 worked well for me. In one case a very underrated transformer caused arcing and killed my output tubes at higher volumes. And that brings me to the second aspect: voltage rating. The primaries are rated for 115V each. When operating with AC the voltages across the coils can be much higher than half B+ due to the inductance of the transformer. Using two transformers for splitting the voltage  (as shown in the schematics) is certainly a good idea but still 560V would be too high for my taste. As soon as the transformers reach their limits of "regular" operation - which is somewhat unpredictable in a context they were not designed for - bad things can happen.

Here's some interesting reading about PTs as OTs:
http://www.edn.com/design/components-and-packaging/4369085/Using-a-power-transformer-at-a-frequency-it-wasn-t-designed-for

[kagliostro]

Ciao Heinz thanks for answering

Previously I know only a part of what you say

OK for the easily saturation and the necessity to use a largely overrated PT as OT, but what about the fact that of the two transformers one semi primary is used for one plate and the other for the other plate (same thing for the other transformer) ?

Do you think this form of connection can solve some of the problems ?

Franco

[Heinz]

The PP cancellation will take place no matter how many transformers you combine. If you use two as shown any imbalance between the two on the primary side will be cancelled on the secondary side, provided the two transformers are identical. However, in reality they are not. So in general, adding transformers will worsen the results. However, I'm talking about theory here. The difference will most likely be inaudible. But again, I think the circuit adds complexity for no benefit.

The above circuit will not solve the idle current issue, either. Let's assume that the left tube has 50mA idle current and the right one has 60mA. This means that L1 and L7 (which belong to the same transformer) are 10mA off. The same is true for L2 and L6. This mis-balance is not an issue for an ideal transformer but it is for real-world applications and for toroids in particular.

The only advantage of the circuit that I can see is voltage sharing (EDIT: and power sharing) of the primaries. However, this only applies as long as the transformers are operating within their specs. When driven into saturation, two transformers may behave even wilder than one and cause higher voltage spikes which are more likely to damage your tubes.

[kagliostro]

Thanks for the explanation

About the reason for the use of a pair of those transformers

I think the principal reason for this choice is the low cost of used toroidal that are more easily available respect traditional OT

which can be bought at a very higher price also if used

Franco

[jazbo8]

This is a typical case of "too clever by half", using two toroids regardless of the "clever" connection, is going to make the task of DC balance any easier, on the contray, it may even make it harder... There are plenty of good quality EI push-pull OPT's at reasonable costs, which makes using toroid (especially the PT's that were not even designed as OPT's to begin with) less viable as it would first appear on paper.

[HotBluePlates]

I also don't see the value in working so hard for a self-split output stage...

Yes, there is a clever way to run it single-ended. But you can run any true push-pull output stage single-ended and still balance d.c. in the output transformer.

[Heinz]

OTs have always been more expensive than power transformers. They are low-volume products (not sound-wise ;-) and production requires much more attention than plain old power transformers. So the idea of using cheap power transformers and save the money for the OT has always been intriguing. I started my tube experiments with cheap tubes and power transformer OTs. The main reason was that I simply didn't want to spend any money for my experiments. In a way I approached the topic "in reverse". I have never tried to build anything specific (e.g. a particular amp model) but instead, I tried to come up with something functional using only parts I had in my box.

So I played a lot with different cheapo tubes and transformers. You can get pretty good sound quality using big power transformer OTs when you drive them with low impedance (low plate voltage, high plate current) at low power. Considering theory, it's actually not surprising. Some power transformers sound better than others but you never know before you try. I have a simple 2W tube amp for my PC using power transformers and PCL82 tubes. It's probably not Hi-Fi but it has a very pleasant sound. There is nothing wrong with a circuit if you like it.

It's true that power transformers are slowly becoming extinct these days. They are superseded by switching power supplies, due to cost, weight and space reduction. However, it seems that toroids are among the few "survivors" so it seems legitimate to consider (mis-) using them. But as discussed above, in my opinion the disadvantages outweigh the cost savings by far. If you can get hold  of a "normal" (i.e. non toroidal) power transformers with the same ratings, the circuit  will probably work a lot better. But a true OT will always be superior.

[jazbo8]

So I played a lot with different cheapo tubes and transformers. You can get pretty good sound quality using big power transformer OTs when you drive them with low impedance (low plate voltage, high plate current) at low power. If you can get hold  of a "normal" (i.e. non toroidal) power transformers with the same ratings, the circuit  will probably work a lot better. But a true OT will always be superior.

Agreed. For low power applications, the slightly overrated toroid PT's could indeed work fine as OPT's, but it gets more difficult as the output power goes up. Since guitar amps have limited frequency response, it really does not cost the manufactuers any more to produce "guitar" OPT's than "normal" PT's, so as we have been saying, there is little to be gained by saving pennies with the toroid or "normal" PT's.

[HotBluePlates]

I also don't see the value in working so hard for a self-split output stage...

... the idea of using cheap power transformers and save the money for the OT has always been intriguing. I started my tube experiments with cheap tubes and power transformer OTs. The main reason was that I simply didn't want to spend any money for my experiments. ...

Heinz's desire to avoid spending money would be a good reason to work hard to make toroids function in an unusual circuit. In fact, a lot of (most of?) engineering in all fields is driven by cost constraints.

If cost was a motivating factor in the original circuit Kagliostro posted, then I retract what I said. I glossed over the statement that the circuit used "[two] 110v:9v toroids".

[drgonzonm]

I am not familiar with using torioidal transformers in SE circuits much less PP circuits.  Since most SE transformers have an air gap to control saturation, raises the question how do you control saturation.   

It is my understanding that the material selection for the cores of a PT and OT are usually different, as are the lamination thickness.

Franco, your willingness to think beyond, the traditional limits is one reason I enjoy reading your posts.  Please continue to push the imagination limits. 

Drgonzo.