toroidal SE output transformer

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The DC bias current for the output stage flows in the primary of an SE transformer. You need some sort of gapin the core to prevent the transformer from saturating or operating in the extremely nonlinear region of the B-H curve and producing loads of distortion. Ferrite is a poor choice for core material, as it has about 1/5 of the usable flux swing of a good silicon steel core. This means that a ferrite core must either have 5X the core area or 5X the turns to equal the low end performance of a silicon steel transformer. It needs to be gapped, too, which reduces the performance yet again.
 
Powdered iron would be perfectly abysmal as an audio transformer core, as the permeability is way too low. The bulk of the current from the output tubes would go to primary magnetizing current rather than to the load where it belongs. Silicon steel laminations of 0.25mm thickness would work better than anything else suggested here so far, but the thickness wou,ld result in eddy current losses at the higher frequencies. This may not be to bad, considering that the amount of power present (or necessary) at high frequency is much lower than at bass frequencies. One could compensate by using a larger core (more laminations) or more primary turns to bring up the primary inductance and also lower the core flux density. Laminations of 0.1mm thickness would be better if they are available. The laminations will need to be gapped in order to be able to tolerate the DC bias current from the output tubes without saturating. This will lower the primary inductance of the transformer, so the primary turns will need to be increased to bring the primary inductance back up to prevent excessive low frequency rolloff. are you trying to do this project on the cheap, or are appropriate transformers simply not available in Turkey? Winding your own transformers will take a lot of effort to yield a good product.
 
transformer

Actualy i can order transformer from abroad but custom makes some problems and its cost may rise up. I found a place in Ankara city. They are doing transformers and tube amplifiers. He told me that; he is using 0,25mm laminated steel and double C core. He says he measured the performance and his transformers are equal or better to popular Tamura or Tango. He asks 80 dollars each for 30w SE transformer. I am now using Hamond 1640SE in my parallel EL34SE amplifier. He said that Hammond bass response is good but high frequency is not good. I checked it and he was right. I want to try local product for my new 813 amplifier project.
Noyan
 
Hi,

It is actually very difficult to compare transformers from different manufacturers as most of them give very few values and almost no one give you the conditions under which the parameters are measured.

0.25mm grain oriented silicon iron in a C-core can give quite decent performance and can be compared to Hammond or cheaper type transformers from Tango or others if the winding arrangement is done in a competent way.

Ask your manfacturer for values of of primary inductance at low AC level and at peak permeability, distortion at low frequency and leakage inductance and of course the measurement conditions for these values, if he can't give you these values he probably doesn't know what he is doing anyway so you better stay away.

BTW as you need a SE transformer it should be optimised for a certain value of DC current so ask him for the spec of a transformer for the value of DC current you are using.

Regards Hans
 
transformer

Thank you!
He said that his product gives 5hz to 50 khz. (20-20000HZ /-1db) While i discusting with him, i saw he has technical knowledge. But as you said not possible to be sure about the product. I gave him circuit anode voltage is 900 volts -90ma.
Noyan.
 
I have done some thinking about the requirements for a SE transformer, and came to the conclusion that 0.25 mm laminations really aren't so bad after all. Postulate that the transformer should have at least a 20Hz-20kHz operating range. For constant amplitude excitation, the largest flux swing will be at 20Hz excitation. The turns and core area should be optimized for a reasonably conservative flux swing at this frequency. For silicon steel, a reasonable value of peak flux swing is 10-12 kilogauss. Cheapskates will push it all the way to 15 kG, which is verging on saturation. With constant amplitude excitatiion, core flux swing decreases linrarly with frequency. Thus the same excitation level that gives 10-12kiogauss at 20Hz will give 10-12 gauss flux swing at 20kHz, three orders of magnitude less. This is why relatively thick laminations can be made to work.
Since there is an AC flux swing riding on top of a DC flux level, the design of a single ended transformer is a little more complex than a push- pull, where the oppositely phased primary windings cancel the DC flux, but the basic principles still apply. The transformer maker you mentioned seems to have his technique down well, as it is difficult to get the wide frequency response you mentioned without interleaving and other winding techniques to reduce parasitic reactances. The price is also right. As has already been mentioned, the only other consideration is to optimize the transformer gap and turns for the quiescent current of the output stage. If you go with this transformer vendor, let us know how it works out...
 
He said that his product gives 5hz to 50 khz. (20-20000HZ /-1db) While i discusting with him

This doesn't say very much, you need to know at what power level and at what DC current this was measured otherwise these are figures without value, you also need to know operating conditions to see that it is the same as in your case, Rp, RL, current.

Normally distortion at low frequency is the determining factor when designing transformers so please ask what the distortion is at full voltage swing and 20Hz. It is not so difficult to design a transformer with wide frequency range but to have both wide frequency range and low distortion is much more difficult. Unfortunately very few manufacturers give full data including operating conditions.

Regards Hans
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.