Question:disadvantages for using a SE OPT on parafeed?

[pauldune]

I have a simple question:

What disadvantages are there for using a (airgapped) SE output transformer on a Parafeed amp?

I already own the SE OPT, so the cost-aspect is not what I mean...

Gr. Paul

 

[Frank Berry]

I may be wrong but I don't think you'll have any problems using an air gapped transformer.

 

[Bas Horneman]

What disadvantages are there for using a (airgapped) SE output transformer on a Parafeed amp?

Depends. What are you going to use the airgapped SE transformer for? As plate load or as output transformer.

Obviously as a plate load it is perfect. As output transformer it will work but you don't have the advantages a non gapped output transformer has.

 

[Shoog]

These transformers will be optimised for use when passing current. Hysterisis will be greater than a none air gapped transformer, and this will blur the sound somewhat.
Not ideal and will likely sound no better than using it as intended in SE duty.
Power Toroids can make good parafeed output transformers if you need a low impedeance plate load - they will likely sound better than the SE transformer.

Shoog

 

[kevinkr]

<snip>
Not ideal and will likely sound no better than using it as intended in SE duty.
<snip>

Shoog

IMHO this SE gapped OPT is not likely to sound as good as the transformer properly employed. For parafeed you should use an ungapped transformer which typically will also have much higher primary inductance than your gapped SE OPT, better for bass performance (frequency extension and LF distortion) in concert with a reasonably sized plate choke. Note that for a given power level the transformer can a lot smaller which makes designing and building a good transformer a lot easier. (Assuming this is a PA)

All that said I am not that big a fan of parafeed at all, although it does have real technical advantages in terms of PSRR which is usually excellent, and good power handling and AC performance from reasonably sized transformers.. Choke design is an important aspect to overall performance and one has to select the parafeed cap carefully and while electrical theory would indicate that location in the circuit is almost irrelevant in practice that doesn't seem to be the case. (Ground side of transformer, versus signal side - I prefer the ground side.)

Hysteresis about the BH curve 0 crossing could be worse in a gapped SE transformer by virtue of its larger than required core geometry. (As compared to a design specifically meant for use in parafeed configurations.)

Part of the charm of SE transformers is that the magnetic flux never reaches the 0 crossing on the BH curve, and therefore the magnetic domains do not have to flip polarity at the 0 crossing which takes time and some energy and gives rise to the hysteresis previously mentioned. This was the original argument for SE and was quickly forgotten (by some) when parafeed came into vogue.. As far as transformer operation is concerned parafeed creates essentially the same AC conditions as exist in a PP OPT which is something of an issue in my mind. (Perhaps unjustifiably so)

In any case just my opinion.

 

[pauldune]

Ok, and thanks so far.

As for the hysteresis, that is the reason I asked this question.
Problem is, does anybody have or have seen the hysteresis of the same OPT, with and without airgap?
At this point I'm not convinced the hysteresis would be worse for the air gapped one, only that it saturates less, and that the airgapped one should be much larger for the same L. Or is it a fact, that for the airgapped one it is wider?
Can you measure hysteresis? (yourself)

Oh, and its loaded with a CSS, not a choke. (not that that makes a difference for the question).
I also happen to have a PP toroid OPT wich has a 4k p-p primary, and I will compare it to the 3k5 AE-europe SE (when I get it). Ill post the measurements and subjective opinion here when its done.

 

[kevinkr]

Ok, and thanks so far.

As for the hysteresis, that is the reason I asked this question.
Problem is, does anybody have or have seen the hysteresis of the same OPT, with and without airgap?
At this point I'm not convinced the hysteresis would be worse for the air gapped one, only that it saturates less, and that the airgapped one should be much larger for the same L. Or is it a fact, that for the airgapped one it is wider?
Can you measure hysteresis? (yourself)

For a given transformer I would not expect the hysteresis to change appreciably whether it were gapped or not. (I'm not completely sure about this, but I believe this to be the case.) A transformer designed for parafeed operation will be much smaller, and given equal (or better) core materials this should result in less hysteresis about the 0 crossing.

You can measure hysteresis yourself but doing so is not trivial at anything other than line frequency. (I've not done it in recent years.) Some information on doing this can be found here:
589 LABORATORY MANUAL: Block D Transformers (Note that this is targeted at 60Hz based measurements, if you want to do this at other frequencies you will need to change the integrator time constant accordingly.) Test set up here: http://eece.ksu.edu/~starret/589/man/Dfigs.html

 

[Shoog]

I would say that when you compare the performance of your SE to your toroidal PP, the PP will come out a winner. The reason that a toroidal is so difficult to use is its very poor tolerance to DC imbalance which equates to very low hysteresis in the core - the opposite to the SE.
The only thing to be careful with is interwinding capacitance - the secondaries should be tried in both directions to minimise this.


I am no longer a big fan of the parafeed configuration - after trying it a few ways. This is because the quality of the cap tends to dominate and it is not a trivial matter to find a good quality one at the required capacitance and voltages. I had good success with some of the old PIO caps, but they are not that neutral - more vintage sounding (ie soft top end).

Shoog

Shoog

 

[pauldune]

You can measure hysteresis yourself but doing so is not trivial at anything other than line frequency. (I've not done it in recent years.) Some information on doing this can be found here:

OK! I guess i'll just have to try it, because that is way to complicated and expensive for me....Money I could spent on a good OPT. Thats one of the reasons I wouldnt buy a non airgapped transformer, because i want to find out which topology suits ME best For me, if it sounds good, it IS good.

Shoog: , I'll try them both, and see who wins.
The schedule i'm using it in is something I came up myself with, and its a DC coupled, CCS loaded driver-> CCS loaded parafeed output stage.
Because I made it up myself, I wasnt even shure it would work at all, and that the bias setting would work, and that is was stable and didnt oscillate.

But it sets itself beautifully, and the prototype I build with a d3a-triode driver and a EL34-triode, while using very small C-core philips output transformers (10k primary), and only wielding 0,5 watts (because of the 640v DC; it should be about 750-800v) is the best sounding amplifier I ever heard...

If I get the SE OPT and the jj2a3-40's and the new powertrannys, I'll build it asap. And if that works like I intended it, I'll post the schedule here.
(that way less chance getting ridiculed over some stupid mistake)