Power Transformers

[Structo]

I didn't want to hijack Jason's thread about his new power transformers so I started this one.

ToneMerc said this:

Based on the very similar ones I've been ordering at 340/345, I'm predicting about 448-455V under load biased around typical D figures. A tad highter if the line voltage is closer to 124VAC.

Hey TC can you tell me how you arrived at that voltage?

I know we have some ball park numbers to multiply the secondary voltage with to arrive at the B+, but they never seem to work for me.

I thought with SS rectifiers you multiplied the voltage by 1.41?

I wish all the transformer companies would all use the same standard for listing their voltages.
Or at least say if it is RMS or Peak or under Load.

Magnetic components, for example, give a figure of 420v on there Twin HT secondary, which I assume is the loaded value.
But I achieve 435-438v in my amp with it.

[M Fowler]

Must be assuming 1.3 x one leg 355v

[David Root]

Tom usually they specify DC mA at idle or under load, both RMS, not peak. From what I've seen there is 5-7% difference between the two on any given PT.

But they generally don't tell you which loading applies to their published DC mA number. I imagine people have just asked each mfr which they use for their published rating.

I have been told that M-C specify under full load. Edcor do the same. Not sure what Heyboer does, or Mercury Magnetics.

As to the voltages, HT DC voltages are spec'd the same way, so I would assume that if Mfr A specifies 420V DC at full load, for example, the 450 mA is a given as the FL current. So yes in that case you would expect to see more than 420V at idle current.

Further problem is how many tubes you have hanging off the PS. If you have only three 12AX7s and two 6L6GCs, you'll see a higher idle voltage than if you have four 6L6GCs and six 12AX7s and two 12AT7s.

Taking this ad absurdum, changing tubes will alter that because no two tubes draw exactly the same current. What you described at idle 435-438 is in the same range I get with a 100W #124, no buffered loop, 437V approx. with the very same PT.

I think you can guesstimate it +/- 5V if you know the PT, but that is about it in my experience.

If you use the "cheat sheet" method, take 437/0.9=485.6, X 1.4=679.8, /2=339.9 VAC on the secondary. Or do it the other way around, starting with the specified VAC.

Having said all this, it is widely held that iron mf'd by these reputable companies that we use has a service factor of about 1.3 on rated current draw overall. Not sure how that splits out between secondary windings.

[Structo]

Thanks David.

The MC PT is rated at 420v. But under load and rectified I get 435v on the plates.

I realize that a Fender Twin Reverb is a bigger load than my amp.


The schematic of a AB763 shows 460v on the plates and 458v on the screens.

But that amp has 10 tubes, 4 x 6L6 and 6 x preamp tubes.

What I don't understand if it is a replacement transformer for a Twin Reverb, why is the voltage lower than the original?
I understand the wall voltage is higher now but all the schematics show around 460v or higher on the plates.
Also, the heater voltage is high at 6.6vac, again fewer tubes in the Dumble.
The AB568 Twin shows 470v on the plates!

http://www.el34world.com/charts/Schemat ... _AB568.pdf

I'm not saying I want my plates that high as the current tubes probably last longer with a lower voltage on them.

Another question, is it the resistance of the wire in the transformer itself that causes the voltage to sag?
Shouldn't a transformer be able to supply the rated voltage up to the specified current?

Because if the transformer outputs 440v without a load but 420 with a load, there has to be loss somewhere and my guess is that is lost in heat generated by the PT.
Because these Mag Comp transformers get pretty warm.

[ToneMerc]

Some it I bet is core loss from the steel type, I usually request M6 steel laminations from Heyboer.

TM

[David Root]

No, if you get 440V without a load you can't expect to maintain 440V at full load, it must be less because the winding is designed for a certain VA (even though only Hammond seems to publish VA ratings). If you increase the current, the voltage has to drop, V=IR. There is no free lunch. Not to mention eddy currents and other leakages and inefficiencies.

Yes you can stiffen the PS with a big choke and big caps, but that can be overdone with a detriment to tone.

The mfr's SF on a winding VA may reduce the voltage difference, not sure about that.

If you can regulate the HT at 440V with full rated current load you could do it but you are biting into the SF, and I have yet to see an audio grade voltage regulator (not rectifier) chip that will regulate 440V at 500 mA. Although there are ways around that, it gets pretty complex. See Merlin's new Power Supply book.

I wouldn't want main PS voltage regulation anyway because it would interfere with the PS dropping resistors' effect on interstage voltages and the resultant tonal and touch sensitive characteristics of the amp's designer, which has been the subject of a recent lengthy thread here. That is also one area in Merlin's new Power Supply book that I personally don't think I'd use in the main PS line(s), for those reasons.

[Roe]

I'm using webers export twin reverb type PT (W022756EU used on the 6a80). its speced at 640vac, I get around 420v at the plates (when the wall voltage is a tad low)