Sizing PP OPT

[alexg]

I have been asking around on how to size push pull OPT and am getting various answers:

1. Size it plate-to-plage
2. Add the current through both tubes and add 20%.
3. One suggests to use the power listed for a PP configuration of the tube, and get a PP OPT that can handle the power.

In the Philippines, ready-made OPTs (Hammonds, Tamura, James, etc) are hard to get, so most of us are forced to have them custom wound, and when having it wound, I need to give the winder the specs.

Thanks.

[SY]

There have been books (many) written on just this topic. Basically, you need to specify plate to plate impedance, power, and bandwidth. The last two will determine the size, the first will determine turns ratio.

As an example: Want full power at 10Hz? Prepare for a hunk of iron the size of a refrigerator. Willing to live with 20Hz? You've dropped the size by a factor of four and probably reduced the leakage inductance.

If you have to tell the winder how to size the core once you've given him the outline of what you want the transformer to do, that's a good indication that you're dealing with the wrong guy.

[Yvesm]

One more once

http://www.dissident-audio.com/OPT_da/Page.html

Look at RDH4 Chapter 5 before.
It contains unvaluable informations.

Yves.

[alexg]

What I really meant by my question is:

If I bias one tube of the PP pair at 40ma, should I get an OPT that is rated at 40ma or at 80ma? or some values in between?

Thanks.

[dsavitsk]

http://www.audioasylum.com/forums/tu...12/123015.html

[ray_moth]

Dsavitsk, that link discusses SE transformers, which are subject to core magnetization because of the net DC in the primary. With PP, that is not the case.

What is meant here, if I understand the question correctly, is what primary current should be specified so that the gauge of wire used will be sufficient. If 40mA is the quiescent current per OP tube, then that will also apply to each half of the primary winding. However, it's not as simple as that, because it also depends upon the class of operation of the OP tubes.

If the amp is Class AB or B, the current at full signal can be several times as great as the current with no signal. The primary winding needs to be able to cope with the maximum possible current. For example, EL34s in Class AB1 could be biased to draw 30mA per tube at zero signal but may well draw 100mA per tube or more at full signal. In such a case, it would be sensible to design the primary to tolerate at least 120mA (or even 150mA) in each half, to avoid the danger of overheating if the amp is run at full power. You don't want the OPT primary to be the weakest link in the chain - it's not a fuse!

[Yvesm]

Originally posted by ray_moth

If the amp is Class AB or B, the current at full signal can be several times as great as the current with no signal. The primary winding needs to be able to cope with the maximum possible current. For example, EL34s in Class AB1 could be biased to draw 30mA per tube at zero signal but may well draw 100mA per tube or more at full signal. In such a case, it would be sensible to design the primary to tolerate at least 120mA (or even 150mA) in each half, to avoid the danger of overheating if the amp is run at full power. You don't want the OPT primary to be the weakest link in the chain - it's not a fuse!

IMHO, the best approach is to compute the DC resistance of the primary winding.
So you can find how much power it will dissipate according to the mean AC and DC current flowing thru it.

Much better that any rule of thumb based on "acceptable" current density, I think !

Yves.

[ray_moth]

I never mentioned current density. Alex wants to be able to instruct his transformer maker about the primary current that should be allowed for. Probably the easiest way to find that out will be to look up the max. plate current in tube data sheets for the type of tube, class of operation, load impedance, B+, etc. that he has in mind.

[Yvesm]

Hi ray_moth,

No, YOU didn't !

But guess what the winder will do ?
He should be informed about DC and AC currents to take its decision about wire gauge according to the way the heat is extracted from the windings.
It could be quite different if the design uses bells or if potted in some more or less temperature conductive material.

Not to tell about ambient temperature, probably much different in Manilla than here

Yves.

[retailer]

I've wound 5 output transformers for guitar amps and I guess the same applies to all OPT's, I use a figure of 3 amps/square mm for the current loading of both the primary and secondary. For the valve anode current I take the max anode current given in the valve data book and add a 25% margin for safety. So for an EL84 this works out at about 75mA, using the 3 amps/square mm figure, wire size is about 0 .18mm for the primary and up to 1mm diam for secondary depending on the secondary impedance. I don't have any 0.18mm wire so I use 0.2mm wire which I have plenty of. For guitar amps I don't use elaborate interleave just a simple P-S-P-S. I always test with a sine wave and run at full power for at least 2 hours, sometimes the anodes start to glow a dull red after an hour or so like this ( you can see it in the dark) but I have never had a transformer failure. They get warm but never so hot than you can't hold your hand on it. I size the core (which is a reclaimed power tranny core) to give me a 75%-80% fill of the winding window. When I wound my first OPT for a PP 6V6 amp I only had 0.16mm wire that I had pulled from an old telephone relay so I used that. The amp is still going after 2.5 yrs and being used weekly in a recording studio, although it almost never gets used at full power. Initially I was concerned about the wire size so I ran some tests on the current carrying capacity of enameled wire. I soldered about 5 cm of 0.16mm wire ( at 3 amps/square mm it will carry 60mA) to some leads and then wrapped the wire around a digital thermometer thermocouple, I slid this into an empty matchbox and then wrapped an old rag around it for insulation. Using a variable power supply I ran 60mA through the wire and after 2 hours temp had gone up only 0.5 deg C, I increased current to 100mA and after 2 hours temp had gone up only 2deg C. 3 amps/square mm is a safe figure to use I can't speak for anybody else but it works for me. I have unwound 2 failed OPT's and it looked to me that the failures in both cases were caused by arcing, in one instance between the laminations a primary wire an in the other instance between primary wires placed too close together.