Hello all.

Please excuse my relative ignorance, on this topic at least...

I'm working on a design for a valve audio amplifier. Unforunately, it's not at all straightforward for me to post schematics here- my internet access is only at the library and hence quite limited.

I'm using a Lundahl power transformer, LL1649. A datasheet is available at http://www.lundahl.se/pdfs/datash/1648_51.pdf

(my apologies- as usual I can't get the url feature to work- this is the correct address, only there is an underscore as follows- "1648_51"- this page can easily be reached from http:www.kandkaudio.com/datasheets.html

This is a C-core transformer. To quote the datasheet :

Magnetic stray is​

extremely small if secondaries of the two coils are loaded identically.

I'm using 3 preamp tubes, each drawing .3 A at 6.3v, and a couple other variable loads that will be run off the 6.6 volt taps. All of these loads could run off AC or DC, so I'm thinking AC for simplicity's sweet sake. So how to load the two transformer coils identically?

One idea I have is to parallel two of the 6.6 volt taps, one from each coil, for 3.1A+3.1A=6.2A at 6.6 volts. (On the datasheet diagram this could be, for instance, paralleling taps 17/15 and 8/7.This could supply adequate current and an appropriate voltage for all the low voltage needs and place an "identical" load on the two coils, hence keeping the magnetic stray "extremely small."

Is there anything wrong with this scenario?
Thanks for your help.
Ted

 

There's no problem with paralleling the secondaries. Just be sure to keep the phases correct. If you connect them out of phase you'll get zero volts.

 

If you connect them out of phase you'll get zero volts.

...and a buttload of current, since it's just like shorting them.

 

There's no problem with paralleling the secondaries.

If and only if the secondaries are matched (same number of turns) to give same voltage. If not, you will have circulating currents.

Sometimes the windings on a given transformer are not identical even if they have the same voltage rating.

 

The mere fact that the windings are terminated at different physical points on the core as a matter of building them causes a difference of a tenth of a turn to a 4th of a turn. Hook them up with no load and measure the current in any one wire. That will tell you whether they are going to fight. Do not use a clamp meter or a Hall Effect magnetic flux measuring probe. The magnetic field from the core will mess up the measurement.

 

You can also measure the AC voltage between the two windings (with the other ends connected together) to see what the voltage difference is between them.

 

Thanks so much for your comments. What an good forum this looks to be!

The mere fact that the windings are terminated at different physical points on the core as a matter of building them causes a difference of a tenth of a turn to a 4th of a turn. Hook them up with no load and measure the current in any one wire. That will tell you whether they are going to fight. Do not use a clamp meter or a Hall Effect magnetic flux measuring probe. The magnetic field from the core will mess up the measurement.

So what kind of values of these measurements will tell me if they are "going to fight" or not? If the two are identical? Close? Within .1 volts of each other? Or will there be a gross anomaly if there's a problem?

Thanks again!

 

Any circulating current generates waste heat so you want a low value, perhaps no more than a percent or two of the rated current.

 

I'm still not clear on how to distinguish a circulating current from the usual current. My best guess right now- with no load there is no current other than a circulating current. Is that right?
Ted

 

...and a buttload of current, since it's just like shorting them.

I've connected secondaries backwards before. I got 0V, 0A. Well, actually I didn't measure the amps, but these were 200A rated secondaries I was using to melt little bits of steel wire. When I had the secondaries connected properly and I was melting my wires, the transformer wires got super hot. When I had them backwards there was no heat, hence I assume no current.

 

I guess that must be right - equal emf on both sides. Like two batteries in parallel.

 

Ive just done some reading and what I read agrees with you wayneh. It seems its impossible to connect secondaries in parallel out of phase without nuclear meltdown occurring. So I guess my memory failed me, and what I actually did was in series. I apologize for the misinformation.

 

I think we got thrown off by the "out of phase" idea. How can they be out of phase when they share the same core? The choices are series, parallel and antiparallel. The latter is like taking the series leads and shorting them, which sounds bad. I guess you could call it out of phase.

 

in 3 phase transformers, all 3 phases share the same core too. Out of phase seems alright terminology to me, 180 degrees out of phase.