Maximum copper losses in power transformer

[Elerion]

After making a couple of test transformers, I'm designing my first power transformer for a real tube amplifier.

As I only have EI96 laminations, and 0.25mm, 0.4mm and 1.4mm wire, I have to use 0.4mm for the primary. This follows 4 amps/mm^2 criteria. Quite high, it seems (I tend to aim for 3 amps/mm^2).
The rest for the windings are more comfortable at 2.5 amps/mm^2.

Will this be acceptable?

Calculations throw about 7 W of copper losses, and almost 6 W of iron/core losses. The transformer is 129 watts.

It seems reasonable. Maybe will get a little bit hot.
But, what do more experienced people think about it?

Thank you!

 

[kevinkr]

I suspect it will get quite warm in normal operation which would argue for using insulation and wire rated for 130C (class B).

The core seems adequately sized for the proposed power level. I checked here:

Using core VA ratings to determine transformer dimensions Marque Magnetics Yeah I know link address is silly.. lol but the article at the end of the link is real enough.

Tony T may have more useful input.

 

[rayma]

Calculations throw about 7 W of copper losses, and almost 6 W of iron/core losses.
The transformer is 129 watts.

The core loss is about equal to the copper loss, and the total loss is about 10%.
Sounds ok.

 

[infinia]

This follows 4 amps/mm^2 criteria. Quite high, it seems (I tend to aim for 3 amps/mm^2).
The rest for the windings are more comfortable at 2.5 amps/mm^2.

1 mm^2 ~2000 circular mils
or 500 CM per Ampere is the correct magnetics designers baseline ` 30 to 40 degrees heat rise. It's perfectly fine to lower the current for the same copper but your overall build will be bigger not only for much lower I^2*R loss BUT the additional benefit of more outside surface area for cooling. Depending on your POV the glass will be filled at the half point. Most engineers will like to serve it in a smaller container "the glass is obviously too big" ( interpretation 'if your build allows that much copper probably could squeeze it down into the next smaller core size.)

 

[Elerion]

Thanks for your opinions. Very useful.

It's perfectly fine to lower the current for the same copper

I'm not sure I understud what you're trying to tell me.
You mean using a lower gauge wire (thicker)? (so the amp/mm^2 lowers, and reduces copper losses?).
How could I use a smaller core if using more copper?

 

[infinia]

it's all good
Your secondary has thicker gauge than the design guideline suggest.
Should run cool depending on loading duty cycle and high line voltages. For audio the crest factor says that duty cycle is very low.

How could I use a smaller core if using more copper

Easy by using smaller gauge as the guideline suggests. Since you choose to use doubly large gauge on the secondary's and everything still fits, then a smaller core might have been possible.
Note the 500 CM per ampere is for a 1st guestimate on the design, duty cycle , V regulation and air flow considerations have to be balanced as well.