Is 24 AWG too wimpy to sustain a 300mA current in a tightly wound transformer even for a couple minutes? Perhaps I miscalculated the current. Also I suspect it is aluminum not copper. I understand it is not even close to the recommended circular mills per amp (anything from 500 to 1000). But it is getting way too hot in just a couple seconds.

 

In normal conductor scenarios 24awg is good for 2amps, I would not have though that it would have heated up in such a short time, even though a Txfmr secondary?
Aluminum would be slightly less ampacity.
But I think there is something else at play, like a shorted turn maybe.
Max.

 

There is no secondary yet. I am rewinding an old transformer using old wire. It appears OK not damaged or anything. So this is a no load current in the primary. It is way to high for the magnetizing current but before I add extra turns, I am trying to understand what is going on. I just didn't expect it to get so hot at such a relatively low current. Unfortunately I don't have a multi meter that can measure >200mA AC so I measured voltage drop across it and from that I determined the impedance to be about 400 ohm. That, along with the copper loss of about 10 ohm, corresponds to the inductance of about 1.2H. When I calculate a theoretical inductance (air core u=1) and then divide the two numbers I get permeability of about 600 which seems very low. Perhaps it is close to saturation. Still doesn't explain high temp.

 

How about estimating the length of the wound up wire, then measure DC resistance, and see if is within a reasonable range for that gauge at that length?

 

?? Need some more details, what is the application? what is the primary voltage? what secondary? What is the VA?
If energizing just the primary with no secondary load, the current should be minimal.
Max.

 

But it is getting way too hot in just a couple seconds.

How sure are you of the current? Measure it!

 

How about estimating the length of the wound up wire, then measure DC resistance, and see if is within a reasonable range for that gauge at that length?

Yes, I measured it. The core size is about 3cm x 3 cm. 400 turns, 60 turns per layer, 7 layers. So the total length is about 45m. Resistance should be about 6 ohm (for aluminum). It measures about 9 ohm but I think it is not very accurate. So sounds reasonable.

 

?? Need some more details, what is the application? what is the primary voltage? what secondary? What is the VA?
If energizing just the primary with no secondary load, the current should be minimal.
Max.

I am doing some experiments with power MOSFETs and I need a quick and dirty power supply of about 10V at about 5A. Shouldn't sag too much, so preferably minimal regulation. The input voltage is 120V. The bobbin size is 3x3 cm. The core A is about 3x4 cm. I don't know the Bsat, so I roughly calculated about 500 turns for the primary. I only wound 400 turns just to see if I could get away with that (trying to minimize copper loss). . Just now I added 100 turns and now it is working fine, not heating at all. But now my concern is, I can't probably get 5A, as it will result in about 400-500mA in the primary and it was not doing too well at 300mA....

 

How sure are you of the current? Measure it!

Unfortunately, I have three different meters, and I just realized that none of them is suitable for measuring AC in the 300-1000mA range! Many basic multimeters can only go as high as 200mA fused. I also have a clamp meter, but the current is too low to be measured accurately. So I connect a resistor in series and measure voltage drop. From that I calculate impedance and current. I just ordered another multi meter (sigh). Instead of having a boat load of cheap ones, I should have bought a good one, but I didn't really need to measure AC current in that range before.

 

I added a secondary winding (about 50 turns of 18 AWG, but I think it is an overkill).
Pretty decent low regulation I think. Take a look please, only a few sec

Never done this before and it is so much fun. Once I get some more resistors, I will try to decrease resistance a bit to see how hot it will get.
I also discovered this cool thing called spice (or pspice or whatever it is called). So I no longer need to do tedious AC calculations with phase shifts etc. I just go here:
http://www.ngspice.com/index.php. No need to draw schematics, just enter parameters directly. The output graph matches exactly my measurements!
I bet I can reduce L1 copper loss to about 3-4 ohm and this will further improve regulation.

*trans

vin 1 0 sin(0 120 60)
Rl1 1 2 8
L1 2 0 1.4H
L2 3 4 12mH
Rl2 4 0 .03
K L1 L2 0.999
Vi 3 5 AC 0
R1 5 0 1

.TRAN 1ms 2000ms
.end