DIY Power Transformer for U67 mic PSU

[Whoops]

Hi,
I'm building an U67 type microphone and I needed a power transformer for it's PSU.
A suitable transformer would have a secondary of 210VAC (for B+) and a another secondary of 18V to 20VAC for the Heater supply.
Some people use 2 different transformers when DIYing the U67 type PSU.

The PSU output voltages and consumption are aprox. 210VDC 1mA for B+ and -6,3VDC 200mA for the Heater.

I looked in my shop and I had a suitable sized transformer with 2 secondaries of 18V each.
I though about re-winding the secondaries of this transformer for the voltages I needed while maintaining the Primary winding intact which was already 230V (I'm in EU),
so I went for it.

here is the donor transformer:

 

[Whoops]

After 30 minutes in the oven at 100ºC I dismantled the transformer,
here are the lams:




Then I proceed to unwind the secondaries while counting the number of turns, this allowed me to calculate how many turns do I need to wind for the 2 new secondaries. here is a photo of my High Tech professional winding Setup:



So each of the 18V secondaries had 101 turns.
Doing the math... for a 210V secondary I needed 1178 turn
I wanted to rewind the 18V secondary also with thinner wire than the one original used for 833mA. I targeted the 18V secondary for 400mA (2 times the consumption needed for the circuit just to be safe)

 

[Whoops]

So trying to rate the transformer for 400mA on the 18V secondary and 20mA on the 210V one (consumption is 1mA, so 20mA will be more than enough)
To choose which wire to use I checked the Wire/Amperage tables, for 400mA it was 0.41mm wire and for 20mA 0.09mm.
I had stock of 0.4mm and 0.10mm, so it's close enough.



So I the winded the 2 secondaries 1178 turns of 0.1mm wire and 101 turns of 0.4mm.
After the winding I mounted all the lams, tested the transformer and it went straight to the Varnish bath,
here is the Tx draining after the varnish impregnation



and then it went to the oven for 2 hours and a half at 100ºC

 

[Whoops]

The transformer turned out pretty good,
18V secondary measures around 19V loaded which is fine.

The other secondary measures 220V loaded instead of 210V, so it's 10V above my original target.
It's no big deal for this circuit but maybe there was something that could be better in the way I calculated the number of turns needed so that this error wouldn't occur.
If anyone knows why this happened please let me know so I can learn.

Here is the photo of the final transformer,
I know it looks burned but it's not, it's just varnish. I think it was to thick I should have thinned it to make it more liquid, well, next time I will do that I'm just an amateur DIYer that is learning by doing.




Tested it in the U67 PSU and it's working great
I'm pretty happy

 

[thomasdf]

Impressive ! Great DIY spirit !

 

[alphasnk]

That's beautiful

 

[volker]

If anyone knows why this happened please let me know so I can learn.

You calculated your new number of turns from the existing winding, but that winding will only be at 18VAC when the nominated current is drawn on the secondary side. If it is lower, the voltage will be higher accordingly. But it got you in the ballpark, nothing to complain at all for a "prototype". If you would wind it more often you could simply adjust now and have it perfectly accurate. For that reason I often include +-5% taps on the primary side with one-off transformers. That means I can slightly adjust up or down, one of them will be close enough.

Also, the wire diameter you got from a table does not take into account the resulting DCR of the winding. It is only to ensure that the current density isn't too high and the winding doesn't overheat. These ready-made small transformers are usually run at very high B-fields, because that means less turns and possibly a smaller core, which both saves money for the manufacturer. If you used the same number of primary windings, or you used the original winding, you also use less copper on the secondary side and the resulting DCR is lowered. And for the HV winding you used a much higher current and thicker wire, so again less DCR and less voltage drop when it's in use.

When designing a transformer, there are several parameters which you can tune. There is always a bit experimentation necessary because of the interaction between them. I think yours turned out really well, especially considering the approach you took.

 

[Whoops]

Thank you so much for your message and insights Volker.
Now I understand why is not a straight calculation, I knew I was missing some details but overall it was worth to try out.

In this case having 220v instead of 210v is completely fine, it it was much less than 210 then it would not be good.

The transformer is performing well, I already tested it in the U67 PSU and it works great, no problem or heating.

Thanks