How did I wire an OT for a single ended power transistor.

[mac]

I just wired my first output transformer. It looked challenging and for patient people, but after all it was not difficult.
After reading some posts and searched the web I decided to rewire a transformer I had. It was not a wall wart nor a PT, but it had the right size for a 5W Class A transistor amp I designed.

A. The first thing I had to figure out was the output impedance of the power transistor, a darlington pair made of a BD139 driving a TIP41C. I had tested the amp with a PT I have, 220V : 47V + 9V, 50Hz.
The secondaries looked good for an OT, I posted about it some month ago,
http://www.diystompboxes.com/smfforum/index.php?topic=76108.0

I used the 47V as the primary and the 9V as the secondary and at 0.25A - 0.3A it sounded the best. No saturation and good bass response.
I searched the web "transistor output impedance" and combinations, but I got bored after the first page. So I decided to use what's left of my brain. How to estimate power transistor output impedance?
The voltage drop across the TIP41C at 0.3A was 33V , and 3V for the limiting emiter resistor and the primary resistance. That's near 110 Ohm = 33V/0.3A.
I needed an independent calculation so I used my Valve Jr data. There are 280V across the EL84 at 50mA giving 5600 Ohm, which is the VJ's OT primary impedance. Cool!

B. Then I had to compute the required inductance L that let low freqs pass through it. Since the primary impedance I had estimated was Zp = 110 ohm then L = Zp/(2*pi*f) being f = 80Hz the guitar low E string freq. That gave me a minimal inductance L = 0.22Hn.

C. How many turns do I needed for the primary. I used the solenoid eq.,
L = N*N*u*A/s
N=number of turns
u=4*pi*10E-7*4000 wb/A/m = 0.005026 wb/A/m [Newton/A/A] [kg*m/coul/coul]
A=the solenoid area, or the bobin hole area in square m
s=the lenght of one of the EI lamination loops in m, known as path lengh.
A note on the path lenght. You will find in physics books that "s" is the solenoid lenght. This is true since the magnetic field B outside the bobin is weak and divergent. But in a transformer there is an iron path outside the bobin where B can not be ignored.

My data was:
A = 0.015mE+2 = 0.000225 m*m
s = 0.1m
L = 0.22Hn

Np = 140 turns

Since the solenoid eq is not exact, and because the estimated inductance let pass only half of the bass freqs near 80Hz, I decided to use 4x the inductance I had calculated before, L4 = 0.88hn.

Np = 2*140 = 280
Doh!

C. Since the secondary impedance Zs = 8ohm, then
Ns = Np* sqrt(Zs/Zp) = 280*sqrt(8/110) = 75 turns.

D. I had 0.4mm diameter wire. Instead of counting turn by turn I wired one layer and counted 40 turns: 7 layers, 280 turns. Nice! Did the same thing for the secondary: 2 layers, 80 turns.
How did I count the turns without a magnification glass? I used a piece of a magazine cover, and passed it perpendicular to the layer. Each time the paper's edge gets between two wires it makes a click.

I turned on the stereo and began wiring. By the time David Oistrakh was playing the last compass of Mendelssohn's Violin Concert I was finishing the primary. 28 minutes or so. Not the right music to do this since you pay attention to what are you doing and not to the Stradivarius. Some chill out next time! 

E. Putting back the lamination was tricky. It was not insulated and it did not have holes for screws in the corners. I imagined that loose laminations will begin to dance when playing due to the forces induced by the coil.
Solution: I went to mother's. She was not home, cool! I went to her bedroom and "took" nail barnish... then to the kitchen... Perdon mama!!!
I painted the E's one by one as I was putting them in place. Hopefully the barnish was going to insulate and keep things together.
The last 2 E's did not want to enter but with the aid of a screw driver I made a little space in the middle and inserted them but not completely. So I went to my car and searched in my toolbox for the sledge hammer. If one thing I learnt in all those years at my construction sites is that everything can be fixed with the hammer's touch!
I ended with my hands and nails in a fancy pearl tone. But I had also stolen acetone from mother's closet!
I let it dry for a day.

F. Next day I crossed my fingers, connected the wires and turned the amp on. Set bias at 0.3A.
Creamy, creamy... and loud!!! Cool tones for a couple of fets and transistors... on the breadboard!!!
The nail barnish worked fine, solid lows, no saturation.
Honestly, I did not expect much from it.
Now I have to experiment a little more to get the most from it and then build the head.

I hope someone finds this useful.

mac

[Skruffyhound]

Nice information - Thanks