Hybrid amp

[Wavebourn]

0.3 mm gap in ETD-29 core
Primary 13+12 turns
Secondaries 6 turns each
Power winding MC 5 turns
Wire 0.51 mm except 3,4 that is 2 wires 0.7

 

[kees52]

Thanks

 

[Wavebourn]

You are welcome.

By the way, here is my result with 60H current sensors:

http://www.diyaudio.com/forums/tubes-valves/307117-copper-belly-belt-se-opt.html#post5065559

 

[Wavebourn]

Here you go:

 

[kees52]

Hi wavebourn.,

I see that the coils are different in henrys, but for 1 to 1 not such a problem but make a differential signal I need to pick them close.
Do you now where I can get this core? I think it is a big one, needed for the inverter welder LVI coil.
After that I go on with the amps, mother is well again and so I need to catch up.

regards
kees

 

[Wavebourn]

As I said before, it does not matter that cores are different. When I stack 2 cores together, I am getting half of an average inductance at both sides. They all have the same 1,000 turns, but slightly differently wound cores.

No, I have no idea where to get cores for a welder. But yesterday I saw cheap switched mode welders in Harborfreight store. High frequency ignition on a distance and electroinc control. Nice!

 

[kees52]

I have received these cores, and did try, whoever mine inductiemeter do not measure them I did even set to 20 Henry but also 20 mH did not measure, but there is a coil because I did get signal from it.

I have done some windings short them and measure, it distorts heavely on 4 Khz, but works oke in the 20 to 800 Hz range, these cores are not oke I think, but I get some more like you did mention however these are not here yet.

Did you als use windings this way and short them? or just one winding? I get also loss, more then half less signal on output.

Strange with this coils is that mine inductiemeter who has 20 mH, 200mH, 2 Henry, 20 Henry settings but do NOT measure them, it just do not let see a value this means she are very high inductance like yours in previous post, mine meter can not see.

regards

 

[Wavebourn]

I use a single "volume" turn, all around.

Here is the RF version

 

[kees52]

I use a single "volume" turn, all around.

Here is the RF version

I had seen that a single turn did better them more turns, I think you use a "unrolled" litze wire, used as one turn but wires are all over the core.

I have to say, I use 3 stacked coils, test was 2 because I need a balanced output, I did measure that 20 hz is easely doen, has a resonant on 8 Khz (distorted) and go to 20 Khz.

I do wait for the others, these are the blue ones you have, but less easely to set on board, Maybe I go wind a core transformer with a high permeability who is used for this, I have ordered some cores, if this succeed I can more easely implement that this stacked cores.

Your picture is not readable.

regards

 

[Wavebourn]

Yes, due to peak on 30 KHz I had to load on a Zobel that I adjusted.

Picture is readable:
rod
nut
lid
body
winding 1
toroid core
static shield
body
winding 2
insulation tube
lid
nut

 

[kees52]

Picture do not open, just spinning wheel from diyaudio.
Did you zobel the peak only? because in audio area fase can get affected.

 

[Wavebourn]

I adjusted it for flat frequency response at 20 kHz.

 

[kees52]

Have you test fase behavior? because when it is bad behavior you get distortions, special oneven ones, or IM..
For me I need three stacked cores for getting a balanced output, so I go try when receiving high perma cores to wind three
wires, I have here enough thin copper. I now I have to be aware of dc, but because the input is J-fet this is not a problem.
Make a house of thin metal to prevent fields get in.

regards

 

[Wavebourn]

Balanced, do you mean bi-polar in respect to the middle tap?

No, I did not test phase shifts; obviously they present; but I see no connection with IM distortions that are the result of non-linearities.

Yes, I shield them by copper foil with adhesive conducting layer, then cover by ABS plastic, and hide in permalloy box.

 

[kees52]

This way I do mean, need 3 cores, but get good balance is troublesome that way, IM get in the amp, special when fase is not oke.

I have ordered some high permability cores, whitout offset voltage and Jfet inputs it do well, for tubes depent on how it is build, dc on such core and we have trouble, putting a slit in it works but then permability is lower.

When winding three wires in the same time capacitance is better distributed over the windings, it needs just time and not much beer or wodka to get job done, for signal transformer this can work fine, for bigger ones not..

 

[kees52]

Maybe you do it this way? it do also work with two cores.

regards

 

[kees52]

I have the transformers, now I go wind them, and use a copper sheet between the layers and on earth for capacitance lowering.

I built first a simple winding machine.

How wind best someone tips?.

regards

 

[Wavebourn]

For a phase splitter I would recommend to wind 3-filar winding. Minimal stray inductance, plus additional capacitive coupling.

 

[kees52]

Do this give not capacitance problems? did read that winding it in layers do work fine, the seller says wind first the primairy, then a copper foil against capacitance and wind the secondary bifilair.

I get 45.3 milihenry with 35 windings, I need for shire a winder,, I go make this with stepper motors, a strong one so I can drive it easely for thicker wire also, and use a pic.

regards

 

[pieter t]

For a phase splitter I would recommend to wind 3-filar winding. Minimal stray inductance, plus additional capacitive coupling.

Great! 180 degrees phase rotation with maximum capacitive coupling through trifilar winding
I am following this with interest