Current feedback Mosfet Circlotron

[jpatay]

When the current decreases, the amplifier switches from class “AB” to class “B”.

 

[UltimateX86]

indeed, but by selecting the components you would have less "step"

 

[jpatay]

Prepared housing for the amplifier-circlotron.
Massive radiators, this is good. And already installed protection for two amplifiers.

 

[UltimateX86]

Hi,

This amp did not have a name yet.

To stay in the circle domain, here it is:

 

[jpatay]

Hi,

The amp did not have a name yet, here it is:

(Ray of light in the dark kingdom).😉

Maybe it's better to give a name - Seb?

 

[jpatay]

Do you think it is possible to divide + 35 V(V2, V3) into Schottky diodes?

 

[UltimateX86]

I tried simulation a long time ago and it significantly increased the distortion

Maybe we should try in real

but you're right we should improve the preamp supply

 

[jpatay]

Later I will try diodes. Now I wind two windings on the transformer.

 

[UltimateX86]

(Ray of light in the dark kingdom).😉

Maybe it's better to give a name - Seb?

In fact I play with the French translation

Radius = Rayon (fr)
Ray = Rayon (fr)

and Ray sounds like a name ... Ray Charles

the image above is his signature it seems to me

 

[Joe Berry]

we should improve the preamp supply

With my circlotron amplifiers I found that adding snubbers to the main floating power supplies reduced RF noise on the supply a lot, and this made a big improvement in sound quality. I would start there. I would not add diodes as shown above, as this reduces the signal-canceling effect of summing the two supplies.

 

[jpatay]

In fact I play with the French translation

Radius = Rayon (fr)
Ray = Rayon (fr)

Sorry, the translation is incorrect. I thought it was a ray of light or a name.

 

[jpatay]

With my circlotron amplifiers I found that adding snubbers to the main floating power supplies reduced RF noise on the supply a lot, and this made a big improvement in sound quality.

An interesting solution. I have not seen this before, and therefore I liked this circuit.

 

[UltimateX86]

With my circlotron amplifiers I found that adding snubbers to the main floating power supplies reduced RF noise on the supply a lot, and this made a big improvement in sound quality. I would start there. I would not add diodes as shown above, as this reduces the signal-canceling effect of summing the two supplies.

Thx Joe, i will try snubber

Maybe we could use one diode after summing

 

[jpatay]

I have a problem. Radiator height - 8 cm.
Had to reduce pcb.🙁

 

[Joe Berry]

Maybe we could use one diode after summing

I've not tried it but it may help. If you try it, please let us know your results.

 

[UltimateX86]

jpatay in my opinion the two input point are too far apart and you shouldn't use two ground point

Could you try to integrate the diode? with maybe a strap for shunt

 

[jpatay]

jpatay in my opinion the two input point are too far apart and you shouldn't use two ground point

Could you try to integrate the diode? with maybe a strap for shunt

I will use one ground point.
Do you want me to turn on the breadboard with the diode, as in the circuit?
https://www.diyaudio.com/forums/att...-current-feedback-mosfet-circlotron-diode-png
OK.Tomorrow I will try with one diode.

 

[UltimateX86]

since you are doing a new pcb and you wanted to try, this is the opportunity. with a strap in parallel with the diode to remove it

 

[Joe Berry]

Regarding snubber design, the best way is to use a test jig like that described here. In my early designs I just used 0.1 - 0.47 uF across each transformer secondary, and even that made an improvement.

 

[jpatay]

The diode is installed.Current - 90 mA, stable. At 50 kHz there are no "steps".
What else can I measure?
The background level at the amplifier output is 5 mV (10000 uF).