Hiraga 20W Class A, the restart

Hello,

I am restarting a Hiraga amplifier, 20W and A class.
Original one.
My problem is, now, that the output offset voltage is not at all stable.
Here are a few measurements:

We can see that voltages are différents with and without the cover.
To set-up the offset voltage, you have to remove the cover.
When you insert the cover back, the voltages are not any more the same !

At one time, I opened a window near the amplifier, you can see the result.


I'm looking for any idea to stabilize these voltages.

What do think of a DC servo compensation ?

Is it a good idea ?
Is it possible ?
Did somebody already tried it ?
Any schematic ?

Thank U in advance for your help.
Regards.
Alain.


Merci par avance pour votre aide.
 

Attachments

  • Capture_Hiraga_Classe A_20W_variation tension de décalage en sortie_eng.PNG
    Capture_Hiraga_Classe A_20W_variation tension de décalage en sortie_eng.PNG
    17.9 KB · Views: 554
Hello,
I did have a 30 watt Le Classe A. If i touched one of the driver transistors ( And make its temperature go up) the dc output voltage also did change. What i did do is mount both drivers ( isolated!!! ) to a thick piece of aluminium so they will have the same temperature.
Dont know if this will solve YOUR problem.
greetings, Eduard
 
Hello,
What do you mean by matching the input ?
Do you mean the input transistors ?

I don't know if they are matched or not.

The amplifier comes from "La maison de l'audiophile" and it shoulb be a first release.

Yes, the input transistors should be matched for hFE. I did this for the 30W version and the output DC is pretty stable. Although I did not try with the window open
icon10.gif
!
 
Hello,
I did have a 30 watt Le Classe A. If i touched one of the driver transistors ( And make its temperature go up) the dc output voltage also did change. What i did do is mount both drivers ( isolated!!! ) to a thick piece of aluminium so they will have the same temperature.
Dont know if this will solve YOUR problem.
greetings, Eduard

Yes, with this topology, the drivers need a separate heatsink, away from the output semis. I know about this issue having worked with this:

http://www.diyaudio.com/forums/solid-state/270239-i-call-infinitron.html
 
Last edited:
Hello,
But they get reasonable hot and keeping them at the same temperature did work in my 30 watt le classe A.
Another tweak was changing the two 1k8 1/2 watt resistor with 1 or 2 watt. I did use all the original parts from the shop in Paris but these 1k8 ( close to the drivers on the board) i did change into Allen Bradley which did give a nice improvement.
Greetings, Eduard
 
Hello,

This resistor is supposed to dissipate about 0.2W.
It's not my prime concern.

Today news:

- capacitor of 220pF in parallel with the 33k at the input, measured at 218 and 222pF.
- Electrolytic Nichicon MUSE BP 10µF, in place of 0,1µF, for the 13V zeners.
Noise reduction, across the zeners, from 4mVpp HF to less than 1mVpp without HF.

But the offset voltages always vary from +/- 150mV.
I think it could be hard for the bass speaker, wich didn't have a capacitor in serie in the filter.


Peak to peak output voltage, on 8Ω: 18Vp, or 20W rms.

BW: more than 100kHz, my BF generator stop at 1MHz.

Power transistors cases temperatures = 55°C.

Andre C. Russell, who propose the "Ovation SX amplifier", kind of more sophisticated Hiraga, recommands to match transistors and zeners.
- Tom D Kite Busts Some Audio Myths! ... ier-v2-10/

More easy to say than to do !

I have found some matched BC550/560.
I think I will try, at least for the VAS, wich give the voltage gain.
The input transistors are in common collector, so no voltage gain.

Concerning the matching of the zeners, what could be their influences on the PSRR of the first stage ??

There is no Zoebel cell, is it normal ??

Il n'y a pas de cellule de Boucherot/Zoebel en sortie, Est-ce grave ???

I always think of a DC servo.
Some thoughts ??
 

damon1983

Member
2013-05-04 2:58 pm
Hello,

This resistor is supposed to dissipate about 0.2W.
It's not my prime concern.

Today news:

- capacitor of 220pF in parallel with the 33k at the input, measured at 218 and 222pF.
- Electrolytic Nichicon MUSE BP 10µF, in place of 0,1µF, for the 13V zeners.
Noise reduction, across the zeners, from 4mVpp HF to less than 1mVpp without HF.

But the offset voltages always vary from +/- 150mV.
I think it could be hard for the bass speaker, wich didn't have a capacitor in serie in the filter.


Peak to peak output voltage, on 8Ω: 18Vp, or 20W rms.

BW: more than 100kHz, my BF generator stop at 1MHz.

Power transistors cases temperatures = 55°C.

Andre C. Russell, who propose the "Ovation SX amplifier", kind of more sophisticated Hiraga, recommands to match transistors and zeners.
- Tom D Kite Busts Some Audio Myths! ... ier-v2-10/

More easy to say than to do !

I have found some matched BC550/560.
I think I will try, at least for the VAS, wich give the voltage gain.
The input transistors are in common collector, so no voltage gain.

Concerning the matching of the zeners, what could be their influences on the PSRR of the first stage ??

There is no Zoebel cell, is it normal ??

Il n'y a pas de cellule de Boucherot/Zoebel en sortie, Est-ce grave ???

I always think of a DC servo.
Some thoughts ??
It's very strange... What did BJT's you use?
And give schematic, what you've done.
 

damon1983

Member
2013-05-04 2:58 pm
Hello,
It's really fig 1, from this site.

"La maison de l'audiophile" add a 2.2µF capacitor at the input and a 33K resistor to ground at the other side of the capacitor.

The four power transistors were destroyed by the owner, collector-emettor short-circuit.
I don't know why yet !

You've done it without changes in orig schematic?
What kind of BJTs you used?