I think the amplifier will work fine, but in a first time, without any offset automatic compensation.
As designed in your diagram, the loop gain at very low frequencies is extremely high, several millions, and will certainly cause important oscillation at subsonic frequencies. And problems occurs when the correction is feed at the input (depending on the impedance of the source) it's better to feed the correction in the feedback circuitry, or in driver stages.
You have to compute or experiment this carefully, in order to not substitute a tolerable inconvenience for a severe problem : with first choice components, and correct adjustements, Hiraga 20 W needs no offset compensation (in my opinion).
Regards, P.Lacombe
As designed in your diagram, the loop gain at very low frequencies is extremely high, several millions, and will certainly cause important oscillation at subsonic frequencies. And problems occurs when the correction is feed at the input (depending on the impedance of the source) it's better to feed the correction in the feedback circuitry, or in driver stages.
You have to compute or experiment this carefully, in order to not substitute a tolerable inconvenience for a severe problem : with first choice components, and correct adjustements, Hiraga 20 W needs no offset compensation (in my opinion).
Regards, P.Lacombe
Hi,
I do not want to talk you down, but I tried cascodes for some years, and while giving more treble clarity, they smeared the bass and sometimes even the midrange. I fooled myself for a long time thinking that cascodes improve the sound.
Then, the zener at 13Volt and the resistor at 20k in the emitters of the input transistors gives bad common mode behaviour, try 20 Volts and 39 to 47k, should be better. This was corrected in later schematics of the 30W super, published by L'Audophile.
BTW, the 2SC1775/2SA872 sound best at 600uAmps idling.
If you want to improve the 30W Super Hiraga, why not try this:
- feeding Q1,2 by current source
alternatively (this is heavily butchering the original circuit):
- omitting Q1,2, instead feeding correcting currents into the emitters of Q3,4 from the opposite rails
regards,
Hartmut
I do not want to talk you down, but I tried cascodes for some years, and while giving more treble clarity, they smeared the bass and sometimes even the midrange. I fooled myself for a long time thinking that cascodes improve the sound.
Then, the zener at 13Volt and the resistor at 20k in the emitters of the input transistors gives bad common mode behaviour, try 20 Volts and 39 to 47k, should be better. This was corrected in later schematics of the 30W super, published by L'Audophile.
BTW, the 2SC1775/2SA872 sound best at 600uAmps idling.
If you want to improve the 30W Super Hiraga, why not try this:
- feeding Q1,2 by current source
alternatively (this is heavily butchering the original circuit):
- omitting Q1,2, instead feeding correcting currents into the emitters of Q3,4 from the opposite rails
regards,
Hartmut
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.