I’m having issues getting equal bias while achieving 0 offset. I can get R11 to .56v and R 12 to .61v and achieve 0 mV offset. But then if I adjust the trimmer so that R11 = R12 at .61v then Dc offset is around 50mV.
When I put the amp together I could have swore the resistors had equal value within .01 ohm. Any advice. Is it better to have r11 and r12 equal at .61v and have the 50 mV dc offset? Or better to have different bias on R11 and r12 and O mV offset?
When I put the amp together I could have swore the resistors had equal value within .01 ohm. Any advice. Is it better to have r11 and r12 equal at .61v and have the 50 mV dc offset? Or better to have different bias on R11 and r12 and O mV offset?
50mV offset is really good.
To answer your question, it’s better to have zero (really low...) offset and slightly dissimilar bias than equal bias. This has to do with the Vgs differences between the N and the P channel Mosfets, which is normal and expected. Offset and bias are adjusted by the potentiometers, but much more then the resistors are at play - the Idss of the Jfets, the Vgs of the Mosfets, the absolute values of the source resistors, and the variations in temp coefficient of the bias thermistors.
Also, you’ll never get the offset totally zeroed and stable, tempature changes by local air currents will make things move around a bit. Next time you adjust it, blow on the jfets and watch the offset. 🙂
To answer your question, it’s better to have zero (really low...) offset and slightly dissimilar bias than equal bias. This has to do with the Vgs differences between the N and the P channel Mosfets, which is normal and expected. Offset and bias are adjusted by the potentiometers, but much more then the resistors are at play - the Idss of the Jfets, the Vgs of the Mosfets, the absolute values of the source resistors, and the variations in temp coefficient of the bias thermistors.
Also, you’ll never get the offset totally zeroed and stable, tempature changes by local air currents will make things move around a bit. Next time you adjust it, blow on the jfets and watch the offset. 🙂
Last edited:
Ok thanks for the answer. Yes it does drift around some +/- 12-15 mV.
I’m running a little hot as my heat sinks are 65c after 1 hour and the output mosfets are 85c right on the mosfets. Is this in the SOA?
I’m running a little hot as my heat sinks are 65c after 1 hour and the output mosfets are 85c right on the mosfets. Is this in the SOA?
And yes one of the harder areas of this build was getting the toshibs Jfets (fakes and idss variance)Could it be a mismatch in idss of is it the p channel mosfet issue it’s IRF that Nelson spoke of?
85C is too hot, turn it down. (or add fan) 65 on the transistors, 55 heatsinks.
This isn't what he's talking about in regards to the IRF9240.
This isn't what he's talking about in regards to the IRF9240.
I have a variac that I’ve been using to go from 24v to 20v. I honestly can’t hear a difference but it does get it down to 55 ish on the sinks. I also have a fan that I really can’t hear that zaps the heat off the sinks totally. Just reluctant to install fans. May install it not sure yet.
Where are people buying these big heats sinks?
Where are people buying these big heats sinks?
Probably a silly question, but is there any harm in omitting the leds and associated resistors? It seems to me these are purely decorative and serve no real function. I'm building v3 boards from the store..
Thanks!
Greg
Thanks!
Greg
Thanks Mr Pass. I currently have the IRF chips running. Seems I read somewhere that there was an issue with the IRF p channel mosfet and that you preferred the Fairchilds that are NLA(don’t want to misquote you so that’s why I say “seems”). Would it make any difference in sound, something one could hear? Thanks for making these designs available.
If you use the Harris IRFP P channel parts, you don't have that minor issue, and
what's a little 2nd harmonic among friends anyway?
My favorite parts are the Toshibas, but they are getting hard to find and $$.
The Fairchilds will work, and you can explore other parts at your leisure.
what's a little 2nd harmonic among friends anyway?
My favorite parts are the Toshibas, but they are getting hard to find and $$.
The Fairchilds will work, and you can explore other parts at your leisure.
> Also 2SK1056, 1057, 1058 and 2SJ160, 161, 162
1) They are not pin compatible with the Toshiba's. So are not 1:1 substitutes on existing PCBs.
2) They have about 1/3 the Yfs of the Toshiba's. But you can make up for it by using 3x matched in parallel.
Otherwise good choice. 🙂
Patrick
1) They are not pin compatible with the Toshiba's. So are not 1:1 substitutes on existing PCBs.
2) They have about 1/3 the Yfs of the Toshiba's. But you can make up for it by using 3x matched in parallel.
Otherwise good choice. 🙂
Patrick