I connect entire chassis to the ground including heat sinks. Of course transistors must be electrically isolated from heat sinks.
The devices are isolated. Not sure if you saw my original question, but the reason I ask is because my amp is built of a wooden frame that the heat sinks and circuit boards are attached to. Is there a reason electrically to tie both heat sinks to a central ground plane under the amp? I was planning on just using a piece of copper for this. It will be easy enough to connect the heat sinks to this point, but i'm not going to bother tapping them if it's not necessary.
In this case I would not bother to ground heat sinks. Just make sure you do not have interference noise in your amp.
Ok, that is what I thought. I'm hoping that the layout is smart enough to help avoid RFI. All of the wiring runs are short. I'm using shielded wire for the inputs, and the circuits are sandwiched between the heat sinks. Worse comes to worst I guess I'll have to attach metal panels to the wooden frame for shielding. I hope not, I really like the looks of this anachronistic beast ;-)
Thanks for your help!!
Thanks for your help!!
Ok, I've fired her up and nothing blew up. The 2.5A fuse seems fine, and my rails are +/- 25.9V where they leave the PSU (I'm using a 600VA toroid for stereo, and a rather large capacitor bank).
I tried to start adjusting the pots to set the bias, and the voltage measured across R11 and R12 doesn't seem to budge. I have the pots that come with the techdiy kit. At any rate, how much turning will I have to do before the voltage drop across R11 and R12 will start to go up? I haven't turned them tons, but don't want to go whole hog if there is a problem. I had zeroed them out completely before starting this process.
Is this normal? Does it take a substantial amount of adjustment to begin seeing voltage appear across R11 and R12?
I tried to start adjusting the pots to set the bias, and the voltage measured across R11 and R12 doesn't seem to budge. I have the pots that come with the techdiy kit. At any rate, how much turning will I have to do before the voltage drop across R11 and R12 will start to go up? I haven't turned them tons, but don't want to go whole hog if there is a problem. I had zeroed them out completely before starting this process.
Is this normal? Does it take a substantial amount of adjustment to begin seeing voltage appear across R11 and R12?
Last edited:
At the beginning I would recommend to measure the voltage at the amplifier output and start adjusting one FET. As soon as output voltage starts leaving 0V DC that FET has started conducting. After that you can start adjusting other FET as well, checking voltage across R11/R12.
After that, when both FETs are conducting, you can adjust both potentiometers to get required bias.
This is how I set both channels up in my F5.
After that, when both FETs are conducting, you can adjust both potentiometers to get required bias.
This is how I set both channels up in my F5.
Last edited:
Ok, both channels are at between .589 - .591 volts per device and dc offset is within 1mv. I'm going to let them warm up and check back in after a couple hours. So far everything is running very cool. The heatsinks are barely warm and the aluminum bar holding down the devices is a little warm, but I can leave my finger on it indefinitely.
Can't wait to hear music!!!
Can't wait to hear music!!!
35v is a safer rating. On my F2 dual positive (24vdc) PSU, i miswired the (2 x 0-18) transformer and had 37v on one of the supplies. Thankfully i switched it off before any of the magic smoke escaped. i was using 35v caps. 25v caps would have probably failed in this situation 
Better to build something that is safe and reliable. Having said that, the service manuals for the First Watt amps state that the PSU caps are 15000uf/25v. Maybe someone more experienced can explain the logic behind Mr. Pass choosing this rating for supplies with 24v rails
cheers,
mymindinside
Better to build something that is safe and reliable. Having said that, the service manuals for the First Watt amps state that the PSU caps are 15000uf/25v. Maybe someone more experienced can explain the logic behind Mr. Pass choosing this rating for supplies with 24v rails
cheers,
mymindinside
Probably because
1) he has several thousand of them, and/or
2) the rail voltage are within specifications of the caps
Of course mains voltage variations could create some issues. I would suppose using 23V rails give reasonable headroom for 25V capacitors, as the last two pictures here show.
6moons audio reviews: FirstWatt F5
1) he has several thousand of them, and/or
2) the rail voltage are within specifications of the caps
Of course mains voltage variations could create some issues. I would suppose using 23V rails give reasonable headroom for 25V capacitors, as the last two pictures here show.
6moons audio reviews: FirstWatt F5
One reason to ground the heatsinks is that ungrounded they act like a sort of floating capacitance and/or an antenna. This can in some cases raise the noise floor of the amp, inject unwanted ultrasonic (RF freqs too) signals and in the worst cases cause parasitic oscillations - especially on peaks...
Copper connections to aluminum parts are problematic, fwiw. You may want to look into the way that copper and aluminum are connected in electrical circuits (home/industrial wiring)?
Someone asked about the resistors, the power resistors?
It's a long long thread - but YES it matters!
In this amp the resistors seem to have a definite effect on the subjective quality of the sound coming out. The power supply does as well.
Someone else suggested Shottky power rectifiers - I'd suggest soft recovery type myself... although as with everything else in this amp, ymmv and you may like it better one way or the other.
I have heard a single F5 implementation go from somewhat subdued and maybe "tubby" to incredibly dynamic and open merely by changing the sorts of things mentioned just now... same semiconductors, board, heatsinks, layout, power transformer, AC mains lead, etc... So, imo there is no "F5" sound, there's your F5 implementation's sound!
_-_-bear
Copper connections to aluminum parts are problematic, fwiw. You may want to look into the way that copper and aluminum are connected in electrical circuits (home/industrial wiring)?
Someone asked about the resistors, the power resistors?
It's a long long thread - but YES it matters!
In this amp the resistors seem to have a definite effect on the subjective quality of the sound coming out. The power supply does as well.
Someone else suggested Shottky power rectifiers - I'd suggest soft recovery type myself... although as with everything else in this amp, ymmv and you may like it better one way or the other.
I have heard a single F5 implementation go from somewhat subdued and maybe "tubby" to incredibly dynamic and open merely by changing the sorts of things mentioned just now... same semiconductors, board, heatsinks, layout, power transformer, AC mains lead, etc... So, imo there is no "F5" sound, there's your F5 implementation's sound!
_-_-bear