Output Mosfets Q6:
Voltage D to S = 19.4
R17 = .47R
Voltage across R17 = .43
I = .43V / .47R I = .91A
19.4V x .91A = 17.7W
The lab supply says it's drawing ~90W, so my calculation of 17.7W at each Mosfet seems a little low.
All Mosfets have essentially the same D to S voltage and same voltage across the source resistors.
Voltage D to S = 19.4
R17 = .47R
Voltage across R17 = .43
I = .43V / .47R I = .91A
19.4V x .91A = 17.7W
The lab supply says it's drawing ~90W, so my calculation of 17.7W at each Mosfet seems a little low.
All Mosfets have essentially the same D to S voltage and same voltage across the source resistors.
Last edited:
nitpickers
under load, 18Vac secondaries will result with approx 22V5dc rails
compute from that, if you have different secondaries
non-cascoded input JFets, I'm hysterical when seeing them having more than 22V of Uds ........... and when I see them having more than 100mW of heat
totally contrary to FAB mob
compute from that, if you have different secondaries
non-cascoded input JFets, I'm hysterical when seeing them having more than 22V of Uds ........... and when I see them having more than 100mW of heat
totally contrary to FAB mob
Toshibas? They can take 32 volts, and be as happy as an uncascoded F5T. Breakdown is @ 40 volta. But MZM is a nitpicker, he will recommend you either order a different transformer, cascode it, or risk meltdown at some point. And he’s probably right as always.
or you could use one of MZM’s J boards which have JFET cascodes right off the bat
The PS is an SLB which I thought would drop more voltage. I ordered a transformer with too much output voltage. Looks like I drop back 10 and punt.
edit: I ran that PS with an M2X (Norwood) and had no problem, but I’m too much of a scaredy-cat to use with Aleph J.
What does it look like when a JFET goes poof? What happens to the rest of the amp?
edit: I ran that PS with an M2X (Norwood) and had no problem, but I’m too much of a scaredy-cat to use with Aleph J.
What does it look like when a JFET goes poof? What happens to the rest of the amp?
Last edited:
The OS just dies. Probably not healthy for the speakers, but not dangerous. I bet you would even struggle seeing it visibly.
don’t be a fraidy cat: give it a go
If you have 30 volt rails, deducted for load to approx 28V5, each device is dissipating approx 26 watts. I hope this is in a 5U and not a 4U. Otherwise I suggest you back down a little bit. Say to 0A7 per device / 0V33 drop or so. At least for starters
Add a couple of Watts for the dissipated heat across .47R resistors.
The most important part you should be worried about (with +/- 30V rails) is that R5 (use a higher dissipation rating for this resistor)... Next -> the JFETs. And then, the MOSFETs if you want to keep close to 1A current through each of them. Re: MOSFETs... you could match the pairs tightly AND you could match the .47-ohm resistors to ensure close to identical power dissipation on all 4 MOSFETs. Re: JFETs... you could put a little heatsink on them. I personally would not go that high with rails voltage...
My rails are 24-25V DC loaded, the current is 1A across each .47-ohm resistor, dissipated power 200W. No C1, so initial bias around 250mV.... settles nicely to around -10 to + 10mV, and it stays stable. 5U case. I matched the .47-ohms resistors.
My BA3, and the one 6l6 built in his build guide have 31 volts on the rails, it has run for years. If they let go, I'll put another set in.
Russellc
Are you using an Antek trafo? What's the va rating and secondary output voltage?
The arrangement of Q1A and Q1B is called a differential pair or long tailed pair.