I referred to 1N4148 as a fast example for checking operation is the same because it was first to grab in simulation. That's cool you tried a diode helps in shorts indeed. I can recommend a Schottky SB140 (40V VRMM like in the one you used) instead of the Zener from now on. 1A rated should withstand the C2 discharge. Good one, thanks.
*About the desoldering problems you described before: I have a cheap desoldering pump in the tool box that does almost a decent job (not like a desoldering station but helpful at times). I got it for something like thirty Euros once on Ali or Ebay, can't remember. Its black and a bit slimmer but its having the same metal parts and same action of the blue one in this video: https://www.youtube.com/watch?v=UhyTpcBY710 or in this one https://www.youtube.com/watch?v=pCECO5GuPkU
Just in case it may help you next time you need desoldering something. Although I would recommend much better solutions like a Hakko gun I also know many DIYers that don't like to invest so much if for few occasions now and then.
Just in case it may help you next time you need desoldering something. Although I would recommend much better solutions like a Hakko gun I also know many DIYers that don't like to invest so much if for few occasions now and then.
Thanks a lot! I am still with the mechanical solder sucker, and was never thinking of such a device... I am just an occasional hobbyist, probably that's why! Definitely on my wish/shopping list 🙂
Yep not the best and not the sturdiest but its something to have than nothing. Proper professional tools are more like this: https://www.youtube.com/watch?v=K2UWxi9V00o
Thermal drift solved
Hi Salas,
Just finished the last tests round to get rid of the thermal drift.
Replacing the FET by a resistor had a too bad regulation Vs mains voltage variations as well as setting a very low VGS (20mV) but at a lesser extend.
Finally the best result was obtained swapping the FET for a different IDSS one, rule is quite straight forward looking at output voltage after startup : Negative variation -> decrease IDSS / Positive Variation-> increase IDSS. Current setting is of second order as I only allow variation from 1.8 to 2.8mA to keep dissipation in R9-R10 OK with 2W resistors.
Voltage at startup reach 550V, raise at 551v at 2 minutes and go back at 550v after 10'. This is with the FET installed BELOW the board. Same FET on top of the board next to Q5 showed same behavior but with a larger 3V sweep. Sensitivity to temp variations while blowing on the fet with a tube has almost no effect.
This is the end of my 600V mod, no arcing, stable. Time to do the same with the 350v positive supply and install them in the amp 😀
Thank you Salas for the nice designs you share.
Laurent
Hi Salas,
Just finished the last tests round to get rid of the thermal drift.
Replacing the FET by a resistor had a too bad regulation Vs mains voltage variations as well as setting a very low VGS (20mV) but at a lesser extend.
Finally the best result was obtained swapping the FET for a different IDSS one, rule is quite straight forward looking at output voltage after startup : Negative variation -> decrease IDSS / Positive Variation-> increase IDSS. Current setting is of second order as I only allow variation from 1.8 to 2.8mA to keep dissipation in R9-R10 OK with 2W resistors.
Voltage at startup reach 550V, raise at 551v at 2 minutes and go back at 550v after 10'. This is with the FET installed BELOW the board. Same FET on top of the board next to Q5 showed same behavior but with a larger 3V sweep. Sensitivity to temp variations while blowing on the fet with a tube has almost no effect.
This is the end of my 600V mod, no arcing, stable. Time to do the same with the 350v positive supply and install them in the amp 😀
Thank you Salas for the nice designs you share.
Laurent
That's 0.18% drift before thermally settling in your 550V example. Not bad 😀 for such a simple & high voltage Vref when you are determined. Congrats & happy listening 🙂
Hi,
I have two SSHV2 for each channel sourced by CLC filter with output 350V DC. To have better (more responsive) regulation yesterday I replaced:
R9, R10 from 68K to 75k,
Q6 2Sk117GR with IDSS 3,5mA to 4.0mA
and replaced Q4, Q5 to new KSA1381
Heatsinks Q4, Q5 one of SSHV2 heats up highly (they are realy hot) whilie heatsinks of second SSHV2 are moderately warm.
How could I reduce Q4, Q5 heat one of my SSHV2? What should I check?
Regards,
bern
I have two SSHV2 for each channel sourced by CLC filter with output 350V DC. To have better (more responsive) regulation yesterday I replaced:
R9, R10 from 68K to 75k,
Q6 2Sk117GR with IDSS 3,5mA to 4.0mA
and replaced Q4, Q5 to new KSA1381
Heatsinks Q4, Q5 one of SSHV2 heats up highly (they are realy hot) whilie heatsinks of second SSHV2 are moderately warm.
How could I reduce Q4, Q5 heat one of my SSHV2? What should I check?
Regards,
bern
Yes Q4, Q5 they are new from Mouser. I replaced all four KSA1381(both SSHV2) to the same series. It seems that during testing (with dummy resistor) all 4 were moderately warm. After momentary connected to target system I returned to dummy resistor and now Q4, Q5 of one SSHV2 are hot. Unfortunately I do not have a scope.
No. Both SSHV2s have exactly the same resistor values and other parts.
I've checked soldering of suspected SSHV2 one more time and all seems to be ok
Having a friend with oscope to check the rails look same on both? Having it on Kelvin output mode? If yes try regular two wire mode also and if something changes to the better its most probably oscillation. See http://www.diyaudio.com/forums/power-supplies/192625-sslv1-1-builds-fairy-tales-346.html#post4595171
Hi bern,
Didn't made it for the 350V yet but for the 550V, Q6 IDSS is at 5.38mA measured with 5V VDS. DS current @1.8mA, 150K for me, so around 100K resistors for you.
Tempco is quite sensitive to IDSS, 5.78mA IDSS showed an obviously negative Tempco.
Laurent
Last edited:
bern,
I forgot to tell you that you can certainly keep your original 68k resistors. Running a little higher current (2.57mA) you may need a little lower IDSS than I needed at 1.8mA.
Laurent
I forgot to tell you that you can certainly keep your original 68k resistors. Running a little higher current (2.57mA) you may need a little lower IDSS than I needed at 1.8mA.
Laurent
Thanks Salas,
I replaced Q3 IRF840 and my heat issue with Q4, 5 disappeared🙂
Laurent I finished my SSHV2s with R9, 10=75k resistors and Q6 2SK117GR IDSS=4.3mA 🙂
Is it possible to use a SSHV2 directly with a CCS-loaded tube or do I need some kind of isolation, lets say a resistor and a capacitor to ensure stable workings for both units?