Q6 NJFET is most probaby gone. 4 Ohm is too low RDS for its type, unless something almost shorted parallel to it causes that reading. So test it again out of board for RDS with DMM across its outer pins. If that will read several tens of Ohms, measure it for IDSS also.
(http://www.diyaudio.com/forums/pass-labs/228897-jfet-matching-sorting.html#post3347479)*
*A 9V battery will also do if not having 11V
(http://www.diyaudio.com/forums/pass-labs/228897-jfet-matching-sorting.html#post3347479)*
*A 9V battery will also do if not having 11V
Over specs supply
Hello Salas,
Would you mind to tell if it's possible to go up to 540VDC output with some mods or would it need deeper mods out of this project scope ?
Current drawn would be 40mA per channel, one for both if possible.
Edit : Forgot to tell it's a negative supply, the positive one is 350VDC, regulated too.
Thanks for your help
Laurent
Hello Salas,
Would you mind to tell if it's possible to go up to 540VDC output with some mods or would it need deeper mods out of this project scope ?
Current drawn would be 40mA per channel, one for both if possible.
Edit : Forgot to tell it's a negative supply, the positive one is 350VDC, regulated too.
Thanks for your help
Laurent
Last edited:
Tighter sound with more resolution due to much lower impedance than passive ripple filtering, much cheaper, less reliable than passive when mishandled.
Can use higher VDS NDMOS but there are no PDMOS, the polarities sections must be stacked to appear negative also. Wider creepage distances and cuts must be used in the layout to avoid arcing, the ref resistors must be scaled up, the BJTs must be sinked.
Thanks Salas, in other words that would need to study and route a new board.
Would IRFBC40 be able to replace IRF840 ? What about DN2540, do they need each to widstand the max voltage ? Say 600v ?
I'm not that used to transistor electronics, feel more comfortable with tubes.
Laurent
M1 (top one) withstands the HV. There are some IXYS that withstand much high voltage but slower. IRFBC40 is a 600V device which comes close to 840 other characteristics. To surely know that an HV regulation device with changes will work reliably at 540V its not an off hand prediction, it takes experimentation and testing. Small things even like how and where you put any protection like zeners may cause it to blow.
I guess you're right, such a fast and dead silent regulator board might be prone to oscillate during modification process. I may give it a try later on if and when this schematic is fully understood, which is not the case yet. Dissipation is another trade off, I might also concentrate the shunt regulation only for the stages that deserved it best.
I'll stick for the moment with my crappy series regulators.
I 've some other work to do with a few Reflector D to order and a DAC to set up. Thanks Salas for all the work.
You could try the higher VDS devices and higher Vout with an existing SSHV2 board if curious and super careful to see what happens, its just not tested for 540VDC out and alternative parts. I would leave the zeners out and use bigger output sink and good insulation in such a test. The mA setting must be conservative, just covering the full sinewave swing in mA of the load plus 20mA spare. M1 should have equal or higher Vgs (off) than M2 to allow enough CCS flow. The usual DN2540s cascode for 400V spec SSHV2 tops out at circa 110mA on the test bench. I wouldn't aspire to cover an output stage, and would surely start with a dummy resistive load, also some kind of soft starting for DCin like an RC or a variac.
Thanks for the tips, just ordered 2 of the boards. I'll keep that aside for at least a few monthes.
Need first to set up back my electronic workplace, then cook the Soekris ladder DAC that will hopefully be a serious source, and finally go further on a tube amp development started some ten years ago. I'll keep the thread posted in case of success.
Technically that would work. I used that when in an urgent need to try a new idea. I'd like to find a more elegant way for a "final" version. Thanks anyway.