@grataku
Hi, V1.2 is valid. No other iterations since.
See post #3,273 for a contributed pcb approach. Naturally capacitors size & tube sockets are not much flexible for choice when going with a pcb.
First page PSU architecture is valid. Shunt reg is subjectively better but trickier to set and hotter than Maida.
Good luck with the Valve Itch and let us know your progress.
Hi, V1.2 is valid. No other iterations since.
See post #3,273 for a contributed pcb approach. Naturally capacitors size & tube sockets are not much flexible for choice when going with a pcb.
First page PSU architecture is valid. Shunt reg is subjectively better but trickier to set and hotter than Maida.
Good luck with the Valve Itch and let us know your progress.
Hi, the SSHV2 shunt reg needs to set its main CCS for each app, gets hotter i.e. needs bigger sink and air circulation.More space. Its Jfet Vref can drift if the ambient delta swings much. Also generally more difficult to troubleshoot if damaged having both a series CCS and a parallel voltage section. A "super reg" architecture as some old techs used to say. Subjectively its different than series regs, you may like it more or not. Many diyers like it.
The Maida is series type, it basically uses an LM317 and a Mosfet shield, has no CCS. Shunt discrete or series integrated you decide which way you wanna go.
The Maida is series type, it basically uses an LM317 and a Mosfet shield, has no CCS. Shunt discrete or series integrated you decide which way you wanna go.
The schematic used back then was from an Italian blogger's link long gone down. But I think todd's Maida is exactly the same: https://www.diyaudio.com/community/attachments/todds-maida-mosfet4-pdf.208001/
Pete Millett also made a Maida. Google it up. Neurochrome's should be the most sophisticated Maida. But hey, they all are Maida.
Maybe I should know this but does AC heater on normal, non DHT tubes, increase the noise?
So, we could use 220K for R6 to be in the 280-300V output, right?
Is it also good to have the the LM317 on the heatsink? To cite Millett:
Finally, what heatsink size are we talking here?
Peppe, I looked at the neurochrome site Maida and I am going with approx the same size or maybe 1.25x that. I would definitely heatsink both active devices.
By the way, that RC between output and ground is for helping LM317's stability. A cheaper 47uF electrolytic capacitor alone could do that trick instead of the 0.47uF 630V Wima MKP and the 4.7 Ohm resistor to give it artificial ESR. Maybe the original link's guy didn't have a high voltage electrolytic handy. Or he wanted it to sound different (kinda speedier/cooler), who knows.
I like Pete Millet version of the MAIDA, too.
http://www.pmillett.com/HV_reg.html
Was comparing the parts, Pete uses 317HV type, and a different series mosfet part. The suppressor parts p6ke39a for the LM317 seem like a nice touch, too.
Seems to me the fqpf2n70 series device has better SOA that IRF840 and higher max Vds but higher RDS on. I am not smart enough to figure out what the actual RDS is for the device at its actual operating point and how that affects dissipation.
Ale Moglia recommends the STF3LN80K5 3ohm RDSon. He says that the lower input capacitance is better, even in a series reg. I have lots of these.
http://www.pmillett.com/HV_reg.html
Was comparing the parts, Pete uses 317HV type, and a different series mosfet part. The suppressor parts p6ke39a for the LM317 seem like a nice touch, too.
Seems to me the fqpf2n70 series device has better SOA that IRF840 and higher max Vds but higher RDS on. I am not smart enough to figure out what the actual RDS is for the device at its actual operating point and how that affects dissipation.
Ale Moglia recommends the STF3LN80K5 3ohm RDSon. He says that the lower input capacitance is better, even in a series reg. I have lots of these.