Simplistic MosFET HV Shunt Regs

Hi Salas,

I got a monster heat sink this time around for it :)
Hope the image works.

photo1_zpsd665ab59.jpg
 
SSHV Input decoupling

Helo @all
i have one external PSU and use the SSHV2 in one extra chassis for dac and tube output stage. my cables are between psu and sshv2 are 50-60cm long. do i have to decouple the inputs of the 2 sshv2 for the b+ and what values and kinds/brands are recommended for this? does it touch the sound?

thanks a lot @all

best regards kokonut alias

Karsten
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Yes, decouple at the raw DC input connector with 0.1uF-1uF MKP high voltage capacitor. The sound can harden if EMI/RFI will pass when picked by the inductance of the long raw B+ cabling. Or even worse it can trigger some oscillation in the cascode CCS, sometimes catastrophic. Use compact size industrial types for that decoupling. Smica can also do. Also don't forget to have bleeder discharge resistors across the main PSU reservoir.
 
Hi Salas,
I remember that I read somewhere in this thread that the raw power supply design is not critical as long as it has low ripple ( 150-200 mVrms max).
I am planning to use the SSHV2 to supply for Kevin's Muscovite phono stage.
Two SSHV2's, one for each channel to supply 40mA at 300V.
My question is : do I need design the raw PSU a certain way ( in Kevin's design he has a tube regulator and recommend a choke of 5H and low DCR(under 100 ohms)).
Do I need to design the LC filter (PSU) with a low cut-off frequency and do I need to address the resonance of this LC at F0?
Thanks,
Radu
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I don't think that you need design it any special regarding the SSHV2. If you want to save space and money on chokes you can possibly make simple capacitance multipliers to chain before the regs. One member was happy with a 2nd order HV one. http://www.diyaudio.com/forums/power-supplies/192625-sslv1-1-builds-fairy-tales-289.html

"Because in HV same value capacitors as in LV are too big and expensive while the current is small vs LV we can afford to make the resistor values higher and restore comparable filtering ability when keeping relatively low voltage drop across them. Which means to drop 15V with 2k2 2k2 220u 220u @ 50mA which is usually available to spare in HV. If you can't spare more than 5V drop, make the resistors 560R. This will make it not as good but still good." from #2880 in that thread

Based on a schematic from #2834. Use MJE15032 on a tiny sink in this case.
 
Thank you so much Salas.
I will try all three options and compare the sound. I believe is more important to get good, quality caps on the SSHV2 ( C1 and C2). The phono stage should be isolated from the mains by the SSHV2.
My desire is to build the Muscovite and your Simplistic and compare. It is going to take a while, but I will get back when have results.
Cheers,
Radu
 
Andrew,
Agree with low Z.
I mostly like to play with tubes. I tried many types of supply using different components ( EI, toroid, r core main trafo, silicon and valve rectifiers,crc, clc, clclc, lc, lclc, cap multiplier, different types of caps...) and use PSUD2 to verify design. I have some experience with SE power amps and I can say that at the end of the day it is very important how you design the psu. In my experience, for tube SE amp, I like the sound of EI trafo with tv dampers rectifier, clclc design with polypropylene in oil motor run caps and low DCR chokes.
Now, in the case of preamp ( phono stage here) I assume that SSHV isolates the raw PSU from the stage and there is no issue like with power amp where the singal "modulates" the power supply. This a good reason to have a very low impedance psu.
Cheers,
Radu
 
Most amplifiers modulate the current from the PSU.

The very few topologies that do not modulate the current are even more rarely adopted.
That includes Valve/tube as well as solid state and includes pre-amps, RIAA amps and power amps.

One of the biggest differences between ClassAB amps and all the other categories is the amount of current modulation.

Take an opamp (usually ClassAB output stage), with a 4mA quiescent current. 10Vpk into 2k load is 5mApk of output current that current has to come from the power rails, each rail supplying on alternate halves of the waveform. half a mA could be the output bias the other 3.5mA is front end stage quiescent. The rail current will vary from ~2mA to ~7mA alternately on each power rail. Opamps were invented and implemented in Valve/Tube days before transistors were made.
 
Last edited:
So, from from I understand reading your response is that I should treat the design of the raw PSU with utmost care.( not to rely on SSHV2 to clean up the mess).
In my case the output stage of the phono stage is delivering 3Vpp into 100k load which is 0.18mApp on the power rails (single ended).