Valve Shunt Regulator Design Walk-Through

Dave the bass · Post by **Dave the bass** » Wed Jan 09, 2013 2:11 pm

pre65 wrote:
I'm trying to get it working by tonight.

Shouldn't you be soldering and constructing instead of going onto DIY HiFi forums on the interwebz? :}

DTB

pre65 · Post by **pre65** » Wed Jan 09, 2013 2:13 pm

Indeed I should, but only having two bwain cells, they get tired very quickly and need frequent rest periods.

thomas · Post by **thomas** » Wed Jan 09, 2013 3:10 pm

Kill 'em!

Ahh it shouldn't do that?!

What size is your dropper resistor before the shunt and what current are you using and shunting? At a guess it might be 30v high, but only for a few seconds? I'd build it and worry later....

Mike H · Post by **Mike H** » Wed Jan 09, 2013 3:47 pm

Don't mention Zeners, don't mention Zeners, don't mention Zeners, don't mention Zeners, don't mention Zeners, don't mention Zeners, don't mention Zeners, don't mention Zeners,

DOH!!!

Dave the bass · Post by **Dave the bass** » Wed Jan 09, 2013 5:45 pm

thomas wrote:
Kill 'em!
Ahh it shouldn't do that?!

Doesn't matter if it did.... they're not my chokes arf arf arf arf!

I agree with what you say about time/over voltage situation MrT, I'd do the same... build it and power up carefully, DVM in hand and monitor current and HT plus constant observationisms.

DTB

Paul Barker · Post by **Paul Barker** » Wed Jan 09, 2013 9:15 pm

Valves don't tend to have a problem with this but capacitors need to be of sufficient peak voltage rating to handle it or you get that noise you used to get when you put caps in a gun and played cowboys and indians. A sorot of loud popping sound at start up. Which even the thickest person reading this will appreciate is best avoided.

So you select your components accordingly. But the valves are the least of your worries.

Mike H · Post by **Mike H** » Thu Jan 10, 2013 8:16 pm

Just FYI, I saw his diagram today and it seems the VR may not be getting enough oomf to start it up properly, unless the preset is turned full up. Or down. Or something. Looks like VR is not getting enough current so I suggested an extra resistor between it and the HT input (before the series dropper resistor, that is) to make up the difference.

Based on me seeing similar in another series reg design that uses a VR as the V reference in series with the cathode of the control valve. Or could have been a neon.... especially since in that case the control valve is passing very small current (as it's only the controlling amplifier for the series pass valve).

Hope that makes sense

Post by **Nick** » Thu Jan 10, 2013 8:41 pm

Yep, VR tubes have a minimum strike voltage and a minimum sustain current. The VR150 octal type needs 5ma or so. Some of the 9 pin voltage reference tubes will work at a lower current as they are designed to be a reference not a regulator.

IslandPink · Post by **IslandPink** » Tue Jan 22, 2013 10:19 pm

Can I get a bit of advice from the brainy people on here ?
Jono Noble has been effusive in his praise of his recently completed 'T-Rex' amp as per Dick Olsher's site :
http://www.enjoythemusic.com/magazine/e ... 5/trex.htm
Now, the Bench/Broskie power supply is of interest to me . It appears to provide two channels of 300B & drivers, but only uses about a 50mA shunt in one KT88 .
http://www.enjoythemusic.com/magazine/e ... x/fig2.gif

Now, it would be a rather nice thing to build, with potential application to my forthcoming amp, also for even bigger valves than the 300B , as it seems more efficient than a full-on shunt reg design.

Couple of questions , which I couldn't immediately answer by reading the Broskie blurb alongside :
(i) Is the KT88 sufficiently low impedance to provide the main signal return path for most audio frequencies ( & not the electrolytic cap ) ?
(ii) what happens when the 300B's demand rises or falls beyond the 50mA shunt current in the KT88 ?

Waiting with bated breath ....

Post by **Nick** » Tue Jan 22, 2013 10:26 pm

1) best to simulate it and see, but I suspect that C6 is passing most current about 100Hz, but thats ok.

IslandPink · Post by **IslandPink** » Tue Jan 22, 2013 10:42 pm

"about 100Hz" or "above" ?

Post by **Nick** » Wed Jan 23, 2013 12:18 am

I stand corrected, above 2k. Quick sim seems to show that the green line (cap current) passes the blue line (kt99 current) just before 2kHz

Mike H · Post by **Mike H** » Wed Jan 23, 2013 1:22 pm

Very interesting!

IslandPink · Post by **IslandPink** » Wed Jan 23, 2013 6:23 pm

Very interesting, and a very nice piece of work there - thanks a lot Nick .
That suggests that the shunt must have a pretty low impedance , say 1-2 ohms, through most of the audio band ?
Does the final cap have a realistic ESR - what is this ?

I'm encouraged - and it's interesting that Jono is really enthusiastic about the overall transparency and the midrange tone - saying it beats all his previous SE efforts with 2A3, 45 etc .

Did you use KT99 in the sim ?
I'm wondering what I'd use if I built it.
What options are there ( not too expensive mind ) that are like KT88 or KT99 ? - I can think of 6550 , any other Russian ones ?

IslandPink · Post by **IslandPink** » Thu Jan 24, 2013 8:34 pm

IslandPink wrote:Does the final cap have a realistic ESR - what is this ?

Ping Nick - just this one piece of info will do ... tells me a lot about how low the Zout of the KT99 is .

I also notice Paul's point from earlier about ideally removing the final cap , but even if it could be just reduced, to extend the band over which the shunt valve dominates, then that would be OK .

audio-talk

Valve Shunt Regulator Design Walk-Through

#46

#47

#48

#49

#50

#51

#52

#53

#54 T-Rex

#55

#56

#57

#58

#59

#60