Howdy Salas,
'put it back in tonight. MJE5731A Vbe = 0.493VDC
As you already know, so far your original choices for ccs devices sound best to me in this circuit. While in some ways it's easy to say that the later version is better, I still like the character of the original circuit and wouldn't complain if it had stopped there. Still exploring , I'll post if I find any more good ones.
That Vbe is not extra chunky to be biasing the JFET steadier so that BJT can be better in parameters its poor data sheet does not cover. At least for a 400V TO-220 we can not say no to the range extension & reliability it brings when going over 300V.
From the start the core circuits proposed are things made and evaluated and liked in real DIY systems. Not theoretical. The first is nice enough but wilder and easier to go astray. Because of several BJT's there are base currents and tempcos in the ref. Especially when it goes higher than 250V. Lower it can be behaving. The JFET ref is more straightforward and has more even Zo and stable characteristics. They are brothers anyway. I prefer the SSHV subjectively also. But I have listened to it in many different projects and I am talking statistically. Going the subjective way in one particular application someone can say the first one helps my tone here, or the SSHV helps another application, but that is individual happening and the job of the thorough inquiring DIYer to find a coherent whole for a given system. Still does not say much about electrical parameters of the circuits. Conclusion: Have fun, that is why I made them, they all certainly help, mix and match, listen a lot, watch your hands with high voltage.
From the start the core circuits proposed are things made and evaluated and liked in real DIY systems. Not theoretical. The first is nice enough but wilder and easier to go astray. Because of several BJT's there are base currents and tempcos in the ref. Especially when it goes higher than 250V. Lower it can be behaving. The JFET ref is more straightforward and has more even Zo and stable characteristics. They are brothers anyway. I prefer the SSHV subjectively also. But I have listened to it in many different projects and I am talking statistically. Going the subjective way in one particular application someone can say the first one helps my tone here, or the SSHV helps another application, but that is individual happening and the job of the thorough inquiring DIYer to find a coherent whole for a given system. Still does not say much about electrical parameters of the circuits. Conclusion: Have fun, that is why I made them, they all certainly help, mix and match, listen a lot, watch your hands with high voltage.
DOH!
The avatar was so small that I didn't recognize the H.M.'s "blue nude" painted in 1952. I always thought it was a subjet from an ancient greek pot or jar.
Touché, Monsieur Salàs!
I'd like to understand your thinking a little better. What Vbe would you like to have seen there?
Do you think the noise is low enough to substitute for a bypasscap/resistor in a DC coupled 45SET amp. I could post a concept schematic. I'm having trouble visualizing how the regulator would hold a steady voltage with the current flowing the opposite direction as normal.
Look in Wrenchone's NLD thread. http://www.diyaudio.com/forums/tubes-valves/171183-no-light-district-p-p-amplifier.html
He is the only one I know he tried the concept, read there maybe it can enlighten you further or to post and make some questions and suggestions.
He is the only one I know he tried the concept, read there maybe it can enlighten you further or to post and make some questions and suggestions.
I read thru it when you recommended it thanks. Looks like I'm treading on new ground.
One last question which regulator is lowest noise, the LVSS or the HVSS ?
There are V1, V1.2, V1.2R for low voltage and the SSHV that can go much higher. All have power Mosfets. V1 has the less BJTs in number and also low noise in type, less gain, so it owes to produce the less self noise. SSHV is few active components also but its MJE must be noisier than a BC550 that you can find in V1. I don't think that the noise level of such shunts that have been used successfully in phono and line stages or DACs can be a problem in the cathode loop of a valve output stage though. The signal levels are many Volt there.
I can't possibly guess at a system that I can't experience and also take measurements as a whole. The reg lowers its Zo with more standing current spared to it up to a point. That is one thing. Interactions and subjective preferences with an unknown to the reader valve circuit that it serves its another. Set it wherever you think its more likeable to you as long as it stays in good working order.
