High voltage regulator

[sparkle]

Hi guys!

Can anyone suggest me some sch's for the high voltage power supply (220V-260V;30-50mA). It should be really good quality unit- tubes or fet's or something - doesn't matter as long it is really good. I have a schematic ideas but don't what to think of them.

Also an opinion for/ against regulators would be nice to hear. Thanks

daniel

 

[ErikdeBest]

There have been lots of threats discussing regulators - theory and practice with plenty of complete schematics. Search for ECL85 - you will find two 250VDC regulators with this tube. In Frank's preamp threat you will find a 300VDC regulator...

Do also give a look at
http://www.glass-ware.com/tubecircuits/High_Voltage_Regulator.html
This one is made with mosfet's.

Also search for "virtual battery", that will give you pretty straightforwarded regulators using mosfets and zenerdiodes.

If you had at least searched a bit before posting this thread....

 

[JPS]

Try this one.
http://cvd.scorpiorising.ca/kits.html

 

[sparkle]

I did- I have at least 15 different sch's including glassware and one or two with ECL85- I said that I don't know what to think of them and that I need suggestions- since I don't want to build them all but the really good one of them.

The sch from glassware is nice and also that ECL85 design- but I don't know how good they are- I need ultimate regulator......

regards and thanks for the reply
daniel

 

[ErikdeBest]

The quality of a regulator can only be defined through you, after you have build it and listened to it (the proof of the pudding....). You can only say: this is the best after you have compared them all.

The only way of saving efforts and money is trusting other peoples knowledge. I think that both Frank and the guys from Tubecad are thrustworth because I know they have made lots of testing to obtain the best. If I understood well, Frank actually uses the 300VDC and 250VDC in his system....knowing that would be enough for me.

Good luck.

 

[sparkle]

The quality of a regulator can only be defined through you, after you have build it and listened to it (the proof of the pudding....). You can only say: this is the best after you have compared them all.

Yes I agree with you- but the money/time factor is here significant...

The only way of saving efforts and money is trusting other peoples knowledge. I think that both Frank and the guys from Tubecad are thrustworth because I know they have made lots of testing to obtain the best. If I understood well, Frank actually uses the 300VDC and 250VDC in his system....knowing that would be enough for me.

Good luck. [/B]

Oh, I trust in the knowledge of the people of Glassware/TubeCad - they are great guys with many nice wicked ideas that bring smile on our faces with their projects. That is why I have been considering to use this projects among others- but I have only a little experience using regulators in tube devices so I don't know how it would sound - from other people experience- and what to use. I like to ask what bothers me .
Many heads many solutions.... I know that there are great guys up there on forum that know much more than me and I respect that and like to ask before I spend my money on something.

All in all I know what to do know- use either Glassware design for 220V. Tubes would not suit my needs (looking fromn the technical point of view). For sound I simply don't know.

Thanks Erik
regards
daniel

p.s. btw- how is working that reg. supply that You have made (with mosfets)- why didn't You use glassware's design?

 

[SY]

The "ultimate" from a performance point of view will have to use a bipolar or similar pass device (highest possible transconductance, lowest rbe) and a very high gain-bandwidth op amp as an error amplifier. Tubes won't cut it (transconductance too low) nor will MOSFETS (much better transconductance, but not as good as bipolars). Joe Curcio has published some designs like this. Check out his website and his preamp articles in Audio Amateur.

 

[GaryW]

Wellborne Labs and Curcio Audio Engineering both have high quality regulator boards that should meet your specs. I use CAE's in my modified PAS3 and it sounds quite good.

 

[DougL]

It should be really good quality unit-

Regulators, like other things, aren’t on size fits all.
The regulator for the first stage of a phono preamp is going to be different than a driver tube in a power amp, for example.

If you could be more specific, we will be less likely to lead you down the garden path.

Doug

 

[ErikdeBest]

Hi Sparkle

p.s. btw- how is working that reg. supply that You have made (with mosfets)- why didn't You use glassware's design?

I used that reg. PS in amp I built for a cousin. I decided for a reg.PS because mains in Brazil vary from 105V to 135V and acoording to my cousin it's working fine - comparing to a SS amp. Returning to your specific question, I can't answer exactly - could as well have built the glassware one...maybe in the next upgrade.

 

[sparkle]

The "ultimate" from a performance point of view will have to use a bipolar or similar pass device (highest possible transconductance, lowest rbe) and a very high gain-bandwidth op amp as an error amplifier. Tubes won't cut it (transconductance too low) nor will MOSFETS (much better transconductance, but not as good as bipolars). Joe Curcio has published some designs like this. Check out his website and his preamp articles in Audio Amateur.

Oh boy- now I know where I have seen this design- in Daniel preamp (that is preety interesting since my name is daniel too). Thanks for that information......

DougL
- I have played with 5687 dual triode trying to make something that will work like I like (line stage for my Onix power amplifier). It came (after 10 or more designs) that a simple grounded cathode works the best for me. Now I run those tubes in paralel with 14mA per channel (7mA per triode) and it is nice. So I thought to play a little bit with power supply since I have only C-L-C-R-and C&R in paralel (before I try a my follower or gary pimm's ccs out of it).

Could You be more specific about the one You said that different ps should be used for phono and different for, say, line stage.
thanks
regards
daniel

 

[Sch3mat1c]

FWIW, here's a design I drew up a while ago. If you bump a few of the parts it can do any voltage range you wish.

Tim

 

[sparkle]

Thanks Tim, that's a nice one... I think that I would rather try something with bipolar's or mosfets because of the things mentioned in the previous posts of this thread....if this fails then I will try something with tubes......
regards
daniel

 

[jewilson]

You can use a LM317 for the high voltage. Just remove the adjust resistor you and cahnge it to zener diodes. This works well, and I have done it many times.

 

[Sch3mat1c]

Bah, tubes are sturdier than transistors. More than 20V across a MOSFET gate = plink... Too much voltage, dissipation or excessive negative base voltage on a bipolar, plink! (Addmittedly, bipolars are quite sturdy otherwise and will work fine here. Just don't touch the heat sink.) LM317's are useless as you lose the 40V input range quite easily over line variation and transformer regulation.

Tim

 

[fdegrove]

Hi,

Fifteen years of experience with "exotic" tube stuff taught me one thing for sure:

First thing to die in a tube pre/amp is those fancy pancy sand regs, the ones you don't have a clue what the heck they are as these guys have of course their references wiped away and have to make a dozen phone calls for to find out what the heck they are etc., etc...

Not to mention the fact that they make your tube stuff sound like sand as well...could just as well have used semis in the first place...

Another advantage is that these tubed regs using those lovely VR tubes make for a nice slow start as well....Great, no?

Since we care about sound, why not use what is reportedly the best? A shunt reg....

Cheers,

 

[SY]

Frank, Tim, I've been using various bipolar-based HV regs for about 20 years and have only lost devices through my own stupidity. If you understand how to protect and stabilize them (and certainly Curcio does), they've got lower source Z and lower noise than any tube-based reg I've ever tried. Solid state does better at DC, there's no way around it.

 

[fdegrove]

Hi,

If you understand how to protect and stabilize them (and certainly Curcio does), they've got lower source Z and lower noise than any tube-based reg I've ever tried. Solid state does better at DC, there's no way around it.

You knew I'd have to respond to that so here goes:

While I'm sure Mr. Joe Curcio understands how to design a fine SS regulator, there are other factors that to me at least are of importance in designing a regulator for audio use as well:

- Low Zout for a regulator is secondary to me for as long as the Zout is consistent throughout the frequency range we envisage of using.
If we consider the fact that series regs are Class B devices heavily depending on NFB and the response time of the error amp it becomes clear that any series reg will be relatively "slow" no matter what we do. This can become painfully obvious with loads operating in classes other than Class A but I digress.

- Noise, no big deal: VR tubes are often quieter than common Zener diodes and far more reliable in operation too. Whatever noise is left it's far less obtrusive than Zener pops and crackles and when operated and filtered correctly the VR tube will be rocksteady noisewise.

- Solid state does better at DC. Fine by me but that hasn't convinced me in view of a tube load attached to it which opearates quite likely in Class A at about 30mA anyway.

- With a tubed regulator, provided it's wired correctly, you can remove the load: no problem.
You can pull the series pass device: no DC pass, no problem.
You can pull the error amp or VR, ditto.

- Despite my stupidity, I never had a single tube regulator die on me yet, ever.

-There's a host of things you can do with what at first looks just like a regulator, or at least I know what you can do with them by using my tubes whereas I'm not certain at all I could achieve the same with semi-conductors...the latter you can easily chalk up to my ignorance in that field of course.

Cheers,

 

[jackinnj]

an IRFPG40 will do very nicely as a high voltage regulator -- the design is shown in Horowitz and Hill --

in Kepco's very high voltage/high current supplies, the 2kV, 200ma variety -- the tubes originally installed are replaced with a bank of FETs.

 

[SY]

Frank, the reference voltage source is, of course, one of the keys to low noise, stable regulation. Different methods have different virtues. One advantage of the zener string is that it can be bypassed more heavily than a VR tube. A disadvantage is that the voltage coefficient with temperature is not so good. But that's less important for a preamp stage- it's not as important that the rail be 245 versus 250 volts, but it is certainly important that the noise be low.

The two versions I use of Curcio's regulators have two different means of generating the reference. The older ones use a FET configured as an adjustable current source. The CS pulls voltage across a rather large resistor to generate the reference voltage. The newer ones use a bipolar/317 combo to generate that voltage directly. In each case, a FET-input opamp can look at the reference voltage through a decently sized RC filter. That also slows the startup.

The output of the regulator is stabilized quite nicely by a good-sized output cap (seried with a small resistor). The result can easily be seen with a scope probe while playing music through the powered circuit- thatis, you don't see much other than DC. I haven't seen (or heard) any ill effect from this sort-of class B operation. The regulator does what it's supposed to do, provide a stiff, low-noise rail that well decouples each stage from the others.