High Voltage Regulator Design

[Seraph]

[/QUOTE]Originally posted by jackinnj

I recommend bypassing the adjust pin of the LR8N3 regulator.

What does bypassing the adjust pin do?

Is there any benefit (or deterimental effect on the regulator) if the cap value is increased?

 

[qq]

I guess I can't understand why you would have D1 in that circuit you linked to unless it's required for that regulator, which is a type I am unfamiliar with. That's not your problem, but it seems unnecessary.

It's to protect against the case where the voltage at the output exceeds the input voltage, such as if there were a large capacitor that remained charged after the power is removed. Or that's my guess, anyway. It's in the app notes for the regulator part.

(btw, that's a .1uF cap, not 100pf, in the schematic, not that it matters)

That's a pretty broad question; you could have used a bad part, used an NPN device instead of a PNP (depending on supply polarity), exceeded specs for your device, wired it wrong .....

I believe this use calls for a NPN transistor, but anyway I mostly wondered why this circuit would work fine when dropping 200v but blow transistors when dropping 50v.

What does bypassing the adjust pin do?

I'm guessing that bypassing the adjust line with a cap might protect against sudden transients that could induce voltage spikes? On the other hand, that might also lead to more ripple? Just speculating here.

 

[kevinkr]

Bypassing the adjust pin is normally done to reduce ripple.. (See LM317 app notes for details.) Don't omit the diode across the current setting resistor (from out to adj) in any high voltage application regardless of capacitance value. A power failure or short on the output can result in large currents flowing back up through the adjust pin in that situation and can result in an exploded part. The diode from input to output is also necessary and provides a path for current from the output cap and reference cap to quickly equalize voltages across all pins in a fault situation.

 

[phi70]

Bypassing Adjust pin with 10uF or 0.1uF?

Greetings,

May I ask if the bypass cap for the adjust pin should be a small value like 0.1uF or 10uF as suggested previously?

I notice that the bypass caps in the input and output are 0.1uF. Please excuse my *density* here.

Cheers,

 

[astouffer]

May I ask if the bypass cap for the adjust pin should be a small value like 0.1uF or 10uF as suggested previously?

The LM317 datasheet suggests 10uF and says anything above will have little effect. The LR8N3 should be similar.

http://www.antiquewireless.org/otb/resto0504.htm

I'm using this to regulate the screens of my latest amp and its working well so far. No TIP50 transistors handy but an old monitor pcb had this, in an isolated package too 😀

http://www.datasheetcatalog.org/datasheet/panasonic/SJD00126BED.pdf

 

[phi70]

Thanks for the quick response

10uF it will be then... sadly have to resort to 450V 10uF lytic due to size.

As for the 0.1uF bypass @ the input and output sides,
any suggestions as to whether a mica cap would suffice or is there a
need to get audio grade stuff like Auricap or Dynamicap or PIO?

Thanks again.

 

[astouffer]

As for the 0.1uF bypass @ the input and output sides, any suggestions as to whether a mica cap would suffice or is there a need to get audio grade stuff like Auricap or Dynamicap or PIO?

No need for anything fancy here. A normal mylar or poly cap should be fine. Save the expensive stuff for the signal path.

 

[phi70]

Thanks a lot for the info

I'm all set.

 

[Hi_Q]

I am considering using a LR8N regulator but with 2N3585 pass transistor as I have load of them, possibly using two in darlington configuration. Any comments on this?
Cheers
Les

 

[phi70]

Seems like that transistor is already good for pretty high current

How many and what type of tubes are you planning on powering?

 

[mod_evil]

I'm going to design a High Voltage Bench Supply.

http://www.novotone.be/fr - Projects - 05 Alimentation HT de laboratoire - Gamme de 50 à 450 V - 500 mA

I saw this design. I belive It will be good. I'm studying to make this Pre-regulator for more voltages and use this in a Pre-amp or something else.

Regards,
Felipe

 

[Hi_Q]

" How many and what type of tubes are you planning on powering? "

I just want to use two regulators, up to say 100mA, in a valve testing configuration, one for plate the other for screen. I have some LR8N regulators and I am a thinking they might be good in this application so I can vary the potentials as required. Has anyone applied an external current limit to the LR8N circuit shown in the earlier replies?
It would be nice to have a bit of protection against the odd faulty valve under test.

 

[steve1112]

This article might help:
(provides a regulated output up to 350 volts and 100 mA)

http://www.antiquewireless.org/otb/resto0504.htm

 

[astouffer]

Has anyone applied an external current limit to the LR8N circuit shown in the earlier replies? It would be nice to have a bit of protection against the odd faulty valve under test.

Wonder if a comparator monitoring a current sense resistor would be quick enough to latch a relay?

 

[Hi_Q]

Seeing as lowering the bottom resistor lowers the output voltage, I wonder if a simple sense transistor in the position shown would work. Looks like a Christmas holiday bread board project. By the way, seasons greetings to all on this Forum and best wishes for 2009, it's just got to be better than 2008!
Les

 

[phi70]

LR8N3 regulator again

Greetings to all,

I am looking at the schematic for the regulator
near the start of the thread again and kept wondering
if the top of R1 (the 2kOhm resistor) should be connected to the output of the series pass transistor rather than the output of the LR8N3.

Shouldnt the LR8 be actively adjusting the output based on the
actual voltage of the output of the whole regulator in response to the draw of the load which may cause different degrees of voltage drop?

Cheers,

 

[dsavitsk]

Notwithstanding phi70's question, anybody spot any issues with this implementation?

An externally hosted image should be here but it was not working when we last tested it.

 

[kevinkr]

I'm going to design a High Voltage Bench Supply.

http://www.novotone.be/fr - Projects - 05 Alimentation HT de laboratoire - Gamme de 50 à 450 V - 500 mA

I saw this design. I belive It will be good. I'm studying to make this Pre-regulator for more voltages and use this in a Pre-amp or something else.

Regards,
Felipe

Felipe, What a great site! Thanks for bringing it to our attention. I've been reading articles and looking at schematics this afternoon and have added it to my permanent bookmarks. Many of the designs I've looked at are quite unusual, and all are very well explained in the accompanying text. (Most have been published in LED or Electronique Pratique.) Definitely worth a look IMHO, but note that you do need a (good) working knowledge of technical French to read this site.

 

[Cobra2]

Notwithstanding phi70's question, anybody spot any issues with this implementation?

Nice! Maybe extend tracks for R5 to a terminal-block, if you want to use a choke instead?
Watch the distance from the rectifier to nearest cap...
And, if possible, make room for holes under the heatsink?
May I buy a couple of the finished PCB? 😉

Arne K

 

[phi70]

Implementation seems right

I realize there is no way I can connect the 2k resistor to the pass transistors output to serve as a feedback mechanism to the adj of the LR8N3. Bummer, which means this regulator itself is only good as a line voltage regulator and not a regulator that can clamp load-induced voltage variation like a Janus Shunt regulator.