HV SS bench top power supply, advice needed

[panos29]

I just finished testing a nice bench top power supply, using the old but trusty LM723! The outcome is a good 85-380V span with more than 200mA capability(limited only by fet's SOAR, the output noise plus hum is less than 0,5mV mostly random noise at all output voltages and currents. The schematic is correct excluding the triode and adding the intersected line.

Only problem the usuall...exploding fets at shortcircuits, thus it cannot be "certified" as a serious bench supply, intented for experimentation.

So my idea is to use a triode as pass element like 6080, 6336 and why not 6C33. I made a possible schematic, but before testing I thought I might ask first, is this viable or not? If not, what might need to be changed-altered(intersected output link is present only when used with fet pass element).

 

[the_manta]

Hi !
Hans Peter recently postet a nice circuit which he developed. Only needs LM317 + FET
see here: http://f23.parsimony.net/forum45451/messages/186404.htm

Regards, Simon

 

[zigzagflux]

Panos:

The LM723 is easily used with foldback current limiting. Just a few transistors and resistors, and you end up with short circuit protection. Have you considered this?

 

[panos29]

I did cosidered this but unfortunately in real world the outcome is very different! Actually there is no way to limit the current efficiently and on time when you short circuit a 400V supply. It just blows everything, most of the times the mosfets and some others the 723 too. Odd as it seems the foldback characteristic works excellent for any voltages under 30Volts.

 

[zigzagflux]

Actually there is no way to limit the current efficiently and on time when you short circuit a 400V supply.

I've built regulated supplies for my PAS-3, and they worked quite well with short circuit protection. Granted, I used discrete op-amps and BJT's instead of the 723, but it can be done. There may be something specific to the 723 that prevents proper operation, but any practical voltage supply can be current regulated.

Maybe you should try something other than the 723, rather than resorting to tube pass elements?

 

[panos29]

Maybe you are right about current limiting using other devices, BUT why not use tube pass elements? Afterall, I got plenty nice pass devices as mentioned in my first post, not to mention I love tubes! don't we all love tubes in here?

 

[astouffer]

How about using a tube like an 811A as the pass element? Seems like driving the grid positive would be an advantage here.

 

[panos29]

811 is nice and crossed my mind already but I would like to avoid it since heater wattage is a bit on the heavy side. Why nobody likes 6080???

 

[zigzagflux]

BUT why not use tube pass elements?

I can honestly say I have no direct experience building tube regulated supplies, but in the hours I have spent investigating the subject, general consensus is that SS regulation is superior (faster, tighter regulation). Similar to the constant current source. Can I build one with tubes? Sure. Which one performs better? FET.

Up to you, man. If you are resorting to tube pass element because of the SOA difficulties, then you have a design flaw, not a component (i.e. technology) flaw.

If you choose tube pass element because you LIKE tubes, that's a different story; then God Bless You, use tubes. But your original post was regarding current limiting and SOA, not personal preference.

jus twying to hewp.

 

[panos29]

I cannot agree more that fets are alot more suitable than tubes in the specific spot but I really like tubes as are far more tolerant devices than SS when it comes to shortcircuiting.

Moreover performace wise, lets say an 6C33 is excellent pass element and does not need any heatsinking. Moreover a tube looks alot better on a chassis than a heatsink .

 

[richwalters]

A 6C33 is an expensive luxury and will soon be getting rare. Try parallelling other tube versions ? topcap 12E1"s are dirt cheap surplus. Or are you against pentodes ?

richj

 

[panos29]

I am not against pentodes at all, especially 12E1s and el509s are excellent pass elements although 6C33 is THE pass element and not expensive and rare at all, yet.

Problem is that I tried the proposed schematic and it does not work, the tube pass element is passing all the available voltage to the output. Shoud try a different topology for providing Vgk to the tube. Any ideas?