Author Topic: psu design questions  (Read 806 times)

0 Members and 1 Guest are viewing this topic.

Offline m3vuvTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1738
  • Country: gb
psu design questions
« on: April 13, 2021, 01:09:47 am »
Hi all,ive just come across this,just a couple of questions before i build it, 1,does it look viable to add bigger filter caps and transformer +another 3 or 4 pass transistors + emitter resistors and change the current sense resistor to get say 20 amps,also will this work as cc/cv or just current limiting?,cheers in advance m3vuv.
 

Online xavier60

  • Super Contributor
  • ***
  • Posts: 2783
  • Country: au
Re: psu design questions
« Reply #1 on: April 13, 2021, 01:58:26 am »
Does it use an lm723?
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline coromonadalix

  • Super Contributor
  • ***
  • Posts: 5795
  • Country: ca
Re: psu design questions
« Reply #2 on: April 13, 2021, 02:20:36 am »
what design(s) are you using, you don't give much to work on ???
 

Offline m3vuvTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1738
  • Country: gb
Re: psu design questions
« Reply #3 on: April 13, 2021, 07:29:48 am »
sorry i forgot to attach the pdf,try this.
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12805
Re: psu design questions
« Reply #4 on: April 13, 2021, 08:50:58 am »
That's an interesting variation on the usual two OPAMP diode ORed CC/CV PSU.  The LM723 is only used to provide a stabilized 7.5V reference to the two OPAMPs, and the whole control circuit is powered by a +/-16V (approx) unregulated supply referenced to the positive output terminal.  It will suffer from the usual defect of overshoot when transitioning quickly  from CC to CV mode and visa versa, as the OPAMP which isn't 'in charge' will be railed due to its feedback loop being effectively open, and its recovery time after transition will be determined by the OPAMP slew rate.  Its also got 47uF of output capacitance so I wouldn't rely on its CC mode for stuff like testing LEDs as the surge current when initially connecting them could be problematic.

Increasing its output current to 20A would be theoretically possible by redesigning the pass transistor array to have many more output transistors, multiple driver transistors and add a pre-driver, but with approx 45V on its reservoir capacitor when unloaded, shorting the output with the current limit at 20A would require the pass transistor array to instantaneously dissipate 900W, dropping by a couple of hundred watts as the unregulated input voltage settles to its fully loaded voltage.  That would make it a dammed good room heater, with a massive heatsink and nearly enough fan cooling for vertical takeoff!

TLDR: High current medium voltage linear CC/CV PSUs aren't really practical without a pre-regulator or tap switching to reduce the dissipation.

I attach the schematic extracted from the PDF for ease of discussion.
« Last Edit: April 13, 2021, 09:31:05 am by Ian.M »
 
The following users thanked this post: m3vuv

Offline coromonadalix

  • Super Contributor
  • ***
  • Posts: 5795
  • Country: ca
Re: psu design questions
« Reply #5 on: April 13, 2021, 10:57:43 am »
the best approach would be switched secondary tranformer taps, to limit excessive dissipation losses

The lm723 could be subbed by a precision reference too ...

For higher current(s)  you need to recalculate  the shunt resistor (R21) and the  pseudo darlington circuit (T2 T3 ...)

But as mentioned  the lm723 as a max vcc supply

There's a limit to the filtering capacitors (C10, C1,C2)   check google for calculations,  you need to add 0,1 uf 100v  in parrallel of them

And you could add an anti reverse protection diode at the output terminals,  it will help if you drive inductive loads

And yes this circuit may dissipate a lot, add temperature controlled fans, or a small fan controller circuit ??

You have many leaked Mastech (and the clones) psu's on the web, you can check how they implement the secondary xformer taps

Add multi turn pots instead of a single turn
« Last Edit: April 13, 2021, 11:04:58 am by coromonadalix »
 

Offline m3vuvTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1738
  • Country: gb
Re: psu design questions
« Reply #6 on: April 13, 2021, 11:31:07 am »
thanks but i alredy thought of all those mods,it seems dooable,also thought of sort of transfo preregulator taps too cut down the pass transistor power dissapation,its just a design im putting together,also thougt of an adjustable crowbar circuit  maybe mechanicaly linked switches to controll it.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: psu design questions
« Reply #7 on: April 14, 2021, 10:23:22 pm »
Hi all,ive just come across this,just a couple of questions before i build it, 1,does it look viable to add bigger filter caps and transformer +another 3 or 4 pass transistors + emitter resistors and change the current sense resistor to get say 20 amps,also will this work as cc/cv or just current limiting?,cheers in advance m3vuv.

That will usually work to increase the output current capability.  However doubling the number of output transistors will also increase the drive requirements which may need to be addressed.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf