1. Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
    Dismiss Notice

60V @ 50 Amps linear power supply repair help needed please.

Discussion in 'Repairing Electronics' started by Chris Wilson, May 18, 2017.

  1. Chris Wilson

    Chris Wilson Member

    Joined:
    Feb 21, 2015
    Messages:
    31
    Likes:
    2
    Location:
    Shropshire UK
    I am having a bit of a saga with my Farnel H60/50 linear power supply I mended years ago. It's taken to blowing pass transistors. What happens is I power it on and often it works fine. I often run it at the full 60V output with a current draw of about 35 to 40 Amps. Occassionally, and seemingly more frequently, it will have uncontrolled volatge output, quite how high it gets I haven't managed to catch, but the 60V meter goes off scale. If I am very quick to turn it off it may work fine again. If I am not quick a pass transistor or two fails shorted and takes out D1 and R5 on the "control board". These machines seem to have undergone several revisions, and although I have schematics that almost match, the manual I have is for an earlier version that used all discrete components, whilst this has some IC's. I think the functionality is similar for all though. I acquired another one at the weekend, but it's sat in my estae car until I can bribe my pal to help get it upstairs to my shack. I took a look at the boards in that and whilst very similar d1 on the control board has a 10nF film cap across it, as shown on the schematics, whilst my control board does not.

    I could really use some help with fault finding as I am still pretty raw at diagnostics.

    I think I need a way to protect the PITA to change pass transistors whilst i try and see if the voltage rise is due to the SCR board, or where the trouble lies. Am I right in thinking the primary winding SCR's should keep the secondary no higher than say 10V above the voltage dialled in, to keep the heat in the pass transistor assembly down? My own thoughts are I should try and isolate the pass transistors somehow and see if I can find why the input to them rises uncontrollably, checking if the SCR's are being fired. The issue is how to get the SCR board running with no output signal from the pass transistor bank. A more off the wall idea, that's cost;y, is to change *ALL* the pass and pass driver transistors from 2N3055's to the higher voltage, higher current limit MJ15003 devices. Then if the output rises to the full rectified transsformer secondary level they should still survive I think? I may be well off though, so yourr advice is sought! I have reasonable test gear including a decent scope. I have changed all the boards electrolytics a few years back. On another forum someone made this statement about why a shorted pass transistor takes out D1 and R5, which lights up like a bulb before opening:
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    "Hi Chris .
    You didn't say if D1 was open or short but (they usually fail shorted due excess voltage but they can fail open if you have put a large current through them )
    Anyway I would say yes a shorted pass transistor could cause R5 and D1 to fail (and possibly p2,p1 voltage adjust pot if your unlucky) due to excess current through those devices .

    Explanation :- If a pass transistor is shorted then V output would then be at max and someone would try to reduce ouput by turning V adj to minimum value ,now since the +-15V aux supply gnd rail is tied to main V+ out (so you will now have 60 V between this and -Vout) so now an excessive current will flow through D1,R5 via P1-P2 voltage adj control (which now is set to a low ohmic value when you tried to turn voltage down )back to -Vout .

    Note that a failed D1 on it's own ( D1 is shorted) could also cause R5 to fail because the output could again be at max V (this time the inputs of IC1 would be shorted together by D1 and so IC1 output would be stuck at one of the supply rails due to it's own input offset voltage ). Again trying to adjusting the output voltage down would cause large current through R5 ,P1-P2 to -Vout ."
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    I found my actual board didn't have C10 as mentioned above. But I am loathe to add components whilst trouble shooting in case I make matters worse.

    I can link to the schematics for both my machine, which seem to match it 99%, and for the earlier one with all discrete components instead of op amps and IC Scmidtt triggers. I have the manual which explains the topology of the earlier version too. I can't remeber if I can post links here though...

    Having rebuilt the pass transistor tunnel heat sink assembly several times, and it being costly and time consuming, I need a better way to diagnose this!!

    Thanks for looking.
    Schematics that match my faulty unit 99% fully, and the manual that explains the board functions, but is for an earlier version that uses all discrete components are at:

    http://www.gatesgarth.com/supply.zip

    The schematics for the earlier all discrete components version are at:

    http://www.gatesgarth.com/H60-50_Diagrams(2).pdf

    All help greatly appreciated. I am as yet unsure if the new one I got at the weekend is functional or not, I just saw it powered up and the voltage control and meter working with zero load on it. I don't really want to mess with it and end up with two knackered ones <LOL>
     
  2. tomizett

    tomizett Active Member

    Joined:
    Aug 11, 2013
    Messages:
    429
    Likes:
    46
    Location:
    UK
    Hi Chris,
    Sounds like an interesting problem - I think a friend of mine had one of those big Farnell units - they're serious beasts. I'd be fascinated to know what you're doing with that 35-40 amps.
    When I get the chance I'll havea look over the diagrams, see if anything comes to mind.
    I'm sure someone here will have some ideas!
     
  3. Chris Wilson

    Chris Wilson Member

    Joined:
    Feb 21, 2015
    Messages:
    31
    Likes:
    2
    Location:
    Shropshire UK
    Hi tomizett. I use it among other things for powering two 1kW quasi Class D LF push pull amps running on the 136kHz ham band, run in parallel into a combiner, so they draw 60V @ about 40 Amps.

    Tonight I did some other tests, so I will take this opportunity to update:


    Rebuilt pass block with new 2N3055 to replace shorted one, tested all others, and tested again as an assembly before fitting. Checked with analogue Ohmeter front panel Volts control pots were not intermittent across their full range of movements. Disconnected bridge and smoothing caps from pass block at red line see http://www.gatesgarth.com/H60001a.jpg Base and emitter links to pass transistor block still connected. Wires to the two pass driver 2N3055's still connected. Input from 10 and 17 on Control Board left connected to bridge output, as well as to 15 on SCR board. Metering with DVM at output of full wave bridge, no load applied, the Volts knobs smoothly increased / decreased bridge output from a minimum of about 8V to a max of about 65V. Seems SCR stuff is working OK? No problems.

    Left voltage from bridge at just over 8V.
    Linked pass block to bridge with THIN croc lead, (at red line on H60001a.jpg) no real spark, no heating of lead but almost instantly Control Board R5 lit up like a bulb. Schematic at http://www.gatesgarth.com/H60002a.jpg Checked the pass transistors en bloc again with Peak transistor tester and all still good. One previous time that this fault occurred the main 50 Amp fuse blew. Cap or diode across output of supply intermittently faulty? I checked D24 when I rebuilt the pass block, again with my Peak tester and it seemed OK. Haven't tested C36 but it's on thin leads from what I recall, so would open them with over 50 Amps and shorted, I'd have thought.

    Can I test Control Board output on pin 6 and inputs on pins 9 and 10?

    Looking at Control Board again I lifted one leg of each of C3 and C5, which are original. My Peak ESR meter showed C3 as 4.9uF zero ESR but C5 was only 47nF on a Philips capacitance meter that's actually in cal still.
     
    • Like Like x 1

Share This Page