Author Topic: Pass transistor gain question  (Read 3409 times)

0 Members and 1 Guest are viewing this topic.

Offline Chris WilsonTopic starter

  • Super Contributor
  • ***
  • Posts: 1081
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Pass transistor gain question
« on: May 25, 2017, 03:39:33 pm »
A pass transistor assembly using 8 off 2N3055, in a 60V @ 50 Amp linear supply needs rebuilding. I was thinking of upgrading to the  MJ15003  devices. Farnell list two different gain types, 25 hFE and 150 hFE, is one gain level more appropriate than the other? It also has two driver transistors, again 2N3055 and i was going to swap those to the higher voltage, higher current capability MJ15003 devices    http://uk.farnell.com/search?st=MJ15003

Thanks.
Best regards,

                 Chris Wilson.
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12807
Re: Pass transistor gain question
« Reply #1 on: May 25, 2017, 03:53:11 pm »
Check the datasheets and you'll find they are the same min. HFE
 

Offline Chris WilsonTopic starter

  • Super Contributor
  • ***
  • Posts: 1081
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Re: Pass transistor gain question
« Reply #2 on: May 25, 2017, 04:01:36 pm »
Thanks, so which should I go for though? :)

EDIT, ahh, wait, you are basically saying they are the same.... I think?
« Last Edit: May 25, 2017, 04:06:24 pm by Chris Wilson »
Best regards,

                 Chris Wilson.
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12807
Re: Pass transistor gain question
« Reply #3 on: May 25, 2017, 04:12:57 pm »
Either!  They both have HFE specified to be between 25 and 150 @Ic=5A.   Farnell's parametric database has a typo for one of them - a data entry slave entered the max HFE when they should have entered the min.
 
The following users thanked this post: Chris Wilson

Offline Chris WilsonTopic starter

  • Super Contributor
  • ***
  • Posts: 1081
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Re: Pass transistor gain question
« Reply #4 on: May 25, 2017, 04:14:44 pm »
Perfect, thanks a lot Ian.
Best regards,

                 Chris Wilson.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14073
  • Country: de
Re: Pass transistor gain question
« Reply #5 on: May 25, 2017, 05:05:00 pm »
Only 8 Transistors for 50 A sounds like rather higher power for each transistor.  2 drivers 2N3055 on the other side looks a little overkill.  I would prefer 9 (or more) transistors driven from a single MJ15003.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19345
  • Country: gb
  • 0999
Re: Pass transistor gain question
« Reply #6 on: May 25, 2017, 05:28:49 pm »
Only 8 Transistors for 50 A sounds like rather higher power for each transistor.  2 drivers 2N3055 on the other side looks a little overkill.  I would prefer 9 (or more) transistors driven from a single MJ15003.
One thing to note is we don't know the maximum voltage drop across the transistors. Hopefully there's a tap changes so the full 60V @ 50A is not being burned off.

Thanks, so which should I go for though? :)

EDIT, ahh, wait, you are basically saying they are the same.... I think?
The actual transistor itself is exactly the same. The difference is the 'G' suffix indicates a lead free package, which you'll need if you intend to sell it in Europe.
 

Offline Chris WilsonTopic starter

  • Super Contributor
  • ***
  • Posts: 1081
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Re: Pass transistor gain question
« Reply #7 on: May 25, 2017, 08:09:06 pm »
Hi Hero999, thanks for the added info. The PSU has an SCR controlled primary that in theory, if not in occasional practice, holds the drop to around 8.5V. Unfortunately it takes it upon itself for a so far un-found reason, to deliver 90 odd volts to the pass transistors, and they don't like it ;) It may be RF getting in, it may be an incorrect set up, I am awaiting a copy of the service manual, I have been struggling with an incorrect one for a slightly earlier model with all discrete components on the 3 PCB's, whilst mine has some IC's and OPAmps. I always find intermittent faults very trying, I struggle enough with something that's stone dead!
Best regards,

                 Chris Wilson.
 

Offline danadak

  • Super Contributor
  • ***
  • Posts: 1875
  • Country: us
  • Reactor Operator SSN-583, Retired EE
Re: Pass transistor gain question
« Reply #8 on: May 26, 2017, 12:21:16 pm »
Some amps were built with screened match pairs / matched beta / hfe
transistors, so be aware of that. Sometimes the package was color
coded/striped to indicate hfe range.

Might be notes in the manual BOM or schematic.


Regards, Dana.
Love Cypress PSOC, ATTiny, Bit Slice, OpAmps, Oscilloscopes, and Analog Gurus like Pease, Miller, Widlar, Dobkin, obsessed with being an engineer
 

Offline Chris WilsonTopic starter

  • Super Contributor
  • ***
  • Posts: 1081
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Re: Pass transistor gain question
« Reply #9 on: May 26, 2017, 01:10:35 pm »
Thanks, can't see anything, I think the failures are related to the control circuitry, but I am running it hard, thanks for the tip!
Best regards,

                 Chris Wilson.
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12807
Re: Pass transistor gain question
« Reply #10 on: May 26, 2017, 01:29:23 pm »
I'd be looking at temporally patching a beefy solid state relay into the mains feed to the primary circuit that cuts off the power if the secondary DC bus gets too high.  You'll need an auxillary supply for it and the comparator that detects over-voltage and a latching circuit so it doesn't auto-reset.  That should at least save the  pass transistors as the SSR should respond in no more than one mains cycle.

Determining what happened to trigger the fault is more difficult - you'll need to run it under load with a deep memory DSO set up to do rolling acquisition with single shot external post-trigger off the cutout comparator, monitoring as many of the critical signals as you have channels for.  For slower signals a data logger may be useful.   Depending on how often the fault happens and whether or not you can find anything that will 'encourage' the fault, it could be on your bench for weeks waiting for enough data to figure it out. 
 

Offline Chris WilsonTopic starter

  • Super Contributor
  • ***
  • Posts: 1081
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Re: Pass transistor gain question
« Reply #11 on: May 26, 2017, 02:05:36 pm »
Hi Ian, the 50 Amp fuse is between the pass transistors and the bridge rectifier output, in the negative side (- ve floats relative to chassis ground). Would testing at max volts but low amps with say a 5 Amp fuse there save the pass transistors, or is it purely volts that kill them? Right now the bridge is showing 93V DC, with the one of the pass transistors shorted, but with the 50 Amp fuse open. I am pretty sure for the SCR stuff to work it needs to read the OUTPUT of the pass transistors as feedback, so it's probably doing as expected with no output from the pass transistors.

I have just taken delivery of some 140V MJi5003 pass transistors, but obviously i can't just keep buying the damned things ;) I'm 90% sure the SCR stuff loses control and the output from the bridge into the pass transistors goes to 93V and then BANG, one or more go short.

I was wondering what these devices would make of 93V in, if I bypassed the SCR stuff to run the primary at full mains volts, to do a divide and conquer assessment of where the issue lies. If I ran plenty of volts out the pass transistors wouldn't have a huge drop across them even without the SCR control. I am not sure if I can divide SCR from pass control though...
« Last Edit: May 28, 2017, 12:25:27 pm by Chris Wilson »
Best regards,

                 Chris Wilson.
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12807
Re: Pass transistor gain question
« Reply #12 on: May 26, 2017, 02:19:28 pm »
The fuse will be too slow to save the semiconductors.  Also, I'd bet its before the reservoir caps, and there's probably enough energy stored in those to vaporise the failing pass transistor junction.    If you wanted to do any good there you'd have to design and build a fast-acting MOSFET based electronic trip.  It probably needs to be able to carry several hundred amps of inrush current so build it beefy!
 
The following users thanked this post: Chris Wilson

Offline Chris WilsonTopic starter

  • Super Contributor
  • ***
  • Posts: 1081
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Re: Pass transistor gain question
« Reply #13 on: May 26, 2017, 03:38:07 pm »
Actually it's after the reservoir caps. Thinking back to when I first got it as a totally none operational unit with one of its three PCB boards missing, i found a cracked trace to one of the OP AMPS on the "Control Board". I wonder if another is cracked and intermittent? My eyes aren't quite what they were, but luckily not too bad for close up work, I'll get a good light on it, or take it outside before this lovely sun goes down, and have a thorough look over the tracks. Thanks Ian.
Best regards,

                 Chris Wilson.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf