Author Topic: SOLVED: HP 6115A Crowbar Circuit Mystery  (Read 2271 times)

0 Members and 1 Guest are viewing this topic.

Offline motocoder

  • Supporter
  • ****
  • Posts: 769
  • Country: us
  • Electrical Engineer
SOLVED: HP 6115A Crowbar Circuit Mystery
« on: December 18, 2015, 10:45:50 pm »
As a fun project to work on over the holidays, I purchased a broken 6115A from eBay. There were several issues with the supply, but I've resolved all of them save one. When the supply has been off for awhile, when powered on, the "Overvoltage" light which is the indicator for the crowbar circuit, gradually starts to light up, and then fades back off. The PS seems to be supplying the programmed output voltage just fine - so the crowbar isn't completely tripping, but here's enough current flowing to light up the indicator LED. This is really just a minor annoyance, and the only reason I'd like to resolve it or at least understand it is to be sure there isn't some deeper problem to which this is just a symptom.

I've attached image showing the crowbar circuit. The full manual is too big to attach here, but can be downloaded from Keysight here:

http://literature.cdn.keysight.com/litweb/pdf/5950-5976.pdf?id=734221

The theory of operation for the crowbar circuit is in section 4-52. Basically there is an LM311 comparator, U1, that is comparing a divided version of the output voltage with a reference value set via the crowbar adjustment pot, R3. If this divided version of the output voltage exceeds the reference value from R3, the comparator output switches from positive to negative. This allows capacitor C14  to discharge through the primary of transformer T1, which in turn causes a pulse on the two secondaries of T1. One of these two secondaries is connected across the Gate/Kathode junction of SCR CR20. This causes the SCR to trip, shorting the output, but also reducing the base voltage on the series regulator transistors, causing the output voltage to drop to a few tenths of a volt.


The indicator LED, DS3, and a current-limiting 1.3K resistor, R31 are connected to an unregulated DC supply. When the SCR trips, there is a return path that allows current to flow through the LED.

Also note that the 1 ohm series current-limiting resistor R34 shown in the schematic is not present in the supply. I've checked another supply that also has this missing. There's a trace on the PCB where this resistor would be, so it was obviously removed in later revisions of the circuit, but not updated in this schematic. Inductor L1 is present, however.

Here are the symptoms in more detail: The supply needs to sit in an off condition for some time. Then when powering it on, you will see the output voltage come up immediately, as normal. The Overvoltage LED will get brighter over a second or two, and then start to gradually get dimmer. I can see it is still lit up to maybe 30 seconds later. There may be some residual light even after that - it's too bright in the room here to be sure at the moment.

Comparator U1 has been replaced, and I think is working fine. I also swapped out C14 and C13 with the equivalent parts from a badly broken 6114A supply that I have (these parts are identical between the two models). I briefly tried swapping out the SCR (CR20). I didn't have an exact equivalent SCR, so I had some triggering issues with the replacement SCR and had to remove it. Notably, the overvoltage light problem remained with the replacement SCR. I checked the three diodes CR16, CR17, and CR18 using the diode check feature on my multimeter, and they all seem fine.  The 7.5V supply shown in the schematic is actually ~7.0V, however, the manual says this is within spec.

One theory that I have is that the owner replaced the LED with a modern variant that is much more sensitive at low current levels, and some leakage current through the SCR is causing it to light up. However, this doesn't explain the behavior where the light first gets bright and then dims. I considered parasitic capacitance in the SCR, CR17, and CR16 causing some current to flow, but the RC constant of R31 and this capacitance doesn't seem like it would be large enough for these long times. Also, why would it first get brighter and then dim (maybe something to do with the PS output ramp-up?)

One more data point - I notice with this PS that the meter is reading slightly below the actual output value, and the pot used to calibrate this can not be adjusted enough to calibrate it correctly. Not sure how this could be related, but I mention it anyway.

« Last Edit: December 19, 2015, 12:39:57 am by motocoder »
 

Offline RobK_NL

  • Frequent Contributor
  • **
  • Posts: 250
  • Country: nl
Re: HP 6115A Crowbar Circuit Mystery
« Reply #1 on: December 18, 2015, 11:33:22 pm »
Either CR16 or CR17 might be leaky. Disconnect one at a time and see if that makes a difference.
Tell us what problem you want to solve, not what solution you're having problems with
 

Offline motocoder

  • Supporter
  • ****
  • Posts: 769
  • Country: us
  • Electrical Engineer
Re: HP 6115A Crowbar Circuit Mystery
« Reply #2 on: December 19, 2015, 12:39:37 am »
Either CR16 or CR17 might be leaky. Disconnect one at a time and see if that makes a difference.

Thanks, RobK_NL - this seems to have been the issue. I am not sure which one was bad, because I replaced both of then with some 1N5404 that I had laying around. It's a big pain to remove the A1 board to replace anything, so I just decided to replace all the diodes of that type. I replaced CR15 as well while I am in there.

No more bogus overvoltage light :)
 

Offline RobK_NL

  • Frequent Contributor
  • **
  • Posts: 250
  • Country: nl
Re: SOLVED: HP 6115A Crowbar Circuit Mystery
« Reply #3 on: December 19, 2015, 09:02:22 pm »
Great! Good to hear that.  :-+

They're really nice supplies. I have a misbehaving 6112A here, but not enough time to fix it.
Tell us what problem you want to solve, not what solution you're having problems with
 

Offline motocoder

  • Supporter
  • ****
  • Posts: 769
  • Country: us
  • Electrical Engineer
Re: SOLVED: HP 6115A Crowbar Circuit Mystery
« Reply #4 on: December 19, 2015, 10:37:24 pm »
Great! Good to hear that.  :-+

They're really nice supplies. I have a misbehaving 6112A here, but not enough time to fix it.

Yes, the quality of these supplies continues to amaze me. After I fixed the main issues with this one (bad op-amp, corroded range switches),  the output was within a millivolt without me having to calibrate it!

Anyway, thanks again for your help.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf