Beaten by Tektronix 7834 LV Regulator problem

[perdrix]

The supply was shutting down because the output voltages weren't correct. There is a resistive network with all supplies connected which should sum to zero at pin 2 (BAL SENSE) of the controller IC.   If it isn't 0V the IC outputs a stop signal on pin 8 which goes to base of Q1254 to shutdown the supply.    This was shutting the supply down, and removal of Q1254 on the Cap/Rectifier board allowed the supplies to run so I can fix the output levels (in theory).

The voltages I observed were

Supply   Measured

 -50V   -50V   
 +50V   +46.9V
 -15V   -14.58V
 +15V   +13.8V
 +5V           +4.275V

-50V is the reference for all the other supplies.   The other part of that story is "reference ground" which is connected to chassis ground with a 47 ohm resistor (R1591).

In my supply reference ground was 0.45V below chassis ground which was definitely messing up the output voltages.

Shorting R1591 and readjusting the -50V rail to exactly 50V I see the following:

 Nominal   Actual
 +5V      +4.75V
 -15V   -14.21V
 +50V   +47.70V
 +15V   +14.32V

Interestingly all of them are nigh on exactly 95% of what they should be!   I'm sure that this fact plus the issue of reference ground being at -0.45V without that shorting link should be telling me what's wrong but I just can't see it.

For reference they should be:
 
Supply                        Limits
 -50                  -49.8 to -50.2
 -15                  -14.85 to -15.14
 +5                   +4.9 to +5.1
 +15                 +14.85 to +15.15
 +50                 +49.5 to +50.5

The raw supply voltages from the Cap/Rectifier board are:  -17.6, +17.5, +54.4, -54.4, +7

David

[perdrix]

Problem solved - one of the sense bypass resistors had a "suspect" solder joint (specifically R1499) once that was touched up, all voltages measured between Reference Ground (TP1592, *not* TP1591 which is chassis ground) and the voltage test points were within specification.   The approx.. 0.5V between reference ground and chassis ground was a "red herring" and would seem to be normal!

David

[David Hess]

The sense bypass resistors do not do anything when the loads are connected because the force and sense lines short them out.  They prevent the regulated outputs from running away if the loads are disconnected.  This safety feature was not included on earlier Tektronix power supplies.

[perdrix]

I was running with a dummy load that didn't have a sense connection so was relying on the sense bypass resistors presence.

PS I've not heard from you on TekScopes since we moved to groups.io.  Did you not approve of the move?
D.

[David Hess]

I was running with a dummy load that didn't have a sense connection so was relying on the sense bypass resistors presence.

PS I've not heard from you on TekScopes since we moved to groups.io.  Did you not approve of the move?
D.

No, I approved of the move but gmail started throwing fits with duplicate and deleted messages which make it very difficult to use email lists.