HP8662A - 20 Volt Linear Regulator supply issue

[DMarchesi]

I have an HP 8662A signal generator with an issue with its +20 volt supply on the A7A1 linear regulator board, i have worked through the service manual troubleshooting sections and am a bit flummoxed by what i find, the supply out of the A7A3 inverter board is 24 volts(so ok according to the manual), and the 20 volt linear regulator section is not being current limited as the voltage across the current sense resistors R27 & R28 is only 0.43 volts(below the 1.5v at which it would limit according to the manual). The red over voltage LED is not lit but neither is the green ok indicator LED . The -10 and -40 supplies are both working perfectly and displaying their OK green LED's

I have 24 volts on pin three of the 20v regulator - U3 which is an LM305 equivalent but 0.02v on the no.8 pin(the regulated 20v output)

so i thought i would isolate with a tiny strip of insulation tape on the outgoing pin that connects to the other boards that use the 20v supply to see if something there was shorted and instead of seeing the green LED light up as i was hoping for the overvoltage shutdown red LED now lights and completely shuts off all 3 supplies

i gather this could be because there is no load now so the output is too high but I'm writing this before i test the output again with the peak hold on my meter(I just tested the 23 volt output from the inverter board and it hits 41.08 volts) but this is with pin 2 of the A7A1A connector insulated so its not supply any other part/boards of the unit

Has anybody out there seen anything similar or does anyone have any pointers as to where i should be looking next

[frsimen]

I don't think there is anything wrong with the input voltage to the supply. Without a proper load it will rise significantly.

You have measured the voltage across the current limit resistors, but you haven't stated what voltage is on the 20V line when it is loaded. If the base of Q9 is at close to 0V, there will still be a significant voltage on the emitter of Q3, but it should be a lot less than 20V.

Assuming that U3 and the two transistors aren't faulty, it appears that U3 is sensing that the 20V supply has gone much higher than 20V and is trying to cut it back. The most likely cause of that is a leaky/ short circuit C16 or a break in continuity between the slider of R39 via R40 to ground.

Paula

[frsimen]

On reflection, my comments in post 2 above are wrong as I missed that you have stated the output voltage, which appears on pin 8 of U3.

U3 configured has fold back current limiting, so once triggered it settles to a lower current than the initial limit current. That is the current which you are measuring across the current sense resistors and is consistent with the lack of output voltage.

You should look for a short circuit/low resistance elsewhere on the 20V line. Tantalum decoupling capacitors are good candidates for this sort of fault.

Paula

[DMarchesi]

Thank you for your reply Paula,

i removed the board from the 8662A and tested it by putting in 23 volts with a 100 ohm 5 w resistor between the 20 v output pin on the board connector and ground and it was outputting 20 volts perfectly so i have deduced it was a problem elsewhere in the 20 volt rail on the other boards somewhere, i used my ohm meter to test all the +20 volt pins on the underside mother boards and it led me to the A4A3 LF Amplifier board with just 0.14 ohms to ground, i removed the board and tested C9 and C4 the power supply decoupling caps and C4 was a dead short, i have replace both C9 and C4 with new 4.7 microfarad tantalums and now i have no status light flashing and all is working as it should, for now until another of those blue devils decides its time to short itself

Thanks for you help

Originally Posted by frsimen

I don't think there is anything wrong with the input voltage to the supply. Without a proper load it will rise significantly.

You have measured the voltage across the current limit resistors, but you haven't stated what voltage is on the 20V line when it is loaded. If the base of Q9 is at close to 0V, there will still be a significant voltage on the emitter of Q3, but it should be a lot less than 20V.

Assuming that U3 and the two transistors aren't faulty, it appears that U3 is sensing that the 20V supply has gone much higher than 20V and is trying to cut it back. The most likely cause of that is a leaky/ short circuit C16 or a break in continuity between the slider of R39 via R40 to ground.

Paula

[Jez1234]

Nice to see a happy ending there! It seems old tants are becoming a major liability... I've had quite a few S/C tant issues myself over the past 10-15 years or so.