Keithly 191 Repair

[at2marty]

A little while back I scored what I think is a decent Ebay auction.  I got a Keithly 191 multimeter along with a Keithly 175 multimeter for around $50 US including shipping.  I briefly checked them out when I received them, and they both seemed to "work" correctly.  They had a few cosmetic defects, which I don't really care about.  I was more interested in getting a couple of bench meters to mess around with.  The 191 is a nice meter with pretty high precision, and best of all, it's kind of "old school".

I finally got around to messing with the 191 a bit.  I noticed that even though it has 6.5 digit accuracy, the least significant 2 digits would never settle down.  They constantly would change readings,  I also noticed that sometimes when I powered it on, I would just get a "-1. " reading that would just flash, and would not go away no matter which buttons I pressed.  If I power cycled it once or twice, eventually it would start to "work".

The troubleshooting part of my mind went to work, and I went back to the fundamentals.  CHECK POWER!

I downloaded the manual for it and began testing power from both a cold power-on and a warm power-on.  The operating voltages are +5VDC, +15VDC and -15VDC.  I noted that the -15VDC was out of spec and the +5VDC was right at the limit.  I also noted that during a cold power on the -15VDC was essentially "not there".

Visual inspection revealed that a filter capacitor for the 5V regulator (electrolytic 100000uF 10V) was showing just a slight bulge.  The others looked fine.  I don't have an ESR meter but on a hunch I decided to replace all of the electrolytics in the power section.

When I removed the old capacitors, I found that the one on the -15V regulator had been oozing crap out of the bottom.  It took some time to clean the mess up, but I was able to solder in new capacitors.

After replacing the caps, I went through as much of the calibration that I could with the equipment that I have.  All digits are now steady and if I apply 190mV on the 200mV setting I get a reading of 190.001.  If I apply 1.9V on the 2V setting I get a reading of 1.90001.  If I apply 190V I get a reading of 190.000 on the 200V setting.

For calibration I used an old Eico resistance decade box (un-calibrated) and my Fluke 3330B voltage standard (un-calibrated).  However, the readings that I get are consistent with the other meters that I have.  I really don't care or don't need accuracy down to the nanoVolt range.

I'll post some photos when I get the chance to do so.  I really have a love for the "old school" equipment.