I’m fixing a hole where the rain gets in
–The Beatles, “Fixing a Hole”
And stops my mind from wandering

For a while I’ve had an old HP 3435A desktop digital multimeter. Picked it up at a surplus place for $60 so I could have a meter that sits on a shelf and is always ready, doesn’t need batteries, has a decent readout, autoranging, basically a bunch of features that my existing meters did not have. Recently when I went to use it, I found that the voltage it was showing was impossibly wrong – I seem to recall it was an AA battery I’d just removed from something marginally working but it read 0.2V – and then shortly after the only thing it would display is “OL”, its code for out of range. Around the same time as that, I saw a video from Great Scott! that talked about a portable DMM that also has a two-channel 50MHz oscilloscope. Now I already have an old analog scope here, but .. well, it’s old, and it doesn’t have a lot of the nice features that newer scopes do. So after getting that new meter in my hands, I decided recently to take the old one apart and see if I could find out what was wrong with it.
On first look through the components I didn’t see anything wrong, though on closer inspection I noticed one filter cap (C403, the third in a row in the photos above) did seem to have a small split on the outer casing but nothing showing an obvious failure. After some measuring I found values far out of spec though; the 7VDC rail was reading around 5.8VDC, and at the positive side of the aforementioned cap I should’ve been seeing close to 15V and instead saw around 9. Ok, so it sounds like the cap might be shorted, or close to it, time to heat up some solder. Lifted the positive side of that one and used the new meter for one of its other purposes, a capacitance tester. What should be a 220uF reading was showing around… 17uF. Yeah, that’ll do it. I then found a 220uF ‘lytic cap rated for 50V instead of the old one’s 35V in my stash, which is perfectly fine to substitute; knowing it’s a filter and not for specific signal work, I didn’t have to worry about equivalent series resistance (ESR) which sometimes is an issue. I then used the fancy new meter to test this cap, not only to make sure the cap was good before installing it, but to make sure I got a good reading from it on the meter to help verify my previous reading on the “bad” cap was a good one. I could have lifted one leg of another cap to test, but… well I didn’t want to 😛
Unsoldered the other side of the cap and had to fiddle with the legs of the replacement one a bit, but once I got it bent nicely and stretched across it made the junction and fit nicely. I didn’t pull the board completely from the case to trim the back end, but it’s a plastic housing so I’m not worried about shorts, and I tried to ensure the leads only went in as far as they had to so there shouldn’t be too much sticking out the back. Next was to see how it worked, and turning on the unit I was greeted with a voltage reading instead of the now familiar “OL” I had been getting. Success!
In the process of testing I had moved at least one of the calibration pots inside, so I went to measure the 7V rail and make sure that things were in spec. It was off by a few tens of millivolts, so a quick adjustment and I was able to get it to 7.002 according to my fancy new meter – which I fully admit, without calibrating that device, who knows if it’s right, but I’m happy to be happy with it. Checked the internal frequency as well and found it riding just a tad high, so barely a tap on the pot for that adjustment and the frequency was hovering right in the proper range now. Time to close up and re-shelve it.
Needless to say, even with my old meter working well again I’m happy with the new one as well and looking forward to the kinds of things I can do with it now that I have a more modern oscilloscope (even if it’s a smaller bandwidth than the 1970s era Tek sitting on the cart next to it). Somewhere around here I’ve got a list of test tools I wanted to make if I got a different scope…
Note: I realized as I was putting this together that here was yet another post that would have benefitted from having more than one photo. I’ll see if I can remember to do better than that on future projects.. and I happen to have one I’m going to do soon. For the second time, since the first one failed. Which is good that I didn’t photograph it.