Knowing no end to my suffering, I realized I enjoyed repairing a friends Fluke 8021B so much, that I bought my own off eBay for $20 and decided to repair it. Checking the meter out before the repair, everything seemed to work just fine, including the historically amazing continuity beeper12.
The unit arrived and was pretty dirty. So I took it all apart, and cleaned the case and buttons in Mr. Clean (as recommended by Mr. Modemhead). The power switch also was a bit finicky, as it required fiddling to get the unit to power on. That in mind, I washed the whole board in copious amounts of IPA (including the switches), and after drying I sprayed all the switches with Deoxit F5 (which is meant for switches). I may try Deoxit Gold instead should things continue to be funky. Calibration, according to my DMMCheck Plus, was way off, about 15%, and readings were very noisy. Closer inspection showed some corrosion around the lower right of the LCD, right where the three electrolytic capacitors are. A search through my parts bins showed I had been smart, and ordered spare 22uF 16v electrolytic caps when I re-capped my Fluke 8060a. Pulling the old caps from the 8021B, they were green (corroding). I used a fair amount of IPA, and scrubbed the top and bottom of the board in that area, and removed the old dull/corroded solder on two of the resistors, and around the capacitor solder pads.
Installing the new ones is easy, and it’s amazing how much better they fit given capacitors have shrunk so much in ~30 years. Putting the unit back together, it still read as off, but seemed “better” and less noisy. Looking through the manual, the calibration is actually pretty straight forward. The attached picture is the calibration procedure. I used my Rigol DP832 and DG1032Z to do the calibration, using my Keithley 196 as the confirmation (since neither Rigol units are exactly transfer sources). The only notes is when doing the AC calibration using a Function Generator, make sure your source impedance is High-Z. Also the setting should be 191mVrms. Once I did all that, the unit calibrated easily, and the other ranges all checked out perfectly, including ohms, and current with reference to my DMMCheck. Awesome.
Now I just need to replace the screen at some point using the same procedure I followed here. But right now, the screen is readable, so I don’t care that much. Good luck if you get one of these units. I thought the issue was going to be the voltage reference, but instead it was just those caps. Now I need to remember to order some more next time I have a Digikey or Mouser order.
- the whole reason I bought it, since it has a trigger time of 50µs (that’s 50 micro-seconds, or 0.00005 seconds). Meaning it only has to see continuity for that long before the beeper sounds. So you can quite literally sweep along an IC pretty fast to find which pin may/may not have continuity to your test point. Thankfully, there’s also a pulse stretcher that means even if you only have continuity for those 50µs, the beeper will sound for at least 200ms. [↩]
- And yes, theoretically the 8060a has a faster beeper, but… okay, fine, I’m a meter collector, and I couldn’t turn down an 8021B for $20. [↩]