CMOS Battery rebuild amendment - amended procedure for cheap DMMs, pre tabbed

ID: 218749

Description:

Steps:

1. No meters worth keeping were destroyed for this guide to show why I did this — the large LCD unit already broke prior and the MCH-108 has a false CAT rating but was not destroyed.
2. Blacklisted meters:
3. iFixit sold: MCH-108/any MCH meter - Reason: Calibration lottery — my 2 from two meters from the RBT kits show 5.012V (CatIII), 5.003V (CatIV); Digital Multimeter IF145-035-1 -- Reason: DT/XL830 class meter; manual ranging 3 1/2 digit average responding.
4. 3rd party: DT/XL830 -- same as iFixit IF145-035-1; UNI-T -- Reason: Slow testing, possible calibration lottery where some are out of spec and not properly calibrated, multiple meters with exaggerated Cat ratings; Harbor Freight 7 function DMM -- reason: 830 class meter.
5. The reason I am putting these meters in their own group beyond slow continuity testers, iffy factory calibrations and accuracy issues? Safety. I tore one broken unit down (don't worry, $10 in 2022 money) and a working MCH-108 from my RBT to show how many shortcuts get taken with these and how dangerous they can be to use beyond DC LV work.
6. These meters also frequently lie about which functions use True RMS (TRMS) - even the iFixit MCH-108 is TRMS AC, Averaging DC -- this is never disclosed.
7. Fake CAT ratings are normal: Even the iFixit MCH-108 is Cat IV on the meter, but has none of the features of a true Cat IV like a Fluke 174/3000 CNX or Fluke 3000 FC.
8. China Export: Most of these have the "China Export" mark, not the true CE mark.

9. The reason I have quarantined these lower tier meters to this guide is they tend to be all over the map for accuracy. To show how variable these get very quickly I put a fresh CR2032 against 4 meters. These meters will work but you need to account for any errors which may occur on them as much as possible.
10. Fluke 115 -- 3.306V DC
11. Fluke 3000 FC -- 3.307V DC
12. iFixit CM-D20 -- 3.309V DC
13. AstroAI DM6000AR -- 3.294V DC

14. This amendment is limited in scope to the following DMMs:
15. Any meter which has questionable accuracy that would fail in the non amended guide
16. Cheap meters from Amazon and Harbor Freight that are built around shortcuts, especially older HF meters where safety was an afterthought in the design
17. Super low cost meters built to a price, such as the ones from Amazon, Walmart, Harbor Freight, and super cheap ones from Ace Hardware

18. Before you start, if you are running the no name pack-ins, discard these batteries and install a name brand pair. Most of the time these batteries in-box are aged as they were sitting for a while and should be replaced to avoid false positives.
19. Discard these batteries
20. Install good name brand batteries (Alkaline or 1.5V Primary Lithium both work) before continuing with this guide

21. Authors note: The minimum recommended meter I would suggest using for this guide (if possible) is 6000 count - a reliable 4000 count meter is fine, but you will need to be more particular about the meter's actual accuracy before using it.
22. As the disaplayed voltage may be inaccurate, inconsistent or set too high to go off on many of these meters, you must test the cell over a longer window.
23. After sealing the cell, you must hold your test leads on the connector for 1-2 minutes and watch the voltage
24. Unstable voltages/regular voltage drop: This is a dead short. Disassemble the cell and start over. The indicator for this is a <0.1-0.2V drop from initial cell testing after applying tape.
25. Unstable reading - last 1-2 digits: Unstable or inaccurate readings (as shown with my DC reference - 10V) indicate a low-quality meter, poor calibration, or a failing coin cell. If available, verify with a voltage reference. If it persists, replace the meter with a higher quality True RMS model (ex: Commercial Electric/HD or Kobalt/Lowes).
26. Minor errors: This is fine as a lot of sub $100 meters use ADCs with issues on the last 1-2 digits (often known insignificant digits). Ignore these last 2 and focus on the first 2 digits that matter (3.X). As long as it is otherwise accurate these last two are rarely accurate to begin with in this class.
27. If you suspect the calibration on the meter is suspect, use a reference like this to quickly verify the meter. If the drift is similar, your meter has a horrible calibration and the rebuild might be okay. A fresh 9V battery will work as well, but it is not a reliable reference grade source.
28. If the meter is really wonky and you have a voltage reference like the example shown, test it to see how many digits off are expected vs a build issue.

29. After confirming there is no assembly short, let the cell sit on your bench for 3-5 minutes before heatshrink sealing and installation. As long as it does not get hot after being allowed to cool to room temprature, it is safe to install.