- This topic has 0 replies, 1 voice, and was last updated 8 years, 9 months ago by Jonathan Bishop.
- March 3, 2015 at 10:13 am #2463Jonathan BishopGuest
I’ve got an interesting failure mode that I seem to be able to replicate where the reported voltage on the high-voltage channel starts ramping logarithmically right as the batteries are giving out. I’m using NiMH Eneloop batteries (so possibly the voltage curves might be different, throwing off any battery SOC detection logic) with the client software running on iOS 8.
I noticed this this morning while I was using the meter to monitor charging of a 24V lead-acid battery pack on my UPS. The voltage started creeping up at a faster and faster rate from ~25v, and I assumed that this meant the charge was suddenly completing. However, within about 15 seconds the reported voltage then zoomed past 30v (which is getting well past the 15v per-battery “do not exceed” charge of a 12v SLA battery). I started panicing, thinking that something was seriously wrong with the charging circuitry as within a few more seconds the reported voltage hit 70v and the auto-ranging switched to 600v. At this point the client software lost the connection to the meter, and I scrambled to unplug the UPS, thinking that something was seriously wrong.
The client software returned me to the scan menu, where my meter faithfully reappeared. I’m guessing that reverting from sampling mode to sleep mode allowed the battery voltage in the meter to come up high enough that it was able to reboot and start broadcasting its presence again.
I went to reconnect to the meter, and that’s when I noticed that the battery status was critically low, something like 12%. It started showing me voltage readings once again, and started out at 25v, which would be the expected voltage for the battery while the UPS was running off of it. Then, a few moments later I watched as the same voltage spike occurred, and I realized that the problem I was observing might not have actually been with the UPS…
I’ve poked around a bit further in the hour or two since, and I’m able to determine that the strange ‘voltage ramp’ only occurs when that channel is in 60V/600V mode, not 1.2V mode. I’d bet that something in the divider circuit is wigging out when the meter battery drops below a certain voltage.
FWIW, I hooked up a separate DMM to the batteries in my meter and measured ~1.81v at the point where the reported voltage did the inverse swan-dive.
Sorry for the long-winded bug report, but hopefully it’s enough to work out what happened!