Got another two of the inexpensive volt/amp/watt/energy meters from mfreelaa on eBay. These are pretty nice, especially for the $12 each including shipping.
I put two of them in series to see how well each tracked the other. There should be a very small drop in voltage for the second on in series as opposed to the first. That does appear to be the case.
The unit under test is pulling approximately 1-1/4 amps at 12 volts. Assuming these ammeters use 100 A/75 mV shunts (since the meters without built in shunts do the same and it’d be easier to keep them all the same) this would imply that the second meter should read about 0.001 volts lower than the first. However, the display precision is only 0.01 A so it’s not practical to see the difference with this small load.
After two hours now, both meters are showing 50 w-h, I’ll have to wait longer to see if they diverge from one another at all. The voltage (refreshed once per second) is generally within 0.02 volts, which is 0.15% agreement. The current reading is looser, maybe 0.2%. The power reading is quite close, but one would expect that the power reading comes from the multiplication of the v and i values. The energy reading is the time cumulative sum of all the power measurements.
Next project is to measure the efficiency of my new MPPT solar battery charger. Just need to have a sunny day!
Here it is the next morning and the DUT has now pulled a total of either 261 or 262 W-h, depending on the meter read. That’s still pretty good internal consistency between meters, well under 1% difference.
So, relatively speaking, these meters are pretty consistent in their performance. Perhaps one of these days I’ll set up a calibrated load to get a better handle on the absolute accuracy. But for now, just knowing that the meters are consistent allows me to do the solar MPPT charger testing.