We are measuring a significant current flow between master and slave inverters on the AC load terminals. With the current test setup - it's about 4 amps (clamp meter measured) on L1 to L1, and L2 to L2, and nothing on N. This is with zero load connected. I have been working on this for a few weeks already. Just received a replacement 18KPV for one of the pair - but seeing the same problem still.
Parallel with 2 18KPVs. 4x wall mount AW batteries shared. 120/240 v split phase. Inverters seem to provide usable output power - load testing so far is OK, but noted relatively high THD along with surge-related drops that will need attention. This problem was noted with other configurations, without solar or grid attached, and even partly demonstrated with a single inverter setup. The present setup has a small solar input, a small (40 amp) grid feed, and is a full parallel configuration again with the new 18KPV.
Here is a monitor site screen while charging batteries from grid at night. No loads attached, but the circled "Consumption Power" changes constantly. With no loads attached - values range from zero to 200 watts by the second.
Current readings on the inter-inverter L1 and L2 are fairly stable when the inverters are loaded the same way. A sample reading:
During this test the batteries were charging from the grid at a steady rate - 120 amps DC collectively. Changing the current, or stopping the charge, changes these measurements.
Now new data has been found. I am wondering what this may mean, if anything? These numbers dance just like the LCD and monitor site readings - and the load power TOTAL is about what the EG4 monitor site shows.
The bouncing consumption power reading that started this investigation may be a net result from the inter-inverter current interaction. The wash affect of the two load power readings (one negative, one positive) agrees with the manual clamp meter amperage checks.
Is there a setting I should check (the distributor has reviewed)? Perhaps I miswired something (every cable has been checked and replaced, and anything optional was removed at some point)? The distributor's elevated tech support sent the replacement 18KPV that didn't change things.
Ideas?
