I found the root cause of the meter inaccuracy, and it was me

After going through the effort to swap out the JLD-404 meter and test the new one with 3 different shunts, I finally discovered the root cause of the current measurement inaccuracy.  The meter comes with the default current input setting on a 5 Amps scale.  The choices are 5 amps, 1 amp or 75 mV. While being diligent to input the correct shunt value at full scale, on both meters I had left the the current input on the default setting.  All I had to do was change A-Sn = 75mV.  That was it!!!

Now another mystery presented itself.  The old meter would always retain the recorded amp-hours, but the new meter would reset to zero every time power was turned off to the meter.  I asked for my old meter back and returned the loaner/replacement.  It is now installed and all is well.  The current is now within 1% of the actual.  A very small fudge factor in the shunt setting and it is reading spot on.

Meter Replacement

After reading that a batch of JLD-404 meters were reported to have incorrect internal resistors, I contacted ECPC about a replacement.  They promptly sent a new meter which I promptly installed. 

Unfortunately, I am having the exact same issue with the meter reading low on Amps.  In the programming set points, I enter APuH = 150 amps (which is the value when reading 75mV across my 200Amp-100mV shunt).  

 

Here is a picture showing that the meter is reading 1.7 amps while the Fluke is reading 2.2 amps (conversion = 1.1 mV times 2 = 2.2 amps) 

 

 

 

I have to enter APuH = 185 amps to get readings which correspond to my Fluke multimeter and to my battery charger meter.  Here the meters are in sync at 2 amps.

 

I have carefully:

1. Checked the wiring installation.

2. Validated the programming entries

3. Measured the mV with my Fluke multimeter directly across the shunt terminals and also using the same wires as the meter

4. Confirmed that there is no current flowing on the meter positive or common wires connected to the shunt

 

From a practical standpoint, the fudge factor is adjusting for the error, but I would like to solve this mystery.