Installation of a High Quality Shunt and Check of JLD404 Amp Reading Accuracy

The Amico 200 amp-75 shunt purchased off of Amazon.com is only rated as +/- 1% accuracy.  I used this inexpensive Chinese steel shunt to verify the golf cart current draw runs below 60 amps with spikes to 140 to 180 amps.  The JLD-404 mV range is 0-75 mV.  With this information I ordered a high resolution brass shunt from Galco.com that is +/- 0.25% accurate.  This FN series base-mount shunt manufactured by Crompton Industries in Mexico is a 200 amp - 100 mV shunt with resistance of 0.0005 ohms.  There is no comparison between the quality of the two shunts.  This heavy brass shunt is over 5 inches long and comes with a thick 6" inch long plastic base mount.

Here is a picture of the new shunt attached to the side of the plastic battery box using 2 short flat head screws.  The large brass bolts are the attachment points for the negative battery lead wire (blue insulation) and the black negative wire to the controller.  Some additional leads from accessories attach to the controller side of the shunt so that all of the current will be measured.  Note that the large studs require 3/8" holes which required modification of my cable lead terminals.  I carefully drilled out the large lugs using a c-clamp and a backer board.  The smaller rings on the accessory cables were snipped and pried open to the required diameter.

The pair of small red and black lead wires in the middle of the shunt go to the JLD-404 meter.

Over the range of the JLD-404 meter, at maximum 75 mV the amp reading should be 150 amps.  This value is typed into the meter in order that the current can be calculated and displayed.  Unfortunately, I observed with both shunts that there is a difference between the mV measurements using my Fluke multimeter and the corresponding amp readings from the JLD-404.

Picture of the JLD-404 reading 2.0 amps (cart headlights on and ignition key on)

Here is a picture of the Fluke attached to the same leads wires with the same load.  The 1.2 mV translates to 2.4 amps using the shunt resistance of 0.0005 ohms (just multiply the displayed mV value by 2X or divide the actual mV by the shunt resistance I=V/R).

Picture with the meter leads directly touching the contact points on the shunt produced the same reading 1.2 mV (2.4 amps)

The current flow was confirmed by using the amp meter function and flowing the current through the Fluke.  I measured 2.35 amps.

I am not sure if the meter error is due to round off or low impedance of the JLD-404?  I plan to test at higher current flows.  If the difference exists at all ranges I may have to try installing a high value resistor in series with the meter leads.  For now I might try using a different conversion factor.

Update:  The meter was consistently under recording amps at both low and high current.  I "calibrated" the meter by entering a higher maximum amp setting into the JLD-404.  Instead of 150 amps, I found that 200 amps brings the displayed measurement in line with the Fluke.  I still would like to find and address the root cause.  In the Feb 24, 2012 EVTV episode at 80 minutes, Jack talks about installing a DC-DC converter to isolate the meter.  http://evtv.projectooc.com/index.php?showid=124

Based on this video, I placed an order with Mouser Electronics for a Mean Well SPU03M-12 DC-DC 3-Watt Converter.  This little encapsulated chip with 4 pins will accept 8 to 13 volts input and output 12 volts.  Most importantly it provides 3,000 Volts of DC isolation.  When it arrives, I will have to solder on 2 pair of lead wires and connect between the incoming power and the meter input.  Hopefully this will solve the accuracy problem and protect the pack from additional current leakage.

Quick mileage test.  Using the amp-hour function of the JLD-404 I drove the cart for 2 miles on surface streets which included both starts and stops.  With myself as the only rider, the EZGO RXV with it's 4.4 hp AC-drive motor and regenerative braking got 1.8 amp hours per mile.  Using 80% or 80 amp-hours of usable capacity for my 100 amp-hour cells, this translates into 44 miles on a charge.  This is in line with my estimates during the pack design.