I replaced the DMOC/GEVCU inverter and controller with a Scott Drive SD100 as noted above because the GEVCU controller was having reliability issues. The Scott Drive proved to be very reliable once the setup was completed. The SD100 has many great features, including a built-in precharge circuit and isolation switch. It provides all the CAN message for the motor performance and has a GUI for setup, calibration and control. The SD100 showed much better reliability and seemed to perform well - i.e. the car seemed to accelerate well. Something that I never did for the DMOC/GEVCU setup was to run the car on a dynamometer to measure the power and torque output. The only benchmark for the DMOC/GEVCU - Siemens motor setup was in a 1978 VW Thing, that was measured to have 160HP and 221 ft-lb of torque. Last summer I found a nearby shop that had several dynamometers and was able to get the 320e tested with the Scott Drive. I was not expecting a great result as I had measured that the Scott Drive was outputting high power, but only for 3 seconds. You can see that in the capture of the GUI screen for the SD100 above. It shows the current and power vs time, but the peaks are very narrow. That is hardly enough time to accelerate the car, although it is something you can definitely feel in the seat of you pants. Unfortunately, the SD100 really is not accelerating the car and was such a problem when testing on the dynamometer that in first gear the wheels would not even turn. The car would just lurch on its suspension and then settle. The Scott Drive shuts down the current to less than half of full current after 3 seconds, even with the accelerator pedal pressed to the floor. The best HP/Torque curve was obtained in 3rd gear, but very is disappointing. See the graph below.
Power torque curve for the 320e in third gear with the SD100.The max power is 128HP and the max torque only 110ft-lb.
This poor power/torque performance was expected based on how the current shuts down from the inverter. There is no adjustment on that parameter, it is built into the firmware. I think Scott Drive does not know how to use their inverters in a car application. The only data I had seen from them was bench testing. That provides a minimal look at the performance. But it seem like their only goal is to over protect the Integrated Gate Bipolar Transistors (IGBT)s. There is no warranty on this inverter. It is more than 6 years old, so why should Scott Drive care how hard I drive the IGBTs? This issue is what convinced me to change to a Rinehart Inverter. The first parameter I looked at for the Rinehart was the current – time limit. Cascadia Motion (who own Rinehart) say in their documentation to limit the full current of 350Amp RMS to less than 30 seconds!! Also, from the documentation is seems like most of the firmware is open – i.e., most of the parameters are adjustable. This should work much better for driving the 320e and the documentation for the Rinehart shows that Scott Drive is wrong about driving IGBTs. The small downside of the Rinehart is there is no GUI for setup and calibration. Those values all have to be entered into the firmware via a GUI with a command line interface. Below is an image of the Rinehart PM100PX inverter.
All of the high current DC and AC wires and the water lines are the same as the SD100. The biggest difference is the analog and logic connections to the controller. The PM100DX has Ampseal 35pin and Ampseal 23pin connectors for all those functions. Ampseal is actually an easier connector to work with than what was on the SD100 and most of the car uses Ampseal connectors. But I had to build a whole new wiring harness and changed most of the wire crossover box connections. The PM100DX is a different size than the SD100 so I had to cut some aluminum plates to make an adapter to the DMOC frame, that sits on top of the Siemens motor. Control of the Rinehart is very similar to the SD100, the major difference is the logic reference. For the SD100 logic was positive (i.e. +5V switching) but on the Rinehart the logic is ground switching. To use the same wires I already had installed in the glove box I had to verify that all the logic wires were isolated so I could just swap the connections. The only wire I had a problem with was the USB connector for phone charging. I had to grab a +12V and ground from under the dash to power that connection.
Unfortunately after spending months on trying to get the Rinehart to work, the inverter failed catastrophically. Check out the article below that is also in the Blog.