emini

Previously, I reported that my Simovert motor controller was happy in the rear of the car with the motor in the front of the car. It turns out this is not the case. While going up and down the driveway did not uncover any problems, going up and down the road has. The inverter shuts down with a range of errors:

• GateDriveUnit
• LCA Err Latch
• Overcurrent
• VC_Over_Voltage

I am hoping that this is caused by a combination of poor motor phase cable routing and poor motor-inverter grounding. It turns out that the Siemens supplied motor cables have a shield which is grounded in the cable glands, effectively joining and extending the casing of the motor and the inverter to envelope the motor cables. Such a shield will do a lot to control noise radiated by the motor cables. My motor cables had none of this and worse the motor was not well grounded -- I'm told that non-trivial current can flow between the motor and inverter through the ground due to "induction effects". Regardless of how real this is, properly bonding the inverter ground to the motor ground seems reasonable.

I am solving this problem with a 50mm² 3 phase neutral screen cable and appropriate glands. The very stiff underground rated cable I have isn't really appropriate and in hindsight, I should have found a supplier of the right cable, but that story is for my next post.

In 2008 I bought a battery. At the time, I knew that I was buying a product with a shelf and cycle life (it wears out even if you aren't using it). I didn't expect the battery to sit around for as long as it has, but here we are 3.5 years later. I did a capacity test of the 36 cells installed in the car and found a usable capacity of 33.5Ah: The nameplate rating is 40Ah which is substantially more. There are several possible causes of this discrepancy:

• Calendar Life
• Cycle Life
• High current abuse
• High temperature abuse
• Differing test methodology
• Low capacity when delivered

Calendar life and differing test methodology will certainly have an effect. I am charging to 3.55V at 1A which is substantially lower than thunder sky's recommendation. My pack is moderately closely top balanced and at the end of discharge, the highest cell was just below 3.1V while the lowest cell was at about 2.8V at 10A discharge. This suggests the cells have varying capacity or are not properly balanced. I also tested 10 cells which have not been cycled or abused with high current or high temperature. The pack in the car has had perhaps 10 cycles, a lot of it discharging at 3C for 5 or 10 minutes at a time. I limited my maximum current to 200A or 5C and kept currents above 120A to short 5 or 10 second bursts. This is not too far from Thunder Sky's specs but does cause significant battery heating. In the summer I have measured over 40C at the battery terminals. The 10 cell pack that sat on the bench has not been subjected to high current or high temperature, probably never getting higher than 25C. This pack achieved 34.3Ah and showed less spread in capacity, with the hightest cell at 3.0V while the lowest was at 2.85V: That these two packs show very similar capacity suggests the high temperature and high current and cycling had little effect on the battery in the car. Cell voltages were measured with the EVD5 BMS, discharge current and Ah counter performed by my EVision and recorded from it's CAN bus by the BMS data logger. The car's battery was discharged with two domestic 240V heaters and a kettle (which nearly boiled dry). The small 10 cell battery was discharged with a steel wire on a wood form in a large bucket of water. The 10 cell battery shows a larger spread of voltages under load because i did nothing to prepare the cell terminals before putting making the connections, the terminal connection resistance was all over the place.