Monday, October 26, 2015

Back to the Shop!

You'll recall that there were several contactor casualties from the precharge drama, along with some mysterious memory loss in the charger and DC/DC controllers. First the precharge situation: after frying the second set of contactors, I wired a couple of indicator lights to check the GEVCU precharge function and noted that they BOTH came on momentarily when the system was powered up, then followed the precharge sequence. That momentary inrush of current was all it took to weld the contactors closed. I also noticed during that momentary start-up period that the cooling fan spun just a bit.

At EVCCON, I checked with Collin Kidder who wrote much of the GEVCU code, and he said he had heard of this from several other guys and that it seemed that in the split second it took for the processor to boot up, the mosfets that were supposed to connect to ground under program control, all connected by themselves. Not good.

Nick Smith suggested taking the GEVCU out of the equation by using a time-delay relay to control the precharge sequence. I found this on Amazon that seemed to fill the bill, and picked up two more contactors from the EVTV store. All mounted to a panel, it looks like this:

The contactor on the left closes at startup connecting the negative side of the pack to the inverter. The gold colored precharge resistor then restricts the current flow from the positive side to allow the capacitors to come up to pack voltage at a lower current. The time delay relay closes after roughly five seconds and connects the positive contactor allowing full current flow. An indicator light is wired to the normally closed terminal of the relay so there is a visual cue on the dashboard while the precharge is taking place. All of this works very well and has kept the contactors healthy.

In the controller situation, Jack Rickard had reloaded the flash memory on my DC/DC controller and re-enabled the WiFi function on my GEVCU while we were at EVCCON. I reinstalled the DC/DC when I got home and BLAM! lost the memory again. Jack asked me to return the controller along with the Volt APM. He found that the plug was bad, and built a whole new controller. Thanks Jack - great customer service!

The DC/DC controller came in a new style enclosure that included tabs that made it so much easier to mount.

While all that was going on, I noticed that the charge controller kept losing its settings. So here we have wiped out the flash memory on the DC/DC controller twice, the EEPROM on the charge controller any number of times, and the WiFi function on the GEVCU. What's the common denominator? The USB cable! I broke out a new USB cable, one with gold contacts and perhaps a little better quality, and everything works just fine now. Moral of that story, don't overlook level 1 diagnostics; Is it plugged in? Is the cable good?

Took son Freddy and daughter-in-law Tiffany for a ride around the block and I'm becoming more and more pleased with the way this car drives. I was wrapping up some of the detail work, installing the hood release cable, running wiring for the dashboard, and generally tidying up the motor compartment when Bob called from Pro Automotive to let me know that they had a relatively slow week ahead and this would be a perfect time to get the car in for reassembly of the front bodywork. Called for a flatbed and now the PorschEV is in for finishing.

After a full year of work, this project is coming to completion. There is plenty of tweaking and tuning ahead, but the end is in sight.

No comments:

Post a Comment