Thursday, September 10, 2015

For Want of a Nail

So near yet so far.

Yesterday was focused on completing the battery pack area in the rear under the hatch. There was a good bit of fabrication involved and I'm really pleased with the results. 

First, the Nissan master disconnect/fuse unit was installed on the left of the pack.
 Then the DC/DC Converter control box with fuse and AVC2 charge control on the right. The red circular device on the edge is an inertia switch to shut off power in the event of an impact. The little red thing sticking out behind it is a switch for the enable signal. I connected the normally closed terminal of the AVC2 relay to the enable line of the GEVCU to prevent driving off with the charge cord attached. Turning off the enable switch will prevent the car from running. It's a rudimentary security feature, and our little secret. 

There is more hiding beneath this shelf: the mid-pack contactor and a terminal strip to connect the J1772 charging port with the charger and controller. The spaghetti will be out of sight down here.

Viewed as a whole, I think the battery bay looks tidy and businesslike.

With all the cabling in place, it was time to power up. Just as I turned the switch, there was a clap of thunder outside and I couldn't help thinking ...

All the pumps were pumping and pack voltage read correctly when turned on, then showed zero when off. Charging started up when the J1772 was plugged in. All good. So I went off for a late celebratory supper with plans to finalize the configuration in the morning and maybe take the first spin around the block.

Alas, was not to be. I calibrated the throttle and power cycled the GEVCU, but noticed that my multimeter was still showing full pack voltage when the system was turned off. Not good. Last night it was zero, now full pack voltage and dropping slowly, eventually getting into the millivolts range. It occurred to me that I was watching the capacitors in the DMOC drain which must mean that the contactors were still closed. Something awry with my precharge routine? I pulled the rear contactor and it checked fine.

Then I pulled the Better Place contactor module and found that the contactors were both locked in the closed position. Contacts welded shut! The GEVCU manual has a full chapter on precharge considerations, and my setup worked as it should last night. What happened?

Here's the culprit: sorry it's out of focus, but you can see the chunk of insulation that's missing. This is the common +12 volt connection for both contactors and the precharge relay. The GEVCU manages the startup sequence by opening first the precharge relay ground line, then the main contactor's ground after the capacitors have reached pack voltage, avoiding the inrush surge current that welds the contacts closed. What changed?
The yellow circle highlights the connection for the wire above. This particular wire was a bit too short to connect with the +12 volt terminals at the left of the box, so I connected it to the normal input side of the diode. Good +12v there, but a rather awkward placement - notice that it passes over the top of the output terminal on the charger side of the diode. I must have nicked the wire when I installed it, and then the bare wire made contact with the charger power output screw. When I plugged in the J1772, power passed through that screw to the contactors without the GEVCU managing the connection. Result: inrush current and welded contacts. I've ordered a fresh contactor, precharge relay and resistor. UPS should deliver them early next week, so all is not lost. Unless, of course, there is collateral damage to the GEVCU, but won't know that until later.

For want of a nail the shoe was lost
For want of a shoe the horse was lost.
For want of a horse the rider was lost.
For want of a rider the message was lost.
For want of a message the battle was lost.
For want of a battle the kingdom was lost.
And all for the want of a horseshoe nail.

No comments:

Post a Comment