July 2009


Here is a charging fixture that I made today to charge 10 A123 cells at the same time. I machined slots into the wood to keep the copper plates from rotating. There is also a 1/2 inch wide slot under each cell to hold them in place. The power supply is set to 3.70v. Each cell can reach its finishing voltage separately from the others due to a one ohm resistor feeding each cell. The other important thing the resistors allow is that a drained cell can be put into the fixture next to a fully charged one. The charging takes a little longer with the resistors, but it gets them all fully charged equally. Then I can load test them one at a time to see what their capacity really is with a 50 amp load. There are currently 30 cells awaiting testing. A group of 10 cells in parallel will eventually find its way into a module that will replace a nicad cell in the #1 truck. A complete pack will only take up the driver’s side half of the battery box. The trick will be to make a BMS board that will fit on top of the closely spaced cell posts like the nicads have.

Here are a couple of links for Thundersky cell load tests. 1, 2 .

The board layout for the 160ah Thundersky/Sky Energy lithium cell BMS is finished. I’d like to scale it down to also fit the 90-100ah cells, but that takes another board layout session since the 160ah cells are so large compared to the 90-100ah cells. For now getting the thermal paths, thermal cycling, and basic operation tested is a higher priority. Making a board that will fit both the 90-100ah and 160ah cells would be the next task. At this point it’s setup to bypass 3 amps. 5+ amps is possible.

The BMS will control the charger output as well as the motor control if any of the set points are hit. I’ll get the BMS to Dolphin interface board going as well.

Here is a paper doll of the 160 ah Thundersky BMS board that I’m developing. Someday I’ll fab a board using the Toner Transfer process. Always a helpful process before paying for a batch.

I ordered 30 used A123 cells this morning. Should be here in a week. I’ve sketched up a diagram of how I can charge so many cells at the same time. I think strapping the whole thing to a 2×4 will work nicely. From the testing I did on Ray’s pack of A123’s, I think they will work great for the scooter and hopefully for the truck. This could be CNC’d and then a lid screwed on. The posts are close together. I don’t know how I’ll get a BMS board across those posts.

Here is a concept sketch of how the A123 module will duplicate the volume of a flooded nicad.

Here is an idea Mike Swift gave me for housing 10 parallel A123 cells in a similar shape to how the nicads are made. It’s tough to see all of the detail on the back of this receipt.

One of the most sobering tests performed by a guy making his own BMS is here. His experiment showed that using the wrong parts can in deed cause a fire. Several designs out there use these same parts.

In another post, this poor fellow had his whole car burn to the ground. What a shame. He did a lot of work on it.

Here is a Prius fire story using A123’s. A bad connection was suppose to be the culprit of that disaster.

This P1 Dolphin board has the Mach220 chip that’s seriously damaged and the main relay fet vaporized from the board. Both instances left some low impedance paths between some traces. The Mach220 removal solved many of them. But the fet traces are damaged inner layers of traces. So the fet will have to be isolated and jumpered back into the circuitry.

All of the power supplies work with the Mach220 removed. So now the debug process starts. With a massive pile of projects on my plate, I don’t know how much more work this board will get until I have more of the projects completed.

This is how the board arrived. The 5v rail powered this Mach220 to death. Amazingly the 5v reg still works.

Here is what a correctly removed Mach220 location looks like. Zero lifted pads or traces. A socket will get installed here.

This fet had to have been jumpered to try and operate the main contactors directly. This is what happens.

The board did not squeal when it was tested again. It works fine now.

The #4 IGBT driver on the 4th board had a bad opto coupler on it. Now it’s in line to be tested in a truck.

Now for the squealer.

Next Page »