August 2008


I’m sending the axle back to the guy that made them.  His name is Frank and he runs http://driveshaftshop.com/

He is going to inspect the axle for damage and reassemble it.  He has given me great customer service and I would definately buy axles from him again.

With my batteries charged, I took it out for its first real test drive. I pulled out of the garage and headed for the parking lot. I gave it a little throttle and it took off quite nicely. I made the first turn and suddenly all the power disappeared. This coincided with the start of a grinding sound emanating from one of the rear tires. After rolling to a stop, I got out to see what happened. My CV joint had come apart at the hub. After a short forklift ride back to the shop, I pull the drive shaft to assess the damage:

It looks like it just pulled apart. I think it is easily re-assembled. However, I’m not sure WHY it came apart in the first place. I’ll need to do some postmortem sleuthing to figure it out.

I disabled comments this morning because I became inundated with spam.  If you want to contact me, just hit the link in the upper right hand corner of this page.

After replacing 50 resistors and finding out that RJ-45 connectors can fail in a very unfortunate way (the terminals inside the socket get bent and short out with their neighbor), all my battery monitors came online. This was exciting, but it was also the point that I looked back and thought about all the things I would like to change. Version 2.0 of the BMS may not be that far off…

In addition to that, I started soldering up the main power system controller. This board will talk to the battery monitors, read the current shunt, and handle all the general data collect for the vehicle. It will report all this info to the MacMini via USB. I haven’t populated it completely yet, but it is complete enough to replace my previous interface to the BMS:

img_2597.jpg

One of my co-workers commented on this setup saying “Why is it that electronics guys always have the jankyest setup for communications?” My reply was, “I needed something quick, and it worked… for a little while”.

img_2596.jpg

Here is the new setup. USB plug on the left and connectors to the BMS boards in the center.

For those interested in building their own BMS, check out http://batteryvehiclesociety.org.uk/forums/viewtopic.php?t=1245

Peter is doing a great job documenting his design process (much better than I have) and the overall design looks quite good. I even stole borrowed a few ideas for my controller.

After spending some time with the battery pack and BMS it appeared that I might have three (out of the 40) bad boards.  The weird thing was that all three were failing in the same way.  I took the boards back to the bench, plugged them in to my bench supply and they worked.  After a bunch of messing around I realized I could make them fail if I adjusted the supply lower than 3v.  Hey wait a minute, I thought I already fixed this problem.   Turns out I have a bug in my “fix”.  I had added a transistor and resistor to bump up the voltage on the LED side of the opto-isolator.  However, I chose too high a value for the current limiting resistor.  I didn’t do the math when I originally tested the “fix”.  I just tried a resistor I had laying on the bench and it worked.  However, it really shouldn’t have.  Now I need to take all the BMS boards off and replace that one resistor on each.  What a pain!

I installed 40 of my 50 batteries this weekend. The last 10 are already charged from my bench testing, so they will go in after I get these 40 charged.

img_2589.jpg

I finished assembling the battery monitoring boards (including the fix for low voltage).

img_2586.jpg

Now I just need to program them and find all the bad solder joints and those darn little threads of metal that seem to plague this board run.

I made some battery connectors similar to the ones Thunder Sky provided with the batteries, but just a little longer. These will connect each string of 9 batteries.

img_2582.jpg

img_2583.jpg

I also started making the 2/0 cables to connect the pack to the motor controller. I hope this weekend to get the pack installed and wired.