Tuesday, July 29, 2008

More on charging

I've been busy and have not blogged my progress lately. I'm trying to get a second working battery (which the TWIKE firmware requires for operation) to be recognized. At this point, I have tried two different BMS boards in three batteries to no avail. I'm afraid the BMS is bad in both batteries. Not surprising given the amount of stuff I found growing on them (mold, etc). I'm stuck with either pulling the BMS boards and doing a component level debug or trying to swap for some other unknown batteries. Not sure what I will do. I did get my order from digikey which gives me a set of connectors for all the 3 connector types used by the batteries. With those, I made a set of extension cables so now I can debug the batteries outside the card. Very handy and much safer I might add. If anyone wants the digikey part numbers for the connectors drop me an email (I'm also in the process of setting up a english based wiki for the twike where I can post such technical details for other to find and to share). More on that when I finish getting it setup. I now also have the means to spy on the RS485 bus to decode the BMS protcol. This is all in advance of attempting to build my own Lithium based battery pack (will have to emulate BMS responses to make car firmware happy). I have the battery service program which will give me a let up on decoding the protocol. The program is all in German but I'm part way through in translating all the screens to english. I'll post that on the Wiki when I'm done. So enough for tonight.

P.S. I'm on vacation for the next 3 weeks so I probably won't have any updates for a while.

Monday, July 21, 2008

Charging update

Last Saturday, my son Paul and I spent half a day working on the twike batteries. After several hours of charging, the Battery2 was at 320 volts and slowly climbing (a good sign) but Battery1 was only at 110 volts. A closer inspection revealed numerous cell venting events and the black areas indicated some fires. I assume these happened over the years when the previous owner tried to revive the batteries. I did get one cell to vent myself when I let a group of 20 cells draw 5A for 100 seconds while trying a technique to revive it. Ok, give up on that battery. We pulled it and decided to replace it with one of the spares that has good looking cells (clean ends, no discoloration) but is lacking a BMS card. I proceeded to drain all the cells to zero in the old pack before beginning the process of unsoldering the BMS to move it to the new batteries. A couple of hours later we had the old board out and soldered into the new battery. I put it back in the Twike and powered up and monitored the voltage. The charger/controller is very smart and will only apply a small current until the voltage comes up to about 340 volts. I'll need to get both packs to that voltage to get the built-in charger to give the batteries a full charge (410 to 420 volts). I'll either have to jump start the cells again (will less current this time) or be patient to see if the cells will get to 340 on their own with the charger. Peter Zeller from Switzerland has been very helpful in answering some questions about the batteries and such. At this point, I have no hope to make the battery pack have a useful range, but I just want it to get to the point I can check out all the operations of the Twike. I need to fix anything that is broken before proceeding. I might have a problem with one of the relays since I notice the DC/DC led stays on even after I push down the red button. The docs I have say this is supposed to remove all power except for the batteries themselves. If so, then why is the LED on? Oh well, more to learn.

Wednesday, July 16, 2008

First attempt at charging

Tonight the epoxy on the console repair was ready to I mounted the console and switch panel back into the vehicle. At that point, everything was ready to power up for the first time. I was not sure what to expect as the batteries are 10 year old (Ni-Cad) and have been setting fully discharged for a long time. I plugged it into the wall and lights blinked, the fans came on, and generally it seems to power up. It put some error message in german on the screen and looped on displaying 4 different screens over about 3 minutes. Once I dug out the user manual, I figured out how to get it to speak English. It is complaining the BMS's need new firmware downloaded (yup, batteries were dead) and it was waiting for the 12V to stabilize. I let it run in this mode for a while to see if it would put any energy into the batteries in this state. The BMS in each battery runs of just the batteries so they have to have some charge to bring up the BMS. Meanwhile, I looked through all the settings I can reach easily and found it has only 58 miles on it. I guess that means not too many parts are worn out (except the abused batteries). After an hour of charge, I pulled the 120v plug and low and behold, it kept running for a while. After another 2 hours on the charger, I measured the voltage of both batteries with a meter. With 280 C sized Ni-Cad cells in series, the voltage is supposed to stay between 280 and 410. Battery 1 showed a voltage of only 90 volts. Battery 2 showed a more healthy 289 volts. I have two other batteries (one without BMS that I can take cells from) that with some luck will let me come up with 2 working packs. This would let me debug the rs485 BMS protocol so I can emulate it if I need to (if I go with Lithium). It also would let me try out the motor and all the electrical devices and see if they work. Not sure why it still thinks the 12v is not stable but may be because of the 90v battery. I'm going to try to build a set of cable extenders for the batteries so I can debug them outside the vehicle (much safer and easier). Please be careful if you are dealing with high voltages such as this. I bought some 500v lineman's gloves so I won't get hit with a fatal voltage if I get across the battery voltage while working on it. I'm off to look up Molex connect numbers.

Monday, July 14, 2008

Purchase and transporting to its new home

On Friday night July 11, 2008, I finalized the deal to purchase a used Twike from its current owner. It was hot and the trailer was just the right size so that with a little padding it was safely on its way to its new home. If you ever need to transport a Twike, the 5' x 9' open trailer from U-Haul (the OA metal one) is a perfect fit. I measured it with a tape measure before I rented it just to be sure. I also used their blanket pads and a inner tube to cushion the Twike for the ride. Right now it is resting at the home of the nice folks at Synkromotive in Northwest Portland. They are the hotbed of EV activity in Portland! It's structure is solid but has a couple of problems. The batteries are shot (10 year old Ni-Cad batteries that were not properly cared for) and some minor parts need fixing/upgrading. The first step is to get some sort of working battery setup to see that all the electronics (controller, charger, battery management system, etc) all work. To that end, I will be trying to condense 4 bad packs into 2 working ones. Tonight I glued back together the two wood pieces that hold the display and keypad. I will reinforce it before I reinstall it to prevent the design flaw that caused the problem in the first place. Once that is done, I can reconnect the display and attempt to charge the existing battery packs. More photos in an album at Fotiki. I also scanned and uploaded photos taken in 1998 in Switzerland by Kent H about the time this Twike was built.