This blog documents the restoration, and conversion, of a 1965 Humber (Singer) Vogue to a fully electric vehicle. The Vogue will be powered by an 11kW(modified), 3 phase industrial AC motor, controlled by an industry standard Variable Speed Drive (VSD) or Inverter. To be able to produce the 400 volts phase to phase the VSD will need about 600 VDC of batteries. A big thanks to the contributors on the AEVA forum: http://forums.aeva.asn.au/forums/

Friday, August 20, 2010

Battery Pack Modifications

The Vogue will be powered from 12 of 48V 20AH Battery packs built by Headway of China.
I must state right up front here that I am extremely happy that I bought these packs. The cells appear to be excellent, the Battery Management System (BMS) boards are good quality (WL086-05) and do their job well (with one caveat) - but the pack construction lets it down a bit. I'm very glad that I had to open each pack - more as the story unfolds.

The reason why I have to change the packs is because they are actually designed for an electric scooter or motorcycle. They are not designed to be placed in series to get higher voltage, in fact when I asked Headway if I could do this they warned me that I could not. Headway did make me an offer to change the whole set-up but I elected to "go my own way" as it gave me more control over how the packs would be protected. I took a chance that I would be able to reverse engineer the BMS and be able to make the changes necessary to run the packs in series.

Once I had a chance to thoroughly investigate the pack and BMS it turned out that the reason the packs could not be used in higher voltage systems was totally the BMS. It uses 6 x 140 Amp 100 Volt FETs in series with the -ve output of the pack. If the BMS is unhappy with how the pack is being treated it "opens" the FETS thereby disconnecting the pack from the -P terminal. The problem is that in a 600V system that would place 550 VDC across the poor little 100V FETS and blow them up.

My modifications consists of:
1. Disconnect the FETs and connect the -P terminal straight to the pack -ve end.
2. Modify the BMS to bring out a connection to my BMS monitor board.
3. Add my BMS monitor board.
4. Physically modify the pack to add an LED and a cable gland to bring out the optically isolated wires for my BMS monitoring system.

Sounds easy. It took about 4 to 5 hours per pack. I have done 11 out of 12 at this stage - still no zaps. I have been working on the packs for about 6 weeks - my evening occupation in the cold weather.

Now the Photos...

The original BMS wiring. Two big black wires from -ve of Cells to BMS. Two more big black wires from BMS to -P terminal of pack. Shown here disconnected and taped to stop accidental shorts.
 


The BMS removed.

Again. I'm happy I bought these packs - great Cells and it will make it a heap quicker to get everything going - but questionable pack construction. This is pretty typical of the terminations from -ve and +ve end of cells. The crimps and soldering are always fine but the heat shrink over the lug???

Big black wires removed.

 I lift off the drain connection to one of the two charger control FETs and use it for my BMS monitor. This is the quick way to do it without damaging the circuit board.

I use the lug off the set of wires I discard and make the big black wire set to go from -ve of the cells to the -P terminal of the pack. This is the modified BMS ready to go back into the pack.

I hit this on one pack only. The BMS sense wire had been stretched when the pack lid was put back on. There isn't much room. One of the things I did for every pack was disconnect a lot of the sense wires and re-route them between cells - just to stop this kind of thing. At this point I should indicate the only criticism i have of the BMS. If a sense wire is disconnected it doesn't know. There is still 'some' cell protection as the 2 cells either side of the missing sense wire are averaged but I'd rather a fail-safe design.

Drill 7/16" inch hole for my M12 Cable Gland. You can see the 5mm hole for the Blue LED already drilled.
(Measure 25mm from top lip of pack, 60mm in for LED, 90mm in for Cable Gland).

Then tap the hole with a 12mm tap so the Cable Gland screws in nicely.
This will be where my BMS Monitor board wires will emerge.

My main two tools. Note the heat shrunk10mm spanner. Better safe than sorry.
The Lithium Grease was for the steel cell interconnects after a scare about Headway cells rusting. (I haven't seen any rust yet.)

BMS inserted and all connections made. I haven't greased up the plates yet on this pack. This was an earlier pack. I got pretty anal about not having insulation touching any interconnect plates as I progressed to later packs.

The Blue LED and BMS monitor test. I apply 9VDC to my BMS enable pins and the light goes on. It shows that all is well as the LED is powered from the optocoupler on the 50V side..

And finally seal up the pack and heat shrink the BMS monitor connector.

One more pack to go...

No comments: