EV: Little, Yellow, Different.
Following the process of converting a 1990 VW Cabriolet into an EV.

A midsummer night’s update

July 13th, 2009 by ahampton

This weekend I still managed to get some time in on the Cabrio.  I finally got the steering column out of the dash – I’ll be working on replacing the dead bearing on that tonight and tomorrow.  The partial dismantling, labeling, zip-tie cleanup, and re-assembling of the electrical system was a success.  I managed to revive the stereo which was shot and keep all of the critical lighting systems operational.  The power windows have been offline for some time so I’ll need to figure out what’s up with them later this week.

Once the steering bearing is replaced I’ll start taking apart the blower to remove the old heater core and probably drop in an electrical one.  Then I’ll work on the old pneumatic a/c controls gutting what I can and replacing them with a black plastic face place.  After that it’s the instrument panel to drop in a digital display for the micro controller that I will have monitoring everything.  I’ll also have some Hi-V warning lights and I have to map out the circuits for the vacuum pump and power steering pump.

–adam

Posted in Uncategorized | No Comments »

More Progress…

April 22nd, 2009 by ahampton

Lots of progress made recently, sorry for the lack of updates.  Something about not wanting to touch a keyboard after work.

I’ve started working with a nice gentleman named Bill Swann who I met at the last Houston Electro Auto Association (heaa.org) meeting.  He’s a mechanical engineer / CNC machinist who has a home CNC table.  He is going to work with me to build the packaging for my batteries.  I’m going to use 38120S cells from Headway (a Chinese manufacturer).  I’ve joined a group buy with the e-bike enthusiasts over at endless-sphere.com in order to help keep the cost of cells low.  Those should show up in early June/July time frame.

In the meantime, I’ve got some big things to get done.  The new motor is in – a NetGain ImPulse9 motor.  All 129 pounds of it (pictures below).  I’ve also got the new adapter plate and transission coupling in thanks to Steve Clunn of GrassRoots EVs out of Florida.  I’ll be working, hopefully with some assistance from friends, over the next few weeks to actually get the motor mounted in the car.  After that I think I’ll be able to fry a couple of batteries on a test drive!

Speaking of batteries, the BMS boards I am building are almost done.  I’m waiting for one more set of parts before they are complete.  They are shown below in their almost completed state. Lots more to type, but I’m tired for the evening so I’ll leave you with some pictures to wrap up the posting.

–adam

Motor and Adapter Plate

Motor and Adapter Plate

Transmission Coupling

Transmission Coupling

BMS Boards (Almost Completed)

BMS Boards (Almost Completed)

Posted in EV Progress Reports | 1 Comment »

A note about LiFePO4 Battery Cells

April 22nd, 2009 by ahampton

I had a conversation with a fellow EV enthusiast about how the Battery Management System (BMS) that I am building works.  I am planning on building a 76.8 volt, 40 Ampere-Hour battery pack as the first pack for my EV.  Our discussion was really educational so I’ve decided to post it here to share with other people.

— snip

Q: So here you have 72 volt x 40 a-hr. Cell count is 24 x 4 = 96 Cells Don’t you need 4 of the x24 BMS boards, for the above configuration?


Nope, only need one of the x24 BMS.  If the cells were Nickel then the answer would be different. Check out the attached diagrams.  One shows the preferred cell connections for Nickel systems, the other for LiFePO4 systems.

Battery Diagram LiFePO4

Battery Diagram LiFePO4

Battery Diagram - Nickel

Battery Diagram - Nickel


The BMS doesn’t connect to each individual cell – it connects to each parallel set of cells.  LiFePO4 is different from Nickel chemistries in that it prefers that cells in a pack be parallel connected before each parallel block is then series connected.  Nickel requires the opposite.  Nickel heats up rudely when in use if one cell in a parallel set has more capacity than it’s neighbor.  LiFePO4 don’t have that problem – they politely balance to the same voltage without much heat dissipation.

When building parallel packs of LiFePO4 care has to be taken to make sure the cells are not at different charge levels before connecting them.  If one cell reads 3.45v on a volt meter and the other reads 2.2v then you should fully charge the 2.2v cell before connecting them to build a parallel block of those cells.  The best practice is to fully charge each cell on a single cell charger before connecting them.  This is how I built my 5p blocks of 18650s.

The BMS in these situations is balancing the charge voltage across these parallel sets of cells.  I can use my BMS on a string of 24 single cells ”(1p)x24s”, a series string of 24 of pairs of cells ”(2p)x24s”, a series string of 24 quartets of cells ”(4p)x24s”, or octets ”(8p)x24s”.

The trade off for the larger packs “(4p)x24s” and ”(8p)x24s” is how long the BMS takes to balance the cells at the end of the charge cycle.  There are two variables here: how out-of-balance the blocks got while in use (because of physical differences in the cells that make up the different blocks) and how much current the BMS can bypass for a fully charged block when it is bringing the other blocks up to the same voltage.  The BMS I am using does a maximum of a 1/2-amp bypass current.

Let me explain how the BMS balancing works with a hypothetical.  Suppose I have a 12v battery pack made up of 4 blocks (exactly like in my LiFePO4 diagram).  Suppose the third block of cells in the batter is “weaker” than the other 3 and after use has a lower voltage.  If I read the voltage across each block I might read values like: 3.31, 3.30, 2.87, 3.29.  The 2.87v reading is clearly the weaker cell.  When charging, the BMS system makes sure that each block is charged up to a “full” 3.69 volts before turning off the charger.  Blocks 1, 2, and 4 will get there first, while block 3 will need more time and current to get up to the 3.69v charge level.

When the strong blocks are “full” at 3.69v the BMS cuts the charging current to 1/2 amp and routes the power around the “full” blocks and keeps applying power to the weak block.  If the weak cell only takes 1/4 an amp-hour to fill up then the BMS will finish balancing the pack in 30 minutes.  If the weak block takes 2 more amps-hours of current to fill back up to 3.69v then the BMS will take 4 hours to finish balancing the pack.  With larger packs the BMS will take longer to maintain a perfectly balanced pack because the potential for each block to be slightly different in capacity is higher.

People in the field who are using this BMS report that after the initial balance they rarely see long balance cycles.  Here is a link to a discussion thread by a guy who is running the BMS with LiFePO4 cells (different cells than I am using, however) on his e-motorcycle.  It has great photos of his assembly process and wiring diagrams.  He is using a product called a “PakTrakr” that is basically a set of volt-meters that record to a file so he can graph the volts from each cell under load.  He also has great graphs of the voltage during charging with the BMS.

— snip

–adam

Posted in EV Progress Reports | No Comments »

Cutting off the gas…

March 2nd, 2009 by ahampton

the gas line, that is!  Last night I was able to disconnect the fuel line from the body.  I had already removed the rubber hosting that connected it over to the gas tank.  I currently have a muffin fan venting on the the gas tank helping push out the remaining vapors and evaporate any gas left standing in the tank.  If I end up completely cutting out the gas tank I want to make well certain that there is no gas left in it when I do.

Once I disconnected the metal fuel line from the where the gas cap connects and from the body it came out as one 3/4m long pipe.  After about 20 minutes with the hack saw I had the pipe cut away from the head that mounts to the body where the gas cap goes.  I found the camera this evening so I actually got to take some pictures.

Here is the gas-pipe head with the pipe cut off.  I have also removed the flap-valve inside the tube.  Note the rough pipe metal around the edges.

Gas Plug head

Gas Plug head

Some electrical tape around the edges should prevent cutting myself against the pipe:

Electrical Tape the Rough Edges

Electrical Tape the Rough Edges

Next I needed to make sure the male electrical side of an extension cord would fit nicely in the pipe.  It looks like there will be ample space for a plug.

Ample Space for a Plug

Ample Space for a Plug

Next I will need to sacrifice an extension cord to power the car.  I think I am going to go get one of those glow-plug ended cords that glows orange when it’s plugged in.  That should look neat to have it plugged in and will also provide a night-light to remind someone in a dark garage that there is a power cord running to the car.  It will also have to be hefty – I want the cord to be able to handle 10 or 12 amps of current while charging.

On the inside of the car I plan on mounting a proper 120VAC housthold outlet in a plastic outlet box just like you find in a wall of a house.  I will probably use a GFCI outlet for an added layer of safety.  I don’t know at this point of I will want to anchor the car chassis to earth ground while charging or not.  I both poles of the traction pack will “float” and not be connected to the chassis.  I will think through it some more and post about it later.

Cheers, –adam

Posted in Uncategorized | No Comments »

The Cabriolet Finally Sees Lithium

March 1st, 2009 by ahampton

This evening I took a (5p)x4s lithium battery pack out to the Cabriolet.  Nothing like a traction pack but it was a fun test.  I hooked the ~13volt battery up where the car battery would normally be connected.  I was able to run the headlights, brights, and tail litghts without an issue.  That’s what I would call a successful test, especially for a pile of battery about the size of 8 decks of cards.  No pictures this evening as I couldn’t find the camera (I’m sure it’s around the house somewhere).

I sold the ADC L91-4003 motor to a local Houston Electric Auto Associate member named Brett. He is working on converting a Karmann Ghia and the motor is much better suited to his scale of conversion.  Best of luck Brett!

The adapter plate for a Warfield Electric Impulse-9 motor has been ordered from Steve Clunnn – an experienced machinist who builds EV parts out of Florida.  He can be contacted at Grass Roots EV (http://www.grassrootsev.com/).  I’m an anxiously looking forward to its arrival.  After it arrives I wil ltake the plunge and purchase the bigger Impluse-9 motor.  That should be lots of fun.

On the battery front my order of 10 VoltPhreaks chargers finally arrived.  As did my order for about 100 High-Current output Headway 18650 lithium cells.  So far my testing has been good.  The Hi-C cells are much easier to spot weld to. I think I have a design for a 40 to 50 AH battery based on, believe it or not, $1.25 sterilite containers that I can buy at the local Dollar General store.  Each container will house a 12.8 Volt, ~50 Amp-Hr battery ( a (40p)x4s cell configuration).  10 of these connected in series should run  the EV just fine.  I still need to get bulk pricing from Headway (the Chinese Cell manufacturer) on how much the full compliemnt of cells would cost.

In other battery news, I have ordered the components for the Goodrum/Fetcher LiFePO4 battery management system that I plan on using.  Some components are on back order (specifically the TC54-2.1v voltage detectors).  The entire lot should arrive some time around March.

On the home front the fence is fixed and the grill works…. now to get the Hot Tub online!

Cheers, –adam

Posted in EV Progress Reports | No Comments »

« Previous Entries