JD's VW Bus Conversion EV project (2 of 2)

[TylerDurden]

I would bolt together a beam from square tube. Gives you lots of options on revisions.

You could also explore "universal" or "multi-fit" trailer hitch framework...

[dogman]

That's true, some square tube, a few bolts, a piece or two of angle iron..... Maybe no need to weld at all. Just use good grade 8 bolts.

[oatnet]

booster_0842.jpg (70.43 KiB) Viewed 152 times

Working hard, I haven't had a chance to post.

Above is the vacuum booster assembly for the brakes I completed a few nights back. It is mounted in the wood box that I got my 10ah cells from Andy/Falcon Ev during the 2007 Alt-Car-Expo, fresh off the boat from Taiwan. I think these are the same cells Justin at ebikes.ca is selling, they have served me well.

On the lower right is the actual pump and oil-filled muffler. The black cylinder on the left is the vacuum reservoir, on the hose is a check valve (left) and the vacuum sensor. Mounted to the left wall are the relays that monitor the sensor and turn on the vacuum pump.

I bought this kit from http://www.evsource.com/tls_braking_system.php for $366 delivered. Noise is reported to be a concern with these pumps, and this is touted as being quiet. I powered it up briefly and it did run quietly, though it sucked so much juice that it melted my test leads. I look forward to mounting this, I'll have to see how much juice it consumes, I may not have budgeted power adequately for it.

========================================

I bought the 20" x 14.75" x .375" aluminum sheet (seen below) on ebay for only $37 shipped. I also got a 16 "x 48" x 0.060" sheet of Phenolic plastic from the same vendor, both local to me so GREEN shipping.

Holes drilled for the controller, and markings for the rest of the parts laid out and ready for drilling.

1-predrilled_0851.jpg (62.23 KiB) Viewed 152 times

Holes drilled for everything and all parts test-mounted! Whew - everything fit.

2-test-mounted_0852.jpg (54.41 KiB) Viewed 152 times

Phenolic plastic cut to shape and a hole cut for the controller.

3-phenolic-0853.jpg (67.12 KiB) Viewed 152 times

Finished assembly.

4-finished-_5870.JPG (69.63 KiB) Viewed 152 times

On the left is a Vicor Power supply. I bought this CHEAP when Doc Bass recommended them a ways back, but never used it - and now the mount that came with it is very handy! I am actually using a different vicor PSU than the one in the picture, but this is good for a test mount.

The Vicor's come with slots for (8) independant PSU that can be any combination of v/a, combined in series or parallel. I will fill (6) slots filled with 48v4.2a PSUs to charge the main pack. These can each be trimmed from 30v to 57v, giving me all sorts of combinations. For example:

I can parallel (4) of them to charge my 48v80ah pack at 16.8a, @0.2c.
I can do 3p2s to charge my future 96v80ah pack 12.6a.
I can do 2p3s to charge a potential 144v80ah pack at 8.4ah (@0.1c)

The remaining 2 slots I have filled with 12v 4.2a PSU that I will use to
a) charge the 12v pack that runs the brake booster, accessories, controller logic etc,
b) drive the EV200 contactor in the upper right, that closes the circuit on 48v-180v main-pack charger so that line is only active when the PSU is on.
c) open a relay that breaks the circuit on one of the motor-power contactors, so that if the charger is on, I can't power the motor.

Elegant, eh?

So the rest of the stuff... The Kelly KDH14500B controller is pretty obvious. On the left of it is the 500a/50ma shunt where I will land the 4/0 cable that supplies V+ from the battery. That will connect to the white EV200 Contactor in front of it via copper strap. This contactor's actuator leads will be hooked to an emergency cutoff switch up front. The 4/0 cable to S1 will connect to the other side of the contactor, as well as a copper strap that connects to the controller's B+ spade. The Controller's B- contactor will connect to the 600a kelly contactor on the left, which will be activated by the ignition.

I figured a mix of contactors would be good, something that kills one might not damage the other. Although the aluminum plate is needed to cool the controller (and a handy sturdy mounting platform), all that exposed conductive metal was making me nervous, that's why I put down the phenolic plastic with a hole cut out for the controller.

Oh, also note the 5 aluminum tubes/steel bolts in the corner and center. I am going to use them to mount a 1" plywood sheet to the assembly. This whole assembly will then flip over (wood side down) so that if it is ever accidentally exposed to water, it will drip down to the wood instead of pooling. I will also mount the throttle assembly/secondary return spring to this piece of wood, the subject of future photos.

Ah, the throttle. Friggen Fernando also CLIPPED my friggen throttle cable, so I don't have much to work with. Arg.

Anyhow, next I have to clean the engine bay, use the copper strip in the picture above to make laminated straps to connect the contactors/shunt/controller, make another laminated copper strap to connect S2/A2 on the motor... Then some wiring, and the whole mess goes into the bus with the brake vacuum pump and I should finally have reasonable control of it. Last bits will be to run an armoured casing with a bundle of wires to the front to connect my JUMBO CycleAnayst, and some control bits (emergency cutoff, brake sensor, etc) and the conversion will be complete - well, until I add more batteries that is.

Then I have to replace the burned wiring to the taillights etc and I can power up the bus's main 12v system for lights and signals. Then I have to repair a torn cv boot. Ah shoot, I forgot about the motor mount.

EDIT 12/14/12: replaced pictures lost when the forum crashed a few years ago.

[Malcolm]

JD
I'm impressed with the progress you've made on this project in such a short time. I'm in the late planning stage of a conversion myself and a little daunted by the work ahead so it's great to read your detailed record and see all the pieces coming together.
I have a question though. I see you have three contactors – two Kilovac EV200 and an Albright. Most people seem to recommend one on the positive side and another on the negative side so that you can completely isolate the pack, but what's the third one for? Is it just added redundancy? Nothing wrong with that of course, I'm just curious.

You've packaged brake servo pump very neatly there. I realise that's partly to muffle noise from the pump and you've probably already thought this through, but isn't there a risk the pump could overheat inside that box if it draws a lot of power?

[oatnet]

JD
I'm impressed with the progress you've made on this project in such a short time. I'm in the late planning stage of a conversion myself and a little daunted by the work ahead so it's great to read your detailed record and see all the pieces coming together.
I have a question though. I see you have three contactors – two Kilovac EV200 and an Albright. Most people seem to recommend one on the positive side and another on the negative side so that you can completely isolate the pack, but what's the third one for? Is it just added redundancy? Nothing wrong with that of course, I'm just curious.

You've packaged brake servo pump very neatly there. I realise that's partly to muffle noise from the pump and you've probably already thought this through, but isn't there a risk the pump could overheat inside that box if it draws a lot of power?

Hi Malcom!

Thanks for the good words!

I fully understand the 'daunted' bit - looking at the total job is stressful, but breaking it into subtasks - "today I am gonna complete xxxx" helps a little bit. Honestly, each of these subtasks has proven doable with my limited skills. Taking an intemediary stop of mounting things roughly/quickly, seeing it work, and remounting them as I now see I need them, has been a big help. I don't need to anticipate everything and get things perfect the first time, because I'll figure it out in the next design; I don't need to worry about reworking something I spent a lot of time on, because the first build is so crappy that I know I am going to have to anyhow.

Good point about the pump. This was a nice modular way to package it, until I decide a better mounting would help. I have seen a few boxed installs (of different kits ), and didn't give it another thought. Open to the top, heat rises... I'll have to monitor it and see. Hopefully I find the run time is intermittent enough that it doesn't suck juice and doesn't heat up.

The 'Albright' contactor is actually a 600a (12v coil) $89 from Kelly:
https://www.kellycontroller.com/shop/?m ... uct_id=456

The third contactor (back right) is for the charging circuit, not the motor power circuit. From the Vicor PSU, I'll run the positive charging wire to the back of the shunt, and the negative charging wire to that third EV200. The other side of the third ev200 will connect to the Kelly Contactor (battery side). Without the 3rd contactor, this circuit would be 'hot' to the battery all the time, a hazard at worst and a potential battery drain at best.

In addition to the 48v-180v charging output, the Vicor has a 12v output that I will use to drive the 3rd contactor. When the vicor it is powered on to charge, it automatically activates the 3rd contactor, which completes the charging circuit. If the PSU gets fried, the 3rd contactor loses power and breaks the connection. An ev200 is probably overkill for the load, but they are rock-solid, cost me only $63 new, and I had a spare (originally intended for my Comuta).

Oh, I am also planning to have that vicor 12v circuit power a normally-closed relay, and put that on the circuit that drives the Kelly contactor's coil - so when the vicor is powered on, the NC relay breaks the Kelly's coil circuit, and the Kelly Contactor breaks the motor power circuit. Effectively, the contactors/relay will swap out the controller and the PSU on the battery circuit so they are never connected at the same time. That way I con't forget it is plugged in and drive away.

While pack isolation is a nice by-product, I am using (2) contactors on the motor power circuit for redundancy and diversity. The coil electronics on the ev 200 are nicely contained, almost waterproof, but when used at 500a levels you are supposed to have it connected to 4/0 cables to dissipate heat so I gather that might be beyond its design usage. The Kelly contactor is not hermetically sealed, but it is a big beefy beast built to handle 600a. (I also bought their $69 400a version for use on my Comuta-Car project, which is identical save for smaller contact points) So if water exposure kills the Kelly, I can bypass it and limp home on the ev200. If the ev200 gets smoked by the load, I can bypass it and limp home on the Kelly. If the controller fails full-open (and the fuses do not blow?), hopefully one or the other will be able to break the circuit.

-JD

[Malcolm]

Sounds like you've got every angle covered. I didn't know that about the EV200 needing additional heat sinking when handling higher currents. I just picked up a couple of them myself, so I'd better read the data sheet... I'd guess they're talking about 500A continuous though, so there should be some headroom there.

[oatnet]

Can anyone tell me, if I put my multimeter on the ends of an individual cell, will I read the voltage of that cell, or the voltage of the entire supercell?

Yes, you can measure the potential (voltage) across any individual cell, regardless of how it is connected to other cells. Just jump in with a multimeter.

You can also charge or discharge any individual cell in the same fashion, while connected to the others. The circuit (say the multi-meter with probes on each end of one cell) is only 'closed' around that cell, so that's all it sees.

Hey voicecoils - FYI I knew this answer didn't fit with my parallel charging experiences, so I asked the question again on another thread, and I got the opposite answer from Fechter:

viewtopic.php?f=14&t=8733&p=133209#p133204

-JD

[oatnet]

Sounds like you've got every angle covered.

I drove myself NUTS considering angles when I was in the design phase. Unlike an ebike pack that I can rip off the frame, If something goes wrong, the whole bus could go up. I tried to consider what was in any way possible - if I wanted to sabatouge the system, how could I do it, and how could I design the system to recover from it? In my ignorance, I have probably attempted to cover far more angles than are likely to occur, I just hope I haven't missed any critical ones! .

I didn't know that about the EV200 needing additional heat sinking when handling higher currents. I just picked up a couple of them myself, so I'd better read the data sheet... I'd guess they're talking about 500A continuous though, so there should be some headroom there.

Yeah, 500a continious should far exceed my loads. Since I have read that brushed controllers fail open (full battery current!) I also think about them handling/breaking that load (mfr says it will break 2000a once). Of course, one of the silicon fuses _should_ blow at that point, but I have made a career of underestimating shoulds.

Anyhow, the source of my rumor-mongering about the 4/0 cables: http://www.evsource.com/tls_relays.php
"4/0 Cable connection(s) made to the terminals of the contactor to help dissipate heat from the internal contacts
Battery current limited to 500A or less"

Datasheet:
http://www.evsource.com/datasheets/tyco/EV200_ds.pdf

-JD

[wrobinson0413]

.

[dnmun]

JD, i would suggest an independent muffler shop to help you make a brace for the back end of your motor. they can make brackets to bolt onto the frame and then bend tubing and cut it to weld onto the bracket plates so your brace would be removable. maybe wait and figure the big picture about the remaining stuff, like the heating ducts running forward, useless gas tank, and the useless starter or other stuff still there and where you may need to run your cooling ducts, or place the batteries or controller, and then let them fabricate something for you while on the lift. this means you will have to be comfortable with standing there and telling them how to do it while it is on the lift and the big problem is finding someone who will let you even be on the shop floor much less under the lift. think discount mufflers who take cash but still can speak enuff english to get the job done, but you can do a lot of it yourself (communicating) with sketches and cardboard mockups on the spot

[oatnet]

I am on 'vacation' this week so I have been busting my hump on the bus from 7:00 am to 9:00 pm - lots of work, took me lots longer than I expected, but well here it is:

EDIT 12/14/12: replaced pictures lost when the forum crashed a few years ago.

0 - I made a 1" wide copper strap to replace the 4' 4/0 cable running from the S2 to A2 studs on the motor; 5-ply was all I could fit on the studs. It is wrapped in 3 layers of black heat shrink, sealed at the ends, and holes cut out for only the contact patches on the motor.

0-copper strap and thermistor_5898.JPG (66.56 KiB) Viewed 1613 times

1 - I finished the assembly with copper straps (each one took @1.5 hours to make), but when I got up the next morning and started looking at hooking up the CycleAnalyst leads, I realized that there was a reason I put the shunt on the negative side in my temporary setup... The CycleAnalyst needs to be downstream of the power connection. Duh. I HATE making modifications to a finished piece.

1-oops_5879.JPG (63.98 KiB) Viewed 1611 times

2 - shunt moved, Kelly Contactor moved back to accomodate it, and two new copper straps made.

2-fixed_5880.JPG (60.54 KiB) Viewed 1610 times

3 - Here is the Vicor PSU in its mount. In top slot is the 24v supply that drives the contactor for the 48-180v charging circuit. 2nd is the 12v supply that will charge the 12v battery for the 12v circuit the bottom (6) slots are filled with (6) 48v4.2a supplies for charging the main battery. These supplies run at 30v-57v, and can be combined in parallel or series, which is what gives me the ability to charge 48v-180v. Currently I am on only using 4 of them in a 4p configuration, giving me 57v/16.8a. 16.8a is a little over .2c charging rate, not fast but gentle on the cells. In the near future I will compliment this with a string of single-cell chargers for balancing.

When I add 14 more supercells to bring the pack up to 90v, I'll make this a 2s3p configuration, which will charge at a 12.6a rate - slower, but relatively lightweight onboard charging.

3-vicor_5881.JPG (88.46 KiB) Viewed 1609 times

4 - The assembly complete save for the 14' 4/0 battery cables, which I mounted later because they would make it too heavy to carry.

4-completed_5883.JPG (88.9 KiB) Viewed 147 times

5 Looking through the rear deck to where the gas tank used to be, and the controller is about to live. Look a little right of center, and you will see the factory-made hole where the mounting bolts will go to secure it to the chassis.

5-the hole_5884.JPG (54.54 KiB) Viewed 147 times

6 Now the battery cables are on. Also note the wire-tied runs of red, white, and black wires for the CyCleAnalyst, each soldered to a polyfuse. This assembly will be flipped upside-down and rested on the mounting board.

6 ready for board_5885.JPG (74.83 KiB) Viewed 147 times

[oatnet]

7 - The UNDERSIDE of the mounting board, super-thick plywood. Note the (5) bolts that will go through the hole in the gas tank area (see picture#5 above). Also note the 5 metal washers where the bolts go for the posts that support the aluminum plate. On the upper left of the board, you can see some lines I scribed that show the natural path of the throttle cable, and screw holes for mounting the PB-6 throttle to take advantage of that path. I have a 2nd (redundant!) throttle return spring, that will be connected to an L-bracket hanging over the end of the board. So, this board serves many purposes!

7 underside of board _5886.JPG (66.42 KiB) Viewed 147 times

8 - This is the top of the board. I cut aluminum tube into (5) pieces 4.5" tall (1/2" taller than the Vicor) and ran a bolt through each one. I put a metal washer at the bottom of each tube to spread the weight of the aluminum plate and keep the tubes from cutting through the wood. Also note on the left side, the heads of the 5 chassis mounting bolts and their supporting washers. The line on the right side of the board delinates the end of the aluminum plate - the rest is overhand for supporting the PB-6 throttle.

8-top of board_5887.JPG (53.17 KiB) Viewed 147 times

9 - The Aluminum plate is now mounted to the board, the chassis mounting bolts hanging off.

9-controller mounted to board_5888.JPG (51.18 KiB) Viewed 147 times

10 - Top View of plate mounted to the board. Heat rises, so I expect heat to rise up from the controller, through the thermal paste, and into the aluminum sheet. With 20x14 of radiating surface, and the mass of .375 thickness, I expect to draw a fair amount of heat from the motor. This would be inadequate for a zilla controller, but with a continious rating of only 200a this should be OK. If it needss more cooling, I might add one of those giant Peltier coolers being sold on ebay.

a10-more controller mounted to board_5889.JPG (65.09 KiB) Viewed 147 times

11 - The whole assembly mounted. Note the throttle cable dangling over the right side of the motor, waiting for the PB-6

a11-installed_5900.JPG (72.93 KiB) Viewed 147 times

12 - A second strip of plywood under the gas-tank hole secures the mounting board to the chassis. The bolts pass through both boards, with the Bus' chassis clamped between them.

a12-underside mounting_5897.JPG (57.92 KiB) Viewed 147 times

Weight I removed:

- Engine/exhaust - 275 lbs
- 12 gallons of gas: 96 lbs
- Gas tank - 20 lbs
- 12v Pb battery - 22 lbs

- 413 lbs so far

-(future) starter/solonoids - 25 lbs
-(future) Fuel pump and fittings - 15 lbs
-(future) Asst heating system parts - 15 lbs

Weight of the parts I put in:

+ The aluminum plate weighed 11 pounds, and with everything mounted to it (Vicor, Controller, contactors, and wiring) it weighs 36.8 lbs
+ The 4/0 cables (2x 14', 2x 4') weigh 27.6 lbs
+ The mounting board, mounting bolts, and riser tube assemblies weighs 11.2 lbs
+ The motor weights 143 lbs.
+ The 128 batteries weigh .75lb each, so 96lb total.

That is 314.6 lbs total so far; when I add in (14) more supercells to fill the battery tray that will be 84 more lbs, 398.6lbs.

+Oops, haven't weighed the vacuum booster yet. That will add weight, but I'll easily still be lighter than stock.

I'll have to add the motor mount to the tally, when I get it made, and compare that to the Moustache bar (30lbs) that suported the ICE and is not included above, I expect to reduce weight there.

Things not added: I don't know how much the wood in the battery tray weighs. I did not weigh the fuses, supercell joiners, or the copper straps at the ends and between the rows, or the S2-A2 copper strap on the motor.

OTOH, there were a lot of misc parts that came out, all of the tin (airflow guide), tons of little random parts.

So my bus, at 90v, will definitely weigh less than it did with the ICE. I wonder how a it will perform compared to a 144v Lead-Acid Conversion that adds 1500 lbs to the stock weight... That is the ultimate answer I am looking for in this experiment.

Still to come:

Throttle
J1/J2 wiring harness
Vacuum booster - i need to convert 3/8" to 1/2" - arg.
Maiden Voyage!

-JD

EDIT 12/14/12: replaced pictures lost when the forum crashed a few years ago.

[Doctorbass]

Nice work Oatnet!!

Lucky you are to realize that project!! i'm still on two wheels

i'm following your project !

Doc

[markcycle]

Wow looks good can't wait to see the maiden voyage. Got to admire a guy who goes on vacation so he can work his ass off on an EV project.
Mark

[oatnet]

Wow looks good can't wait to see the maiden voyage.

Mark, your wish is my command!

I got an inner ear infection on that last day of work, haven't been able to get near it to solder the last three wires until today...

On the maiden voyage, I didn't want to go very fast because the brake booster isn't hooked up yet (I need a 3/8" to 1/2" hose adaptor) and until I repair the burned wiring on the taillights I can't hook up main system power - brake lights, turnsignals, etc.

Changing gears was easy - THE CLUTCH WORKS without a pilot bearing! It pulls away in first gear better than the old ICE did, but it felt like it was running out of ooophm at the magic 25mph mark. OTOH the ICE was slow to build speed too, and I didn't take the time to push it faster without vacuum assist for the brakes.

After making the video, I also soldered up my Jumbo Cycleanalyst, now I finally have access to the information I am used to, the data I need to see how the batteries perform. I think I'm just gonna talk to Victoria and get another batch of cells now, and take it up to 96v since I already have space in the battery tray. I have a feeling I'll get another batch after that and bump it up to 144v... then 180v...

TD, you were right - it was darn near silent inside. I was suprised to hear how loud it was in the video, but I think that is just tranny noise.

It's short, but so is life so enjoy!

-JD

[Dee Jay]

One man wave ~~~~~~ whoo ! ~~~~~

NICE!

I was suprised to hear how loud it was in the video,

Digital video cameras can pick up high freqs sounds really well

J

[headway]

CONGRADUTATION!

Dear Mr.JD,

very pleased to see the video of your electric car with our Headway LiFePO4 lithium-ion battery.

Best Regards
Victoria

Wow looks good can't wait to see the maiden voyage.

Mark, your wish is my command!

I got an inner ear infection on that last day of work, haven't been able to get near it to solder the last three wires until today...

On the maiden voyage, I didn't want to go very fast because the brake booster isn't hooked up yet (I need a 3/8" to 1/2" hose adaptor) and until I repair the burned wiring on the taillights I can't hook up main system power - brake lights, turnsignals, etc.

Changing gears was easy - THE CLUTCH WORKS without a pilot bearing! It pulls away in first gear better than the old ICE did, but it felt like it was running out of ooophm at the magic 25mph mark. OTOH the ICE was slow to build speed too, and I didn't take the time to push it faster without vacuum assist for the brakes.

After making the video, I also soldered up my Jumbo Cycleanalyst, now I finally have access to the information I am used to, the data I need to see how the batteries perform. I think I'm just gonna talk to Victoria and get another batch of cells now, and take it up to 96v since I already have space in the battery tray. I have a feeling I'll get another batch after that and bump it up to 144v... then 180v...

TD, you were right - it was darn near silent inside. I was suprised to hear how loud it was in the video, but I think that is just tranny noise.

It's short, but so is life so enjoy!

-JD

[oatnet]

CONGRADUTATION!

Dear Mr.JD,

very pleased to see the video of your electric car with our Headway LiFePO4 lithium-ion battery.

Best Regards
Victoria

Thank YOU Victoria, for making the excellent, affordable Headway cell; this brings the technology out of the clouds to something everyone can afford to use. Because it can perform at 5c, my 80ah supercells can supply the 400 amps of power needed to drive a 500 amp controller. Because I can discharge them to 80%, I can pull the same amount of watt-hours I'd get from a 130ah Lead Acid battery. Because they are so light, I don't need to lug around an extra 1500lbs of lead-acid batteries, steel mounting racks, or the extra capacity it takes to move that weight. Because they are rated for 1500 cycles, they will last three times as long as lead-acid cells.

The screw-hole end is a great innovation that makes them easy to assemble into a range of shapes - my simple battery tray doesn't even scratch the surface of the possibilities.

If these cells can effectively, affordably, convert a giant vehicle like a VW camper bus, imagine what they would do on a small, lightweight, aerodynamic vehicle...

NICE!

Thanks DJ!

-JD

[GGoodrum]

Very cool, JD. Congrats!

-- Gary

[dogman]

Awesome dude! I can't wait to see how it does for you in the real world. If I understand right, you have somewhere around $3000 in battereies there. It will be very interesting to see how far you go on them, since they are so much lighter than lead. I'm thinking a bug conversion could be affordable if a lifepo4 pack under $5000 could be practical.

[pwbset]

Un-f*&^ing real... as a former westfalia owner that it hands down the coolest damn thing I've ever seen. I will be looking for old scrapped westys now in the hope of someday doing what you are doing! Outstanding work man!!!!

[jorhyne1]

Excellent! I know I don't really post here much, but I have been following this thread from the beginning and I almost get that EV grin just sitting at my computer screen. Can't imagine what its like to be piloting that thing!

[oatnet]

Thanks Gary!

Dogman, with LiFePO4 prices where they are now 5k is way more than you need to build a pack for a bug. What does a standard bug weigh? I think you could also go with a smaller motor - maybe even a 6.7" - so the conversion would end up much lighter than stock. I'd almost like to tackle one, but my wife would kill me.

pwbset, thanks! A westy with a bad engine can be had cheap... I hope you find one!

jorhyne1, thanks for honoring my thread with one of your rare posts, I am glad you have been enjoying it!

I tried charging the bus with my onboard charger; it worked well but the cells are clearly out of balance. So today's project involved setting up the 48v-3.7v20a DC-DC converters. I plan on mounting 30 of these little things - one for each supercell - on the aluminum plate I got earlier this week.

However, they each draw 94 watts, and the Vicor PSU I am using to drive them is rated at 1200 watts at 120v (although I have seen it supply 1400), so until I get a 220v line or two in my garage I can't power all 30, so I am just going to mount 10 for now - enough to balance a full row of 10 supercells.

First order of business was to trim them to 3.65v. I was planning on using a resistor to trim them, until I stopped being stubborn and started listening to Bill (powermed) and used a trim pot instead. It helped that I was able to buy (10) 100kohm trim pots for $2 at the Swap Meet today . Now I can vary my setpoint to whatever I like.

I bought the wrong size drill bit/threadcutter so I decided to just wire up one so that I could balance the cells one at a time.

I was nervous about soldering the DC-DC converters -vicor has warnings about it - but I found that if I coiled a few turns of bare wire around a nail, it was just the right size to slide on the posts and stay put for soldering. It was then easy to build up a ball of solder on my 100w iron, touch it to the turns of wire for 5 seconds, then cool it with a wrench. before soldering, I slid the posts of the trim pot into the turns of wire to mount it.

I trimmed it to 3.65v, connected it to a vicor PSU, and put it in the bus. I burned my finger a few minutes later - that sucker gets HOT - so I rested it on the aluminum plate and it ran cooler. It may take a while to balance all 16 cells this way though <grin>.

-JD

trim pot_6033.JPG (130.27 KiB) Viewed 1347 times

[dogman]

The EV limitation I run into is that I live a bit out of town, and on a hill. So I need about 50 mile range, to drive 15 miles away and get home. And have some reserve for battery life. Nearest store for hardware, or groceries is at the bottom of the big hill. So loading on more lead just makes the hill harder to climb as I found out with bikes. For me, a project like that is at least 5 years from now anyway, since a photovotaic roof is a higher priority if I get my finances squared away. But I trust that from you, I can get real information. Some of the EV bloggers do like to exaggerate the results of thier build.

[oatnet]

since a photovotaic roof is a higher priority

Solar is good preparation for the future, as is any personal power generation. Did you see my post in 'general' on these motorwave microturbines I am trying out? Very cool concept, supposedly effective in light winds, but the company was difficult to work with:

motorwave_6053.JPG (200.47 KiB) Viewed 84 times

Some of the EV bloggers do like to exaggerate the results of thier build.

I really hope to not do that, but I have to admit that I want the weight savings of Headway's LiFePO4 to translate into reduced battery requirements, so I am biased. I am trying to be vigilent and stick to facts; I also want to generate very specific WPM info so anyone following in my footsteps will have LiFe chemistry-specific results to base their builds on, it has been a real challange to engineer this without similar builds to compare it to. Accordingly, here are today's test results I went for an early-morning (pre traffic!) cruise to evaluate my watts per mile. I live in a moderately hilly area, so the results are real-world, not flat-surface. The video is below.


When Justin first introduced the CycleAnalast I bought a beta version of the Jumbo Cycleanalyst (6-wire), specifically for the day I would do this bus conversion. I connected the V+ wire to the battery + contactor. I connected the signal to the controller side of the shunt, and both grounds to the battery - side of the shunt. The shunt is a .50ma/500a, so I configured the shunt resistance in advanced settings as 0.5000 .

Because any of these long 15' 20ga wires could short a huge amount of current and catch fire, I put a polyfuse on all three leads right where they connect. question: Could the polyfuse, and the resistance of the long thin wire, change the resistance and affect the cycleanalyst's readings?

The CA is configured in high-current mode. I am assuming that when I see 50.1 it is really drawing 501 amps, does that sound right? There were a few times it spiked up to 60.x, but my 400a Silicon fuses didn't pop, and Steven at Kelly assured me that the 500a controller would never draw more than 400a, so I think I didn't do something right.

Under peak load the 16-cell array was around 47v, or 2.937v/cell. This corresponds with about 3c on the discharge testing curves for the cells, or 240a on this 80ah pack.

I realized afterwards that I didn't check how many watt-hours the cruise consumed. When I powered the CA back on, I discovered that this version isn't saving data - not sure if it is a beta-bug or my wiring. I went and redid the cruise, this time the bus seemed a bit faster, but the tranny/transaxle fluids I pupt in 6+ years ago were way too viscous, and with the ICE I noticed it was faster after they heated up. I need to replace both of these with a lightweight synthetic.

I adding an extra block for the second cruise to exceed a full mile - and took the start/finish pictures below. If I have the the CA configured correctly, then it turned 100wh right around the 1 mile point. Since it was so slow, I figure this is what it takes to actually move the bus over that distance, and that additional current consumed will be to overcome drag.

On uphills the bus didn't want to go more than 15mph, which was disappointing but I should have expected that from 48v. Now I am wondering if I should go straight to 120v instead of 90v, even though that means building a second battery tray.

-JD

EDIT 12/14/12: replaced pictures lost when the forum crashed a few years ago.

start_6059.JPG (22.57 KiB) Viewed 84 times

finish_6061.JPG (26.79 KiB) Viewed 84 times