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

[pwbset]

Very cool! I may have missed it, but where do you get that copper strapping? All I've been able to find locally is plumbers "copper plated" strapping, which I'll use, but wonder about conductivity losses or whatever.

I also saw last weekend a "new" auto shop in the neighboring town that carries old Type II Westys! They are beautiful! Now all I need is a lot of money.

[oatnet]

Very cool! I may have missed it, but where do you get that copper strapping? All I've been able to find locally is plumbers "copper plated" strapping, which I'll use, but wonder about conductivity losses or whatever.

I also saw last weekend a "new" auto shop in the neighboring town that carries old Type II Westys! They are beautiful! Now all I need is a lot of money.

Thanks pwbset! I went through the same painful search, but I finally found the ebay seller "metallectric" who sells 1/2", 1", and 2" thick rolls. I just went to look him up, but his ebay store says he is 'away' so ebay won't let you even look at his completed auctions. Hopefully he will be back soon, I have wiped out my stock of 1/2" and 1" (see below). I hope you find the Westy of your dreams! I think a splitty would be a very cool conversion, but I think the 72 up versions are the most practical and servicable.

First off though, LiFePO4 Porn - Headway Heaven - (120) new cells arrived today via FedEx. The last batch of cells got here 3 business days after I wired payment, these cells took 5 days, but only because Customs interrupted the shipment in Alaska and SAT on them for two days, I guess maybe they looked like little bombs or something on the x-ray - the last batch sailed right through.

They are lined up like this, out of their white 2-cell shipping boxes, because I tested every cell; most arrived at 3.37v, 3-4 were at 3.38v.

THANKS Victoria, for a great product at a great price point and EXCELLENT customer service. It is always a pleasure to deal with you!

-JD

2 boxes of cells, ignore the reflection.

Headway Heaven 6112.JPG (77.09 KiB) Viewed 3092 times

Here is the last of my 50' roll of 1" copper strap, getting cut into strips that I will laminate together to form the ends of the third row of cells. Note the one strip that has already been marked up, top center, and drill points punched.

end_row_cutting_6108.JPG (51.49 KiB) Viewed 196 times

The individual straps are hammered flat between two pieces of iron, one piece marked, and all of the strips are taped together for drilling. That will have to wait for tomorrow, I am beat from checking all the cells!

end_row_taped_6111.JPG (27.67 KiB) Viewed 1496 times

EDIT 12/14/12: replaced pictures lost when the forum crashed a few years ago, plus deleted (1) smilie since apparently an 8-smilie limit was implemented after I added this post.

[markcycle]

Wow that sure is alot of cells those pictures are almost NSFW because its not good to endlessly stare into the monitor smiling with envy at some point your going to have to explain what your looking at.

I can see that bus burning rubber any day now

Mark

[oatnet]

I can see that bus burning rubber any day now

Now THERE's a thought. I wonder if I can get to that point? That would make a heck of a video... Even if I end up not needing the full 180v the controller can handle, I oughta do it to see if I can do a burnout. Wow.

-JD

[Doctorbass]

Oatnet.. I really like your project!!..

wow all these cells at your home!! I would love having that!!!..

I just think... 100+ hours of intense test!! lol

Be carefull with 180VDC !!.. we want to keep you alive just 100VDC on my arm kick me strong!!!

Alot of Wh!... eee.. may i should replace with kWh !!!

Doc

[Lapwing]

Be carefull with 180VDC !!.. we want to keep you alive just 100VDC on my arm kick me strong!!!

Alot of Wh!... eee.. may i should replace with kWh !!!

Doc

Yep, anything over 48V needs careful attention. Rubber linemans gloves are not overkill.

If you haven't already, this is a "must read"

http://www.evdl.org/pages/plasmaboy.html
http://www.evdl.org/docs/plasmaboy.pdf long version

Perspective is everything, and it helps to learn from others experiences.

I run a 72V system of 220Ah lead acid GC in my boat, and decided to isolate the batteries in 24V blocks. When the 3 contactors are off I have 3 x 24V banks. (easy and safe to work with and charge). Only when the contactors energise is it a 72V system.

[dogman]

Those isolated blocks of 24v are a damn good idea! While 48v might be pretty safe, you still get that tendency to vaporise a fitting when you screw up. I love good common sense ideas like that.

Now if I can just find a clean shirt to wear after looking at the pile of headways, dooling copiously. I can't stop thinking about that same pile of lithium in a geo metro or bug. As they say in Waynes World, someday,,,, it will be mine!

[Fishmasterdan]

Great project, somehow I missed this thread.

Keep up the good work. This looks like a really fun project.

[oatnet]

Yep, anything over 48V needs careful attention. Rubber linemans gloves are not overkill.

I hear you Lapwing!

In the picture below, note the linesmans gloves on the right. Each of the rows of batteries is 10 cells, 30.00v/33.75v/36.50v (nominal/resting/fresh charged). Note the red 2/0 cable connecting rows 1 and 2, with 400a Silicon fuses on BOTH sides. Rows 2 and 3 are connected by a solid bus bar for 60.00v/67.50v/73.00v, which I am thinking about cutting so I can fuse between them. The wire for each single-cell charger is fused right at the battery terminus, and although the unterrminated ends are just floating around right now, there are no fuses in them.

I also have contactors on both the negative side and positive side, energized by discrete circuits.

I spent a LOT of time thinking about every exposed surface, and have taken great pains to cover every potential contact point possible. I don't have exposed terminals like an SLA battery, and I am pretty sure i could drop a wrench anywhere and not conduct current - but I don't want to test it out either. When I put a sheet of plywood on top, this pack should be pretty well enclosed and maintenance free.

For the final mounting of the 180v pack, I am thinking of putting it between the front seats, 5 supercells per row, stacked 9 rows high (13.5"). Each row will rest in a tray that slides into a big steel box, each tray will be lined on all 6 sides with phenolic plastic, and each row will be connected by a silicon fuse. Cooling will become an issue at that point, and I have a lot of other vehicle-mechanical things to sort out first, so it will be a while before I go there.

ANYHOW, I busted my hump last night, and got all of the new batteries installed. I finished about 8:30pm and thought I would take it for a quick test. Clicked on the main contactor, clicked on the control contactor, pressed the pedal... and nothing happened.

So I tried to sort it out in the dark. I thought maybe I had popped a silicon fuse, but everything tested fine. Frustrated, I gave up for the night. I tested it again this morning, found continuity between the contactors, just no power! I ran a wire to test continuity of the negative 4/0 cable, which I had just been moving all over the place, and yep it was OK. Moved the wire to the positive 4/0 cable, which hasn't moved much - no continuity! I wiggled the battery side of the negative cable, and to my horror it came off in my hands, pulled right out of the lug.

So, power off, jammed the cable back into the lug, and viola! 108v. I could not resist taking it around the block to see how it rides. At 108v the bus now has at LEAST as much power as it did with the old ICE - plus with electric torque it is a little faster off the line.

This means I have met one of my design goals; I have successfully converted this bus to match ICE speed/accelleration, and because I used LIFEPO4, it is LIGHTER than when it started. There was no need to use 10c cells because the 80ah supercells will only be running around 2c continious, 40% of Headway's continous rating.

The Cycleanalyst, which is only rated to 100v, was flickering under load, and I could smell something burning, so I unhooked it. Until I can figure out how to get it working over 100v, or buy a new Custom version from Justin, I am not gonna have performance data The cells were ice-cold immediately after the "ride."

Next, I need to fix/replace the 4/0 cable, wire the Vicor DC-DC single-cell charger system, and start dealing with mechanicals like the motor mount, brakes, fluids etc.

Then I am gonna try to soup it up - it will be a real screamer when I bump to 180v, even with the extra 80 or so lbs over stock that will take me to (I have hauled 1500lbs of slate in this bus, 80lbs is cake ).

-JD <----- Sporting the EV Permasmile!

108v finished_6119.JPG

(119.29 KiB) Downloaded 1209 times

Cable on the right, lug on the left.

the culprit_6122.JPG (50.17 KiB) Viewed 1303 times

[Lapwing]

Awesome project. Glad you are taking the safety thing seriously.

Quick recommendation on crimpers. For a couple of years I used a cheap set of chinese made, box-crimpers up to 4/0 for my marine battery installs (what my employer provided). I was never happy with consitancy of crimps, even when I got really good at judging just how much "extra" to adjust the size nut. Had a several slip out in what seemed a good crimp.

When I quit the boatyard to go out on my own, I e-Bayed a set of THOMAS & BETTS - TBM5. What a difference. Heavy buggers but for $50 - perfection. I use only Ancor and T&B lugs, because these two brands fit the tool and have good dipped plating that holds up.

I did a bit of dangle-load testing on 2/0 cable and ran out of weights at 330#.

There is a set on e-Bay now item - 170311308703 http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&ssPageName=STRK:MEWAX:IT&item=170311308703

[oatnet]

[quote="Lapwing"]When I quit the boatyard to go out on my own, I e-Bayed a set of THOMAS & BETTS - TBM5. What a difference. Heavy buggers but for $50 - perfection. I use only Ancor and T&B lugs, because these two brands fit the tool and have good dipped plating that holds up.
quote]

Great recommendation - I'm gonna buy them, I need to go spend time with my wife RIGHT NOW, she has been incredibly patient with me being EV absent!

[oatnet]

I've been a little hung up on a technical detail, and worried that it was a design flaw.

I planned on charging each of my 80ah SuperCells with those Vicor 20a DC-DC converterz, and bought a whole mess of 30a ATC fuses to allow 50% headroom. Yesterday it just occured to me that TWO current flows will be going through each wire - Positive for one charger on the 1st cell, and Negative for the next charger for the second cell, etc, so it could be 40a - with headroom, too high for an ATC fuse. More, actually, since I am trimming v down.

I emailed Doc and Gary about it. Doc came up with a clever idea for putting hooking the converters in series, charging the whole string with that, as well as connecting the charging wires to each converter with a resistor in series to hold it to 10a.

Gary reported back that since the two current flows are going in the opposite direction, amperage is not cumulative.

So I guess I am back to my original design!

-JD

[Doctorbass]

I emailed Doc and Gary about it. Doc came up with a clever idea for putting hooking the converters in series, charging the whole string with that, as well as connecting the charging wires to each converter with a resistor in series to hold it to 10a.

Gary reported back that since the two current flows are going in the opposite direction, amperage is not cumulative.

So I guess I am back to my original design!

-JD

JD,

The idea i suggested to you was what you said but you forgot to mention that i said to still use the balancing wire (all next wires betweeen the main NEG and POS) to balance the cell.. but the idea to use a resistor to limit the current. The entire pack will charge via the main neg and pos.. and the balancing wires will still balance the pack with the single charging method.. but their balance current will be reduced.


I think what Gary said is that the two current flows athat re going in the opposite direction in the balancing wires result in no amperage.. but this occur ONLY when you close the current loop.

EX:

- let supose that you have only two cell and two dc-dc. cell1 with dc-dc1 and cell2 with dc-dc2
- if you charge separatly every cells with their own dc-dc at let say 15A,
every wires will have 15A circulating in them.

now if you connect these combo in serie for 2s1p... that will change nothing right..? ok

now if you only use one common wire for the cell1 and the cell 2 to the dc-dc.. what will happen?

yes... it will have twice current..

Doc

[PaulM]

- let supose that you have only two cell and two dc-dc. cell1 with dc-dc1 and cell2 with dc-dc2
- if you charge separatly every cells with their own dc-dc at let say 15A,
every wires will have 15A circulating in them.

now if you connect these combo in serie for 2s1p... that will change nothing right..? ok

now if you only use one common wire for the cell1 and the cell 2 to the dc-dc.. what will happen?

yes... it will have twice current..

Doc

I'm not sure I agree with this; would it not be possible to charge with the two cells hooked in series, the two converters hooked in series, and no common wire at all?? The common wire is only needed if the cells are not balanced.

[oatnet]

I'm not sure I agree with this; would it not be possible to charge with the two cells hooked in series, the two converters hooked in series, and no common wire at all?? The common wire is only needed if the cells are not balanced.

It is safe to assume that they will not remain balanced due to variations between the cells. Frodus's postings showed that his 1p large format T-sky cells went widely out of balance on each discharge because of cell-to-cell differences. I am hoping that by building 8p with smaller cells, they will average out to be similar from supercell to supercell, but I am still doing single-cell charging to keep them in balance. I also have the onboard charger which charges across the whole pack more effeciently, I'll see what method I end up using later but for now I want the pack balanced every charge.

I should note that this is another place where a BMI/PSI/Lifebatt cell allegedly will outperfrom the Headway. They assert that the automated process resultes in less cell-to-cell variation, which (if true, I'd believe it but I have no direct knowledge) would keep the cells in better balance. I couldn't afford to do this conversion with those cells though, so I have to work with the cells I can afford, I am just pleased to find good quality cells that can do the job are in my price range. Maybe I'll luck out and my 'averageing' theory will hold true.

The idea i suggested to you was what you said but you forgot to mention that i said to still use the balancing wire

Doc, I am sorry I butchered the explanation of the suggestion you were so kind to dream up and share with me my ignorance of electronics is showing, I apologize.

It took actually (3) Gurus to get the concept through my thick head, I went to Guru Fechter for clarification, whose explanation made what you and Gary were saying click. Apparently, even though each charger is connected to each cell individually, they effectively act in series so 20a will be going through the wires at each end of the pack and the balance wires will see 0v - if the cells are in perfect balance. Where the cells are out of balance, some current will go through the balancing wires for each cell.

That is, assuming I didn't mangle his explanation too , but I think now I have a dim mental model of how it works, and it resonates with a phenomenon I had seen before - two packs, wired in parallel, both charged with banks of single-cell chargers; when the first pack's charger failed, the other pack's charger charged both packs, so I thought one got a single-cell charge and the other got a series charge - now I see they both mostly got series charges .

As I wrote to Fechter, every day I see how much more I have to learn.

-JD

[Doctorbass]

JD.. This forum ia a very nice place to learn about EV!!

You just can't imagine how many people i refer to our forum!

What is imoprtant is that you correctly understand what we explain.. otherwise it mean we(guru) don't do our job correctly!!

Good evening!

Doc

[echas]

JD,

You did a really great job with the conversion and this forum! I converted a 240sx to electric and need to get the lead out. The car has 1100 lbs of batteries and is not happy. http://vintage-voltage.com/240sx You have inspired me to go for the LiFePo, but I'm concerned sea freight may take 3 - 6 months to get delivered.

Do you remember how much the FedEx shipping cost you? I counted 240 cells in your pack. Did you ship all cells using FedEx? I'd like to talk to you more about your experiences with these cells.

[oatnet]

JD,

You did a really great job with the conversion and this forum! I converted a 240sx to electric and need to get the lead out. The car has 1100 lbs of batteries and is not happy. http://vintage-voltage.com/240sx You have inspired me to go for the LiFePo, but I'm concerned sea freight may take 3 - 6 months to get delivered.

Do you remember how much the FedEx shipping cost you? I counted 240 cells in your pack. Did you ship all cells using FedEx? I'd like to talk to you more about your experiences with these cells.

Greetings Echas!

You are the first to post it, but your reaction is something I was hoping to encourage, by demonstrating with a big heavy old bus. I think there are lots of underutilized Pb conversions out there, that already have the EV infrastructure in place, the just need a battery chemistry replacement to restore them to ICE weight and handling - and exceed ICE performance thanks to electic torque.

It costs $12/kg - or $4/cell - to ship air. Victoria got the cells to me in 3 days via air - although customs interrupted the second shipment for 2 days. Even better, she matched the price of local Thundersky seller (where I could pick up cells without shipping), so cost + shipping = local pickup cost.

I am happy to talk with you or anyone else who is lurking - just send me a PM with your email address!

-JD

[headway]

Dear Mr.Doctorbass,

How are you?

Did you get our two samples of 38120P 8Ah cells?

By the way, I will be in San Diego from 22nd to 29th, March. Is it possible for our meeting there?

Best Regards
Victoria

JD.. This forum ia a very nice place to learn about EV!!

You just can't imagine how many people i refer to our forum!

What is imoprtant is that you correctly understand what we explain.. otherwise it mean we(guru) don't do our job correctly!!

Good evening!

Doc

[theyerb]

JD-

I'm thoroughly impressed (and somewhat jealous. haha!). Great work

[oatnet]

JD-

I'm thoroughly impressed (and somewhat jealous. haha!). Great work

Thanks theyerb!

Don't be jealous, try it yourself, it is not as challanging as it seems

OTOH, I spent my ENTIRE weekend wiring up a Vicor PSU, the board of vicor dc-dc converters, and the rows ov batteries so I finally have a single-cell charge!

I paralleled (5) 48v4.2a sub-chargers into a Vicor Autoranging PSU, and ran a set of leads, with a 25a fuse in a waterproof ATC fuseholder, with Andersons on the end.

There are 10 (10) DC-DC converters (48v-3.7v) mounted on the aluminum charging plate, to match the 10 supercells in each of the (3) rows in the pack. I ran individual positive and negative 16ga leads to each of the converters, collected them into a 10gau wire, and terminated in andersons. Then I ran a 10ga wires to the output side, splittin each one between the positive of one vicor and the negative of the next vicor, and terminating them in a block of 11 anderson conectors. At the same time I wired trim pots between the '-' and the 'sense' of each converter.

Then I ran a 12gau wire to the fuse on each copper battery strip in the pack, soldered and double-shrinkwrapped, and terminated them in blocks of 11 andersons, matching the block on the aluminum charging plate. Once terminated, I put a 30a fuse in each waterproof ATC fuseholder - 50% headroom over the 20a/.25c load I anicipated. I can also use these connections for a future LVC/pack monitoring system.

I adjusted the trip pots from 3.7-3.75v down to 3.60v. A little low, I know, but I will not discharging this pack below 50% any time soon, and I want to take it easy on the batteries. I saw a recommendation that the t-skys not be taken below 70%DOD for their first 20 cycles, I have no idea whether it is logical but I will follow it supersitiously. Similarily, I will take it easy on the charging for a while, and take it to 3.65v when I start truly exercising the pack.

I then charged each row in turn. I figured that the 10 converters, each drawing 94w, would draw 940w. As you can see from the display on the Killawatt, the bulk of the charge was at 650w, occaisionally peaking up to 700-750w. I don't fully understand why it would increase, but OK. Still, this managed to put out a massive amount of heat; the massive aluminum plate was exteremely HOT, even though it was sitting directly on a cool concrete slab. I had intended to put (30) of these vicors on this plate, I am glad I did not get that far.

When not connected to the pack, the PSU and converters drew 94w. Because it was sitting on 116w for a long time, I charged each pack to 116w. When I have more time I might let a row sit a while and see what it goes down to. Each row took about 1.2 kw, although I have no idea how much of that was heat. The good news is, I am no longer running on chinese electricity - I have the means to power the pack again!

Oddly, (1) supercell charged up to 3.68v, even though the DC-DC was trimmed to 3.6v. This made me think about fetchers point about the string being charged in series, with current flowing through the balancing wires only when the the cells were out of balance... That 3.68v sure looks like a series charge, which makes me wonder about the efficacy of single-cell charging-which I have always sworn by before. OTOH maybe this is symptomatic of my 3.6v instead of 3.65v charge.

Anyhow, to pictures:

The charging plate and PSU, finished and ready for action.

row charger_6141.JPG (46.41 KiB) Viewed 2672 times

The charging plate and PSU charging the first row at 650w

row charger_6147.JPG (55.86 KiB) Viewed 307 times

The batteries, with all three rows wired for single-cell charging. No more wires floppy around madly, a sense of order prevails.

charging_wired_6151.JPG (87.78 KiB) Viewed 307 times

Another view of the batteries, in sunlight.

charging_wired_6159.JPG (104.25 KiB) Viewed 307 times

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

[SRySwe]

I'm impressed!

Maybe it's the overvoltage to the DC-DC converters that causes the heat at charging?

[oatnet]

I'm impressed!

Maybe it's the overvoltage to the DC-DC converters that causes the heat at charging?

Thanks!

I trimmed the vicor psu to exactly 48v. The converters say 48v94w in, 3.7v74w out, so that means 20w of heat from each one. Not sure why they only pulled 65w each, or what the heat output was there.

-JD

[oatnet]

Today I replaced the passenger front brake line, now the bus no longer swerves left when I step on the brakes. Diagnoses was easy - when I first got the bus, it swerved to the right when I stepped on the brakes, and it turned out to be the drivers front brake line. Too bad I didn't replace both at the same time .

I installed the brake vacuum booster. I 3/8"-1/2" adaptor from brass plumbing parts I bought at Lowes, and hooked up the booster to the existing vacuum lines. It blew the 20a fuse I had connected to my main power line (the fuse on the booster itself is 25a) but a 30a fuse seemed to run OK. It never shut off though, so I probably need to replace all of the original VW vacuum line (sigh). Worse, the pump seemed to spew oil; I have to track down THAT leak. The manufacturer warns to never set the oil-filled muffler on its side - and I never have - but gotta wonder about what other problems the oil-filled muffler could introduce. It was reasonably quiet though.

1-3 adaptors_6183.JPG (64.43 KiB) Viewed 2473 times

2-assembled_6184.JPG (28.78 KiB) Viewed 2473 times

3-installed_6186.JPG (33 KiB) Viewed 2468 times

I bought the new high-current CA from Justin. I couldn't easily find a project box that was the right size, so yesterday I ended up cutting up the plastic case of an old UPS to fit and mounting it to fit. So, with the new CA in place I ran another test with the new batteries, and I youtubed video of the run below. The shunt setting was already at .5000 and in high-current mode from the factory, but oddly again the continous battery-side drain the CA was @40a, spiking up to 60-70a occaisionally. It consumed 145wh is about 1.2 miles.

I am concerned that I was only seeing 40a, because if it was off by a factor of 10 - then I would be burning 1,450wh in a mile and this project will be unviable, and a failure. I checked the voltage on the video - at 50a the cells are @86v, so that would be 30 cells@2.8v, reasonable for 5c or 400a, low for 500a. My 400a silicon fuses should have blown at 500a though, and the batteries should have heated up instead of remaining cold.

====================================================
Edit/Update: On a whim, I changed the default shunt setting to .0500; looking back in this thread, I see that I did that with the last CycleAnalyst on page (6). This new one is the high current model, designed for an external shunt, so when I saw .5000 I just assumed it was set correctly. It was not! Now, I see amp values jump up as high as 419a, but not a factor of 10 like .5000 vs .0500. Because there was a lot of traffic I didn't do the full mile, more like .85 miles, and consumed 527wh. That isn't fatal to the project like the 1,450kwh/m I extrapolated from the 145wh/m I got at 0.500, but it is worse than the 500wh/m I considered to be my worst-case scenario. This makes me wonder about the wh/m numbers I have seen posted by folks with lead-acid bus conversions, I'd rather fail and publish the results than lead other builders astray with false numbers. As it stands, it seems pretty likely that I am going to need to go to 180v for the pack to have enough Kwh to do my full daily commute.

On the other hand, that .85 mile course did include (5) stops, so it real-world consumption for my commute might not be so bad. I think I might also find some effeciency when I change the transmission fluid (gear oil) to something lighter, and doing the wheel bearings etc helped comuta-car folks a lot, I might get under 500wh/m. Since my commute includes a trip home for lunch, it is possible efficiency gains and a lunchtime opportunity charge could get me there. Or maybe I just do the morning commute in the VoltsBus and the afternoon commute via ebike. Or maybe I will have to fall back on the original Comuta-Car project.
====================================================

So anyhow, here is today's video:

[MotorSmoke]

Great project, I've enjoyed, like others, reading all the build details.

I have some thoughts about the required to roll down the road.

Are your tires rated for low rolling resistance and well pumped up? I believe rolling
resistance is the major drag at low speed. I'm into road bike racing and I sweat over having
the best tires because it makes a very big difference.
Most interesting for bicycles, the best rolling resistance tires also
have the best traction(on pavement), I do not know if that is the case with auto tires.
Is the motor truely running free? Does a fresh motor require braking in? Have you tried a "roll down
test"?, with in gear and neutral. One can calculate rolling resistance with a "roll down" test as well.