jonescg's NEW electric racebike BUILD thread!

[jonescg]

Cheers guys! I'll call them in a while to see what can be done. I have a few ideas for a pack design, but I night need to change then in light of a good trackbike offer; $5k for a well set up 2004 CBR1000RR with all the goods. It's a little bit dated, but it is a Honda What I don't know is how much room there is for batteries - Gixxers are like wheelbarrows compared to the CBR.

[jonescg]

OK some noodling around on sketchup has resulted in this idea:


A CBR1000RR doesn't have as nearly as much room in it as the Gixxer, meaning 5s wide would be pushing it. So I thought I would come up with a battery mounting plate for a 5p 4s rack of LiPo. This means it is only 220 mm wide, and ~145 mm long with the cells terminated in place. Being so narrow, I can actually fit more of them lower down in the bike, so the pile can be up to 800 mm high. A second, shorter pile of maybe 10 of them can sit in front of the seat.

Now, the reason I only have one of these, and not a whole plate, is because I can Ctrl+C and Crtl+V a whole stack of them, depending on how high I want it. But it also made me think of an alternative means of pack building - individual 5p4s racks can be pulled out and put in as needed. More stuffing around, and a slightly weaker pack, but kind of convenient in other ways.



Oh, by the way, the units are 45 mm high, 220 mm wide and 10 mm deep, at the deepest point. The M4 holes would be threaded so you can tighten SS screws into aluminium bus bars that not only hold the paralleled tabs down, but also serve to join the neighbouring packs above or below in series.

The 10 mm wide strips on the outside would have M6 holes to thread into 10 or 12 mm side walls for the pack.

[jonescg]

OK, so I can make the stack out of 9 blocks of 1.1 kWh each. Plus the larger front cover has more strength now.



This means I can make a front stack 810 mm high (more like 840 mm with a bottom and top) and a back stack of 405 mm (435 mm including top and bottom) and have the L-shaped walls hold it all together.

I just doubt that the tabs will be sufficiently long enough to allow this method of clamping They would need to be at least 15 mm long... Maybe?

[jonescg]

DAMMIT! Who has $5000 I can borrow?

http://www.ebay.com.au/itm/2008-Honda-C ... _500wt_950

I'm worried it might be hot

Otherwise there is this option:
http://www.ebay.com.au/itm/SUZUKI-GSXR1 ... _500wt_950

However the 2004 Blade trackbike is probably the best deal going.

Damn I wish I weren't so broke right now

[voicecoils]

DAMMIT! Who has $5000 I can borrow?

http://www.ebay.com.au/itm/2008-Honda-C ... _500wt_950

I'm worried it might be hot

Otherwise there is this option:
http://www.ebay.com.au/itm/SUZUKI-GSXR1 ... _500wt_950

However the 2004 Blade trackbike is probably the best deal going.

Damn I wish I weren't so broke right now

Ha, there's a few on formula-extreme in that price bracket. I think it will cost quite a bit of money to take and sort of street bike and get it fully race ready. Much easier to just buy one already set up for the track and pull the oily bits out.

$4500, 2002 GSXR1000
http://www.formula-xtreme.com.au/xtremema.nsf/b5c953e8e1b65e94ca256a560030b8d6/b447bdee734e688dca25792700217a7b?OpenDocument
$5200, 2005 Honda CBR1000
http://www.formula-xtreme.com.au/xtremema.nsf/b5c953e8e1b65e94ca256a560030b8d6/13fb7b068ba690f7ca2578f80022e5d2?OpenDocument

[jonescg]

Hi All,
Not much to report, except I'm now a married man and I have no money. Sounds like a familiar story

But, my Turnigy 40C cells arrived!

And they are exactly as the sizes stated. However, the thing that will let me down are the tabs They are barely 9 mm long, so my mechanical termination idea just won't work. It looks like buying pre-soldered packs from HobbyKing and paralleling up the balance wires, making use of bus bars etc will be the easiest way to go, but I will try my hand at soldering these guys.





Interestingly, the positive terminal looks to have a copper tab spot welded to the ally tab, so maybe soldering isn't so bad

[jonescg]

As seen earlier, I have successfully soldered three of these cells in parallel.



This will be made much easier when there is a mounting plate to hold them all in place.



Now that I have the tab dimensions I can make some more accurate drawings. If I go for plates of two rows of paralleled cells, I can have the terminations between blocks in line, much like those on Voltron. However, being an even number, I will need to bump the final voltage up a tiny bit more to make the whole thing sit together. That is, 5P8S I will need a more convenient number to reach 112 cells (415 V fully charged at rest). I can stack these guys together to make a pack as high as I like, resulting in increasing the voltage by 14.8 V at a time.



6 MM holes in the 10 mm plastic will serve as mounting points for the side walls, and a floor will bolt into the side walls with an appropriate number of bolts.

I think it can work

[SplinterOz]

Looks like you might have a plan! I hope all went well last weekend.

[jonescg]

Thanks Tony, we had a blast Married life is good. Not sure why, but I'm loving it.

So a bit of an update from the other battery tech thread, I have been playing around with my Turnigy cells looking for a means to build a big-arse battery pack.

I have concluded that this is still a good way to go:



I spoke to a machinist here in Perth and he said they could easily come up with this exact battery front plate in acetal (POM). Now, acetal isn't good with heat, but I thought about putting some of that single sided copper, fiberglass PCB material (this stuff:


behind the tabs, then tighten down onto the tabs with a bit of 40x40x2 mm copper plate courtesy of 4 or 8 M4 screws and washers. This way there is good contact between the tabs and the front plate, conductive the whole way around, and yet the acetal plate is shielded from any excessive heating by the fibreglass backing. And it's repairable to boot! Only issue is that all this copper will add about 3.5 kg to the final weight of the pack, which is a lot but not too bad all up. I also have the option of putting some kind of protective cover on the final product to stop water, road grime and unsuspecting fingers from accessing 450 volts



I dread to think what machining 14 of these guys out will cost... I enquired about more superior plastics to acetal - for $4000 I can buy a single sheet of PAI I think this method will work...

[jonescg]

Holy CNC shitcakes. I had no idea this sort of thing would be so expensive! The machinist I spoke to quoted me $3000 to do this ONE unit! Let alone 14 of them!! (he charges $100 per hour).

I asked what the slow part was and he said machining all of the tiny slots. Apparently the feed rate is critical, and with acetal, it tends to be quite slow. I have since enquired about getting the board water-cut first, then getting him to machine all of the recesses afterwards. Might save me a few thousand bucks.

Otherwise would I be better off making a mould so I can fill it with HDPE or something? Apparently it doesn't take a thread well, so I could mount M4 nuts behind the front plate, but again, recessing all 32 nuts per plate would take a long time... More thinking required...

[nieles]

what about making potholes on the botom side of the plate, and put a nut in there. make the hole rectangular, so the nut will not turn and will clamp on 2 of the 6 sides
then put strips of fiberglass material across the holes to cover them up and not lose any nuts. this way they will also be isolated from the tabs (this could also be a big plate, but then it also needs all the slots machined in)

i could make a quick 3d model when i am home to better visualize the idea if you want

also when drawing parts always put a fillet on the inside corners.. this way there will be no problems afterwards with the corner radius of the endmill


when you will be using the PCB board between the cell tabs and the Acetal, the slots could be a bit wider, cheaper to machine
then make the slots in the PCB board only 1-2mm so you still have max contact surface

[Jehu]

Holy CNC shitcakes. I had no idea this sort of thing would be so expensive!

Otherwise would I be better off making a mould so I can fill it with HDPE or something? Apparently it doesn't take a thread well, so I could mount M4 nuts behind the front plate, but again, recessing all 32 nuts per plate would take a long time... More thinking required...

Could probably use nutserts, = drill hole, insert like rivets Cant remember where to get them from, Coventrys, blackwoods, or vip fasteners.
will find out, need some myself

[kfong]

You could keep it simple and use a pcb one sided board and just cut a lattice structure for barriers? It does seem like lining up all those tabs will be difficult. Why not just use a smaller pcb board to connect + and - banks together. It would make it more modular and manageable.

[jonescg]

I need to use thick acetal plastic, as the front plate serves as a structural element to the battery pack. This thing is going to weigh over 70 kg, so it needs to be strong. None of this would matter is I knew that the cells I was buying were reliable, but I just don't believe them. I have a feeling I will need to rip the pack out and weed puffed cells more often than I care. Oh, how do LiPo cells go with lots of pressure on them? The cells at the bottom are going to have a good 8 kg worth of pressure on them.

Also, I think sinking heat away from the tabs and onto the terminations is critical, as letting the heat go back into the cells would be very bad. I can possibly make the slots a tiny bit wider but this means less meat for the M4 screw to bite into. A captive M4 nut will occupy most of the remaining space...

Nieles, please to draw something if you get a chance, I'm all for ideas.

[AussieJester]

Checkz this out CHRiS..

viewtopic.php?f=14&t=33945

Dunno if they might be usable and i have no idea if the price is good or the reliability
of the supplier BUT the cells , if they are genuine A123s, have proven to be good performers...

KiM

[jonescg]

Thanks AJ, but like you if you want A123s I'm inclined to go with Cell_Man. I know he and his cells deliver I still need to find a way of making any kind of cell fit, and so far these Turnigy's are looking a bit more versatile.

On the termination front, I tried to find a plastics shop open on a Saturday morning in Perth Who was I kidding, everything's shut. Anyway, I found two bits of acrylic (Perspex) which added up to 6 mm, which is how much I plan on threading on my proposed acetal front plates. I drilled with a 3.2 mm bit, and threaded with an M4 tap. It's so nice threading plastic - just belt it through and reverse the drill. Awesome

I tried a bit of aluminium (ideally I'd use copper) which is 2 mm thick, and a couple of star washers to bite into the plate. The thread went in smooth and tightened quite nicely.

So I added a bit of PCB material (1.5 mm thick) and found that there was about 1.5 mm of missing thread on the other side. Not ideal, unless I can get my hands on some 12 mm long M4 screws.

So it would seem that 6 mm worth of thread through plastic can still provide some reasonably good tension between the PCB, the cell tab (0.25 mm) and the copper termination plate.

All is not lost

[jonescg]

Bodge time!

I got my free samples of 6 mm acetal and some copper plate and started to make a mini-battery to see if my idea is worth pursuing or not.

I drew up the slots on the acetal and drilled out the holes (3.2 mm). I then threaded these holes with a drill and an M4 tap. The cordless drill makes light work of it


So I cut the slots with a cut-off wheel on my dremmel. Not recommended Unfortunately the thinnest router bit I have for the dremmel is about 3 mm, and I was after something closer to 1 mm. Anyway, I also can't make separate slots, I just have to cut right the way through.

And as expected, not nearly enough tab was protruding


But I noticed that I can push the whole cell forward so the compressed bit just behind the tabs can fit into the slot, provided it is 2 mm or so wide. This is good, because the front plate can take more weight, it makes the whole pack 4 mm shorter (critical) and the tab protrudes by at least 7 or 8 mm - enough to punch a 5 mm hole in it for mechanical clamping.

I stuffed around on sketchup again and came up with this idea:



Absolute bitch to machine from 10 mm acetal, but it might be easier if the ridges and the slotted plate were made separately and the two 5 mm sections were fixed together somehow (screws I suppose - nothing can glue acetal).

Soldering is looking more likely...

[bigmoose]

I have been trying to figure out what the geniuses at A123 did with their package development in 2010 for a year or so. I figure that they spent a lot of energy optimizing this design. I have not been able to find a better photo or drawing of it. Note how compact it is at the tab connection area. It appears bolted, and the overall packaging sure looks great. It looks like there is a "shell" or U shaped liner for each cell that snaps together also to form the outer hardpack.

Perhaps an idea here for you... that I am missing!

On another note, instead of machining acetal away, I was thinking if you could imbed threaded brass strips in a fiberglass layup to make this header sheet? That way you could perhaps make the molds from wood even.

Found a video:

[AussieJester]

Great idea BiGMooSE what u reckon CHRiS
Time to lay up some fiberglass!

KiM

[Hillhater]

Chris... do you really need enough tab protruding to "punch a 5mm hole through" .. ?
If you have enough ( 5mm ?) tab to fold down, can't you simply clamp that to the board with a clamp plate covering the two tabs , with the screw in the middle ?
just a thought.

[jonescg]

KiM, you know how much I frocking LOVE fibreglass!

BigMoose, you got me thinking. Since these plates are covering the series connections, it wouldn't matter if they are made of conductive material since that's it's job. But the acetal takes a thread nicely and I stripped screw heads before I stripped thread when tightening into it.

Chris... do you really need enough tab protruding to "punch a 5mm hole through" .. ?
If you have enough ( 5mm ?) tab to fold down, can't you simply clamp that to the board with a clamp plate covering the two tabs , with the screw in the middle ?
just a thought.

That would allow for barely 2 mm of tab on either side of the screw. It might even be enough contact for high currents, but without the screw to hold them in alignment it would be a bit of a PITA to do 550 times. So would soldering, by the way...

Having the tab lay over the screw hole and having the screw go through it means there is a guaranteed connection. If movement, rubbing or a crash at turn 2 was to dislodge a tab, there is the potential for a more serious failure.

I'll have a bit of a go at this idea on the weekend...

[c_a]

check this out:
viewtopic.php?f=14&t=30950

[jonescg]

Hey Christian!

Haven't seen any finished product photos, are they in a neighbouring thread? I like the idea, although I think Aluminium will corrode faster than copper, so that will be something to be aware of. Also, these Batterist cells are easy in that they have huge tabs you can do anything with. I actually like having the tabs on the same side because you can face them forwards for a fresh blast of cool air.

...


Sigh, I have struggled to put something remotely good together, and it's just all too hard. I'd love mechanical terminations, as they are more robust and a bit more uniform. But these tabs are just too small to be practical.

So, it pains me to say, I have attempted to solder the big copper terminations. The 80 W iron does the job, eventually, but I have no idea what damage has been done to the cells by having so much heat on the tabs for so long. I'm charging the 5P2S pack now, and will attempt a few hard treatments to see what gets hot.

Failing this, I will go back to my A123 idea. At least I know that will work

[AussieJester]

::cough:: lay up :: cough :: fiberglass.. I darez ya CHRiS
Just need to let the release agent DRY before laying up the glass buddy...

KiM

p.s Go oooooon you know ya wanna

p.s.s ^^^ Drunk Post

[jonescg]

I'll leave the fibreglass to my last option. Still got a few ideas to try yet



So that's what it looks like now, after attempting to solder 2 mm copper plate to the tabs.



Not an easy task as it sinks so much heat, but I think I got there eventually.



Once the pack is charged I'll try a few high current discharges to see who comes loose first.