[DRIVING] VW Touran powered by Nissan Leaf [FINISHED]

[johu]

Subsequent repair today. Freed the cable from the one point it got caught. While doing so I found that I hadn't fastened the middle bolt that holds the battery pack up. Maybe that's the cars way of telling me to take off the battery cover

Once it was freed I found the enable signal was still shorted to ground. Then I found the cable had gotten caught at another spot as well. CDYCS - Clamp down your cables, stupid.

With that sorted and all bolts now securely fastened all is as it should be now.

Now comes the best part - all that good work will be undone when I install the ID.3 battery pack I asked the pros to sketch up a battery box for me that doesn't leak.

I could theoretically install the 9th battery module behind the axle but am not sure about it. It adds extra complexity and I'd have to throw out the Outlander charger. The Leaf inverter would probably cope, I suppose.

[tom91]

How will the box be made? I do not see how you will transfer the weight of the batteries into the chassis.

Also small hint, you are missing the CMCs which need to be in front of the modules (front being the terminal side)

Luckily the OEM cables can work with 4 modules stacked 2x2. The CMC should be insulated between module sections so wiring order does not matter.

I have done this in the mini battery box and it works fine.

Front box shows better the cables not being in HV module order.

[johu]

Thanks for that.

It's supposed to be 3mm aluminium.

It's bolted to 3 rails:

The one in the front (red) with 3 angle pieces, the middle one (also red) directly with 3x M10 bolts and in the back again with 2 angles (not colored). There are 40mm of extra space, I'm hoping that is enough for the CMCs. Otherwise I could relocate them to the side perhaps?

I think I'll add reinforcement bars where the box is bolted to the chassis.

[arber333]

Thanks for that.

It's supposed to be 3mm aluminium.
......

The one in the front (red) with 3 angle pieces, the middle one (also red) directly with 3x M10 bolts and in the back again with 2 angles (not colored). There are 40mm of extra space, I'm hoping that is enough for the CMCs. Otherwise I could relocate them to the side perhaps?

I think I'll add reinforcement bars where the box is bolted to the chassis.

Sorry to rain on your parade guys but if you want to use fast charging with those cells you really need to think about cooling system. Those cells are culprit to more than one battery fire. They start smoldering just at 48degC and reaction is irreversible unlike NMC chemistry.

Myself i intend to build a central plate by two 8mm plates with the 3rd 5mm plate in the middle. Everything glued by Sika glue and bolted together at the edges by countersunk screws. All together 21mm. The middle plate i intend to cut into a pattern where water will flow through. In the end i will have to weld two 10mm plugs where tubes will connect. So i need to find someone with AC TIG welding rig to weld aluminum.

[johu]

Can I read up on it somewhere?

I'm restricted height-wise to 250mm but the modules are only 110mm.
I will inspect the original cooling plates, maybe I can sort of reuse them

[arber333]

Can I read up on it somewhere?

I'm restricted height-wise to 250mm but the modules are only 110mm.
I will inspect the original cooling plates, maybe I can sort of reuse them
....

If you can use them yes!!! That would be really good. Even if you get to chop it in halfh and get it rewelded.

Let me discuss how i intend to solve this for my ID3 modules.
I will assemble modules in 4 packs 2modules on top and 2modules on bottom inverted. In between i will have a cooling plate which will double as a mounting jig, so assembly will be turnover proof.
Cooling plate i first intended to assemble from 2 plates one 8mm and the other 12mm and bolted together. 12mm plate would have cooling channels cut on milling machine as well as O ring seal channel.
But as i couldnt find a milling machine operator who would machine just one or two pieces for me i decided to make the plate in 3 pieces.
3 piece construction will include two 8mm side plates and middle 5mm plate. This plate i will be able to just cut by my hand jigsaw. There will be no Oring seal instead i will use sika silicon glue on both sides together with countersunk head screws.
I will still have to position 10mm tubes somewhere on the top plate and weld them to have inlet and outlet.

Also dont forget each module is 30kg and 4module pack is 120kg which is quickly too much for a single person and it might cause problems with two not so strong persons. Having more than 4 modules on a single plate will require additional supports to bolt to the chassis to enhance material stiffness.
Modules have 10mm bolt holse and those are quite sufficient to support 30kg individual weight of the module.
Grade 8.8 bolt should hold upwards of 640MPa force converted to G loading should be 1.9T for each bolt = 60G+.
Shear strength is a different animal. You have material capacity of 60% at shear forcess only. And if one bolt goes everything goes! Usually Aluminum box will fail before that bolt. You will now see the weak point really is mounting assembly to a car.

I am still assembling my rear box on Mazda. I will post photos of it here:
https://mazdamx3ev.wordpress.com/
viewtopic.php?t=4210

[johu]

Here it is in all its glory. 895x590x220mm (without spacing - will add 10mm of spacing between the modules and below the lower modules)

And here is the box

I will probably build it from 2 custom U-profiles. One from stronger material that is actually structural and one from thinner material that is just the cover. In the front and back and will be mounted with 2 L-profiles to the existing mounting holes and in the middle to the extra Arber-holes into the beam that holds the seats.

I will assemble the entire pack on the transmission jack and on the lower "cover" U-profile and then put the upper mounting profile over it with 3 studs sticking out to go into the Arber-holes.

I will try to only air-cool the pack. I will add inlets to the front panel that direct the air towards the gaps between the cells and then pull it out the other end with a fan. I learned from e-grandpas latest video that there's vents that block water but not air.
Software will make sure no module ever rises above 45°C or so.

Addition:
The BMS boxes will sit outside of the battery box in their own extra box so they don't get in the way of the mounting blocks. I will make cutouts for its cables and the two main DC cables. Would also like to put a service disconnect somewhere so I can turn off HV from outside the box. Maybe on the side and have it interrupt the high-to-low busbar

[arber333]

Nice photo. That is actually my favourite rtool right now .
Do you need me to come visit .

Let us explore another venue. What about strong central plane where all modules will be bolted to? M10 bolts fix modules in place and you just use sheet metal as cover - not structural. That way if you at one time decide to implement cooling you can do it with a single plate and bolt channel plate in between.

Hm... could you cut the bottom of your OEM box on one side. Use one row only as a central carrying member? Alu is already there...

EDIT: I would not use that last module in vertical position. Such connection could only allow for 75% of SOC use and shorten the life of that module. Unless of course those modules are omnidirectional, but i would get confirmation from OEM for that. Probably not worth the chance...

[johu]

Early forum release to clear things up

[johu]

Nice photo. That is actually my favourite rtool right now .
Do you need me to come visit .

Probably would be good, dad is getting old

Let us explore another venue. What about strong central plane where all modules will be bolted to? M10 bolts fix modules in place and you just use sheet metal as cover - not structural. That way if you at one time decide to implement cooling you can do it with a single plate and bolt channel plate in between.

Issue is how would you connect the central plane to the Arber holes? Also I'm not sure if the modules should be mounted upside down.

EDIT: I would not use that last module in vertical position. Such connection could only allow for 75% of SOC use and shorten the life of that module. Unless of course those modules are omnidirectional, but i would get confirmation from OEM for that. Probably not worth the chance...

I won't. If at all I would mount it behind the axle below the spare wheel. But I don't think I will, it adds too much complexity and also too much voltage at 108S

[arber333]

Probably would be good, dad is getting old
....

OK, let me see what i can arrange at home .

Issue is how would you connect the central plane to the Arber holes? Also I'm not sure if the modules should be mounted upside down.

I can easily make some new holes if you like .

Seriously, in mazda i intend to fit lower 2 cell modules inverted to the upper ones by a simple reason. Inside the modules coolant goo is much more pronounced on the lower side of cells hence you can conclude modules are meant to be transfering heat from the underside.
Some friends from Poland are using those ID modules in inverted configuration and there seems to be no adverse effects.
Cell tabs are on the sides and every second pair of cells is inverted to the first one anyway.

Of course fitting cells up would be even easier since contacts are not in the way...

[johu]

Now I went over the pack design with my neighbour and we found a rather fatal flaw - the output would be 0V!
We are looping - to - and + to + when jumping from the upper to the lower row.

This would work but is clumsy

So now I consider turning the upper row upside down. I could then use the grey heat compound to connect them to the outer shell of the box and thereby get rid of the heat of the upper row. Likewise on the lower side. Maybe with an additional blower or so. That way nothing but electricity leaves and enters the battery box. No danger of leaks.
EDIT: I could even snake some coolant lines along the top of the box.

EDIT: maybe this lets us decide about turning them upside down

[johu]

Cobbled together a DBC file for the MEB BMS with Toms help (https://github.com/Tom-evnut/VW-bms) and decoded the cell voltages of 8 modules

Two voltages stick out a bit. Highest is 3862 mV and lowest 3848 mV . so 14 mV Delta. will take a few more samples and watch these two carefully

[arber333]

Now I went over the pack design with my neighbour and we found a rather fatal flaw - the output would be 0V!
We are looping - to - and + to + when jumping from the upper to the lower row.
.....

Well i am seeing this as well. I am using big M10 through bolts and in my pack two bolts in the middle are going to interfere with simple flat bus bars.
I will have to make insulated 35mm2 cable shoes to bridge those bolts... but i will probably assemble cells in upside position. It is easier to connect everything together when accessible than in upside down scenario.

[johu]

My current plan is first installing the upper part of the box (large U-profile), then bolt the upper 4 modules in upside down. BMS cables will go through holes in the front. Then assemble the entire lower brick on the jack and lift it into the box. The outside bus bars will be connected from the side. This will also be the method to do service disconnect. Unless I can fit in a proper disconnect assembly.

[johu]

And latest iteration of my electrical layout. Upper row is now upside down. +/- will exit on the left (could be the other side just as well) and I will try to fit the service disconnect on the other side.

[johu]

And the first 3d-printed models of the mounting blocks have arrived.

[Aragorn]

Two voltages stick out a bit. Highest is 3862 mV and lowest 3848 mV . so 14 mV Delta. will take a few more samples and watch these two carefully

Given you have a 9th/spare module, might be prudent to swap the entire module containing the lowest cell at this stage? Then theres only one to keep an eye on.

[johu]

Given you have a 9th/spare module, might be prudent to swap the entire module containing the lowest cell at this stage? Then theres only one to keep an eye on.

Aye sir, did just that. Here is what it looks like before and after balancing. Not sure if I got the balancing right as the low cell sits even lower now

And here is some connector progress. Made 5 isolated bus bars from copper, 6th connection will be the ISA shunt that Damien donated a while back and the 7th connection will be a kindof, sortof service disconnect aka the VW pack halves bus bar

[arber333]

Aye sir, did just that. Here is what it looks like before and after balancing. Not sure if I got the balancing right as the low cell sits even lower now
grafik.png

And here is some connector progress. Made 5 isolated bus bars from copper, 6th connection will be the ISA shunt that Damien donated a while back and the 7th connection will be a kindof, sortof service disconnect aka the VW pack halves bus bar

Hm... maybe you should get a 12S connector and extend wiring to connect that JK BMS balancing module to streighten up your cells. Warning you need to limit balancing to 300mA in JK app!
https://www.aliexpress.com/item/1005002 ... 8201405%21

[johu]

Not necessary I would say

After correcting the balancing code yet again (more below) The cells are now even enough (red)

The hysteresis is 4 mV so it will remain like that except the few outliers on the right.

The CAN Ids for balancing are "interesting". Basically you send one message per 8 cells, starting at ID 0x1A555412. And you would just think you work your way through 16 Ids and have balanced all cells. Not so! After Id 0x1A55541B comes not 0x1A55541C but 0x1A5554AB . Not sure what VW was exactly thinking but it's why the rightmost 40 cells weren't balanced on my first attempt.

I will now turn off the BMS because it draws some energy from the cells while sampling and then get back to it after a few days to see how cell 52 has developed.

[johu]

The shunt is in. Don't really like how much metal I've exposed there compared to the bus bars. Will have to think of something.
Will cut down the aluminium angle to size that gets rid of some exposure

[johu]

It is growing. This U-profile set me back 180€ but I think it's worth it.
Still wondering how to best route the HV out of the enclosure, out the side or out the front.

[Proton]

Would Lexan be appropriate? Since is plastic would be easy to heat up and ben in any shape.

[arber333]

It is growing. This U-profile set me back 180€ but I think it's worth it.
Still wondering how to best route the HV out of the enclosure, out the side or out the front.

That is expensive i paid 180€ for welded box but it is only 1.2mm inox.
Well, from experience your pain is just beginning. I had real problems how to cut holes for cable glands into this material. I tried "step drill piramid" and failed... In the end i decide on a suitable "crown drill". Some 24mm diameter and to my surprize i just needed some WD40 and set my drill to slow speed and i had a hole in couple of minutes. While i was at it i could use what remained for a washer .