[WIP] Leaf to VW Type 2 Campervan

[tom91]

What are you seeing?

Burn marks on the input/filtering/balancing circuits on one place, so the last connector in this case

BMS is always connected last and disconnected first. The pack is split in two as you mention, so is the BMS circuity in terms of protections for over voltages, thus no electrical references between the halves.

[Marat]

The LEAF30KW battery diagram (EVB.pdf) is available on Google. The BMS module connectors are completely identical. You can take measurements on all 4 connectors with a tester. Just be very careful, smoke may start coming out of you.

[CorW]

The Service Disconnect has NO influence on the BMS.
The BMS is electrically divided in two, the two halves are coupled in their communication by an opto-coupler (galvanic isolation). The is specifically so you can pull the Service Disconnect while there still is some load on the battery output, because at that instant, the full pack voltage will be across the service disconnect *in reverse*.
The BMS consists of a processor with a UART I/O and 24 ASICs that each handle 4 cells, the processor UART output connects to the first ASIC input through an opto-coupler, from there a simple level shifter connects every next ASIC to the previous one. Between ASIC 12 and 13 is the mid-pack Opto coupler (just like the Service Disconnect is in the middle of the pack), then again at the end of the chain ASIC 24 output connects via an opto coupler to the processor UART input. The processor addresses each of the 24 ASICS and each ASIC inspects the serial message if it has their address, if not they simply repeat the message on their output. If the address matches, they decode the message and send a reply, which will ripple further through the chain until it reaches the processor.
The important part is that each ASIC is connected to 4 cells, so it is *powered* from the 10-16V that the 4 series cells carry and the ASIC can also measure and balance the cells under control of the processor by applying a 430 Ohm resistor across any cell to slowly bleed off some energy and lower the State of Charge when necessary due to good cells having lower self discharge than marginal cells. Or temperature difference or ....
If the 4 cells are not connected in the order expected, the ASIC may see a negative voltage or a too high voltage and since there is no protection in the cell connections, that will burn a small crater in the ASIC(s).
Most common mistake is to use a wiring harness from a different generation than the BMS, because the connectors are physically compatible but the cell connections to the connector pins is different between generations. Poof.
Also if you have modified the bus bars and the cells are not connected in the same order in the string as original, then the BMS wiring will connect cells NOT in the same order as the BMS has the ASICs connected to each other: POOF. Lastly, if the cells are in the correct order, but you modified the BMS harness because you re-distributed the modules (I changed the pack for my truck into 3 stacks of 4 plus 4 stacks of 9 modules, so I had to modify the BMS wiring a LOT!) then your best bet before you plug in the BMS is to grab a DVM and a thin solid core wire or thin sewing needle that you clip into your Voltmeter clips and you *measure* *every* pin in the 4 connectors that plug the 96 cells into the BMS, so you can be *sure* that every cell is connected correctly. The Service manual has a nice schematic for the LBC connections, so use the tools and take the time to avoid letting the smoke out, I have never blown a BMS myself and I completely rewired several packs.

[ianlighting]

Nice detailed explanation, thanks @CorW.

I believe I’m in a good place, although I remain a bag of nerves (probably healthy way to be). Modules are in original order. Original BMS loom. 10 out of 16 BMS module connections are original, unmodified. Final 6 are extended (thicker wire). Irrespective if changes or not, pins on all connectors tested yesterday are as expected and match with the complete module voltages.

I expect I will stare very hard at it again and try to uncover some other cock up, but basically next step is a new BMS…

Assuming all that is good, I guess the only potential slight unknown is whether a 5 series BMS for a 62kWh is fine to swap for my sizzled 6 series 60kWh.


Edit - meant to add for anyone else verifying similar, @Ali Bro did a useful spreadsheet - viewtopic.php?p=56104#p56104
Caveat - I didn’t use all the tabs, check yourself before relying on it.

[CorW]

Ian,
If you want to compare - please share your measurements of the cell voltages of each half of the pack on the 4 connectors that plug into the BMS (one black, 3 white) and I can compare with the 5 series wiring that I have here in my garage, to check that this is exactly what the BMS expects.
The agreed BMS is under way to you.

[ianlighting]

Ian,
If you want to compare - please share your measurements of the cell voltages of each half of the pack on the 4 connectors that plug into the BMS (one black, 3 white) and I can compare with the 5 series wiring that I have here in my garage, to check that this is exactly what the BMS expects.
The agreed BMS is under way to you.

Very generous of you, thanks.

Link below is a slightly modified version of the original by Ali Bro, who is often found around this forum.

I just used the 2nd tab 'Cell voltages at plugs', didnt really use the other tabs. The cell voltages are shown in the yellow boxes. Readings taken with the service disconnect plugged in.

These values shown are based on an average cell voltage of 3.7V. This is accurate for all my cells, except module 9 which is a little lower than the rest. It just means that the shown values from module 9 - 16 were 0.05V lower in reality that shown in the sheet.

The red boxes show the module values with the module number in italics next to them.

Just for a final sanity check, would be good to verify LB14 pin 76. This is a single light brown wire that goes to the service disconnect instead of the cells.

Edit - ah, interesting that the site displays the Google Sheet within the page. I assumed it would stay as a link. Looks readable, but shout if not for you and I'll look at other options.

[ianlighting]

HV cable shielding question.

When HV cable is used as a replacement for OEM busbars within a battery enclosure, what’s the correct thing to do with the shielding? Should it be grounded or fine to leave unconnected at both ends?

[CorW]

HV cable shielding question.

When HV cable is used as a replacement for OEM busbars within a battery enclosure, what’s the correct thing to do with the shielding? Should it be grounded or fine to leave unconnected at both ends?

The shielding is a safety feature. I have had a Leaf where a rodent gnawed through the corrugated orange plastic hose and ate the plastic outer sheet off the HV cable and there was some evidence of it even trying to get deeper into the cable - the metal braid of the shield pushed away to create small holes, but I could not see it having punctured the inside isolation anywhere, however, if it had been punctured, the shield would cause a connection of the HV to ground in *one* place, so no short circuit, but this would trigger the HV isolation detection to fail and prevent the contactors in the battery to stay engaged, so that the car would be safely disconnected from any HV. Same issue might happen during a crash - puncture of (one of) the HV lines, isolation detection fail and contactors open automatically, preventing further problems.
Concluding: shield must be grounded, not important if it is on one or both sides.

[CorW]

Link below is a slightly modified version of the original by Ali Bro, who is often found around this forum.

I just used the 2nd tab 'Cell voltages at plugs', didnt really use the other tabs. The cell voltages are shown in the yellow boxes. Readings taken with the service disconnect plugged in.

I am traveling now, can do some verification next week. Will let you know as soon as I have my hands on the BMS wiring again.

[ianlighting]

The shielding is a safety feature. I have had a Leaf where a rodent gnawed through the corrugated orange plastic hose and ate the plastic outer sheet off the HV cable and there was some evidence of it even trying to get deeper into the cable - the metal braid of the shield pushed away to create small holes, but I could not see it having punctured the inside isolation anywhere, however, if it had been punctured, the shield would cause a connection of the HV to ground in *one* place, so no short circuit, but this would trigger the HV isolation detection to fail and prevent the contactors in the battery to stay engaged, so that the car would be safely disconnected from any HV. Same issue might happen during a crash - puncture of (one of) the HV lines, isolation detection fail and contactors open automatically, preventing further problems.
Concluding: shield must be grounded, not important if it is on one or both sides.

My take on it had been, yes ground HV lines outside, eg inverter, PDM, and to battery, primarily for EMI. But internal to battery was not necessary. Perhaps I will rethink this.

[ianlighting]

It’s alive!

The VW has taken its first tentative steps out of the garage and into the spring sunshine… all is well with the world (ok, drastic overstatement, but I’ll grab any positive moment when I get one).

Noisy vacuum/brake pump, VCU wires all over the place, and holding the throttle pedal in my hand, so plenty still to do…

[ianlighting]

Insurance.

Good news / bad news.

Good news - my annual cost was previously £240 with a low valuation and agreed low annual mileage. They agreed to keep the same price! There’s an admin fee of £37 to switch and there’s only 3 months left to run, but pretty happy and surprised with that.

Bad news - apparently the insurers think I’m dead… Or at least the letter in the post expressed sincere condolences to my estate! Maybe they’ve had a premonition about my EV conversion skills Phoned up and seems there’s a box mistakenly ticked on their system. I have been unticked and the insurance remains valid - hooray!

So will need to up the value, but will wait to get to the the start of the next 12 months before seeing what effect it will have. Don’t know if it works like this, but kind of feel it’s set a good precedent to get in their system as ‘now running on electricty’ as they’ve put down. I wonder if they really understand what they are insuring… The nice lady on the phone was quite interested at the end of the quote. Wanted to know if it was cheaper than replacing with another petrol engine in her broken jaguar

[Jacobsmess]

Who have you insured with?

[ianlighting]

Who have you insured with?

Im with Sterling. Been with them for many years because under a different name (Herts) they did classic VW insurance.

[ianlighting]

Done some initial driving around the block. Made it back alive!

The benefit of living up a v steep and bumpy track is it seems to have dislodged lots of air from the motor stack coolant circuit.

Drove around in 3rd gear, max speed 40mph. I’ve separately tried 4th gear which was also fine to start off in. Helps to have lived with the sluggish behaviour of a petrol T2 because that’s my comparison.

I know some talk about changing up gears whilst driving to get into 4th gear for higher speeds. I’m unsure about this and don’t really know how to test safely. If anyone wants to share their knowledge on this, happy to hear.

Final point for now is leaf spy and the BMS. I’m unsure of the Leaf Spy logic. The motor stack is a 30kWh. The battery is 60kWh. The replacement BMS I believed was from a 62kWh. But Leaf Spy says 40kWh. Is it confused by my mix of components and makes a best guess? Perhaps the BMS was actually from a 40.

I can force it to 60 by switching the settings and saying explicitly it is 60/62 BMS. But by default on startup it is back to 40.

Here’s an image. Yes I know cell 55 is weak. Realised when I bought it, and decided to push that can down the road. I guess I’ve just caught up with the can. Edit - due to the construction of the 60/62 battery, you can’t swap individual cells, so it would be a complete module - think it’s module 9, one of the tall ones.

Edit - also you can see the last set of cells do show a different voltage. Either this is 40 / 60 BMS situation, or because of the cell taps extensions for these. I’ve been warned about potential anomalies with BMS wire extensions.

[Marat]

Leaf Spy shows AHr=106.65 this is probably BMS40.

[Bratitude]

Done some initial driving around the block. Made it back alive!

The benefit of living up a v steep and bumpy track is it seems to have dislodged lots of air from the motor stack coolant circuit.

Drove around in 3rd gear, max speed 40mph. I’ve separately tried 4th gear which was also fine to start off in. Helps to have lived with the sluggish behaviour of a petrol T2 because that’s my comparison.

I know some talk about changing up gears whilst driving to get into 4th gear for higher speeds. I’m unsure about this and don’t really know how to test safely. If anyone wants to share their knowledge on this, happy to hear.

What are the final drive ratios?

And what rpm motor limit have you set?

I’ve done 140km in 3rd in the brat.

[ianlighting]

What are the final drive ratios?

And what rpm motor limit have you set?

I’ve done 140km in 3rd in the brat.

Lots of variations over the years, but I believe the gear data for my bus (stock wheels / tyres) is
1st - 3.8
2nd - 2.06
3rd - 1.26
4th - 0.89
FD - 4.857


There is no config stuff to do with Resolve-EV, so no motor limits to play with. It’s just as Nissan set it.


What max speed do you do in your Brat?


I think the question I need to clarify is whether my reported capacity is incorrect due to a possible 40kWh BMS with a 60kWh battery.
This is a question I would appreciate input from anyone reading with experience. I know in the Leaf world people use a CANBridge to spoof the can data to get a leaf to to display correct mileage. But that’s when swapping a complete battery with its internal BMS. In my case I’ve got a swapped BMS and I’m unclear what affect that has.


It’s early days on my real world testing. I’m slowly building up to an extended run, so if there is a problem I’m not too far from home.

[ianlighting]

Leaf Spy shows AHr=106.65 this is probably BMS40.

Thanks for clarifying. What affect do you think this would have on reported battery capacity do you think?

[ianlighting]

Went for a drive. Went up to 50mph this time, all the while listening to every creak and bang from the road and hoping nothing serious was happening. My poor old bus has all its internal panels and insulation stripped out at the moment, and I can’t shut the internal engine hatch currently due to the way the coolant tank is temporarily poking through. So noisy is an understatement.

My weak cell is a concern. I called it a day on the test when I could see on LeafSpy the weak one was down to 3.50 while the others were up around 3.70. Have decided replacing this module is a priority now.

Turned today into a slow AC charging test day instead. So it’s working fine. I don’t have a fan on my radiator yet, so monitoring temps carefully. I have one of those little infrared temperature gauges you can point at surfaces. After a few hours the hottest part I could find was the rad itself at 40deg. Other surfaces cooler. Inverter was 37. So fan is another priority… Although be interesting to compare the temps when it’s fast 50kW charging on Chademo. I assume it will get warmer than this.

Battery hasn’t got warmer than 23deg.

I’m going to let that run up to 80% or bed time, whichever comes sooner and then probably have another drive. I’m expecting the weak cell to remain weak, but should still have charged and allow me to do a more extended test.

[Bratitude]

Lots of variations over the years, but I believe the gear data for my bus (stock wheels / tyres) is
1st - 3.8
2nd - 2.06
3rd - 1.26
4th - 0.89
FD - 4.857

So 2nd is 10.00 and 3rd is 6.12? kinda awkward ratios

And wheel size is?

My weak cell is a concern. I called it a day on the test when I could see on LeafSpy the weak one was down to 3.50 while the others were up around 3.70. Have decided replacing this module is a priority now.

Trickle charge that cell back up with a bench power supply before you do more testing

[ianlighting]

Trickle charge that cell back up with a bench power supply before you do more testing

Everything I’ve read suggests when you have a cell so far out it’s more likely a manufacturing defect. Even if you get it back to match the others, it’ll drop out of sync again. But I don’t claim expert personal knowledge on this.

[Bratitude]

Everything I’ve read suggests when you have a cell so far out it’s more likely a manufacturing defect. Even if you get it back to match the others, it’ll drop out of sync again. But I don’t claim expert personal knowledge on this.

If it holds it holds. Better off temporarily balanced than how you are currently using it

[ianlighting]

If it holds it holds. Better off temporarily balanced than how you are currently using it

Bit concerned that due to the construction of these modules from the 60/62kwh version of leaf, you can’t access individual cells. The BMS cable breaks out to 9 pins in this case (27 cells for a tall module) the module itself is 3p. So when leaf spy says ‘cell 55’, that’s the best granularity it can provide, but I actually have 3 cells that are ‘55’, and no way of knowing which of the 3 is suspect. If I was to directly charge using the relevant BMS pin, I can’t actually be sure how each of the 3 are performing.

And I need to buy a charger / power supply…


Example image of a medium size module with the front plastic removed. This example has 7 pins in the BMS wiring, for its 21 cells.

[Bratitude]

If you have cell taps, you can trickle charge the cell back up. A multi meter is all you need to find said cell.