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

[vin]

Ok, Does it have an ECU?

It reads the CAN bus. It will work without any CAN messages, but once I put rpm on the bus it must be greater 600 rpm or so.

Hi John,

Sorry about the delayed response, I've just been trying things. I have absolutely 0 experience in can, co far everything has been without it. I guess this is the next mission on this journey. any ideas on where i would start? I'm assuming I will need to investigate a device that will allow me it interface of the BMW?

[johu]

Sorry about the delayed response, I've just been trying things. I have absolutely 0 experience in can, co far everything has been without it. I guess this is the next mission on this journey. any ideas on where i would start? I'm assuming I will need to investigate a device that will allow me it interface of the BMW?

Yes, but please take it to a more appropriate thread

As has become a tradition I will write about my latest Touran trip, 1000 km all around, 9 quick charge stops.
The first one was the best So I added the error flags lately, that signal to the EVSE when the vehicle has detected a malfunction. One of them is deviation in voltage. So say the EVSE reports 380V but the car only 370V then there is a considerable voltage drop somewhere. This condition is signalled after 5s with an appropriate flag. I actually tested the implementation on an ABB triple charger. But in Kirchheim there is some special 60kW charger that "precharges" before starting to charge. This extra time meant the 5s would time out and the charging session wouldn't start.

What this meant in practise: I was stranded with low battery 80 km away from home. To make matters worse I had also forgotten my laptop. So I couldn't just fix the issue by flashing a new firmware to the VCU. Think... Aha, Damien is among the small group of people who can compile STM32 binaries and I have his phone number So I looked up the firmware on github on my phone, found the evil timeout and then called Damien who thankfully was available. He changed the timeout to 20s, sent me a new binary to my phone. Wifi flash and retry - works!

All subsequent quick charges worked and thus were boring. The next one that didn't work was off poor mans charger at my customers garage. It turned out two adjacent cell voltages would read too low and too high, respectively. So say one falsely read as 3.8V and the other 4.2V while really they should both read 4V. Consequently the BMS would limit charge current to 0A. The fault was intermittent, so obviously a bad contact. We ended up hiring a place in a DIY garage and removed the battery cover. First made sure there weren't any loose cell interconnects and then cleaned the connectors and treated them with WD40. Not sure how sustainable that is but it works up to now.

It also seems the SoC display is more reliable again though I have to say when I got home last night SoC was displayed as 12% but the cells were depleted, some below 3V. So still not super reliable.

Also found a strange problem with temperature derating. Basically during fast charging the average cell temperature was displayed as 38°C. Then, when I took off it was still 38°C but suddenly jumped to 57°C. This resulted in a power derating to 33 kW. I'm pretty sure this is not representing the actual temperature because it wasn't overheating last summer (35°C outside temp) and also a hot battery pack can be felt through the rear seat carpet. I'm transmitting the outside temperature to the BMS and during charging the sensor is not powered and consequently I report 0°C. Maybe the BMS is then surprised by the sudden jump and does some sort of "correction". Maybe I should just transmit a constant value.

So, still waiting for a trip without surprises

[mjc506]

So four issues, none of them trip-ending. I'd say that was pretty good

I'm sure you've got this covered, but one thing that jumps out at me is the combination of an unexplained high temperature reading, and an apparent bad contact... Worth carrying a IR camera with you for a while in case it reoccurs? I know I'd be happier to be able to visually confirm no localised overheating (Perhaps a single cell at 57°C may not feel particularly warm through the carpet?)

[johu]

Yes, I would be more worried if the over-temperature was consistent. But like said, the sharp incline occurs after the load has been taken off the pack and it is fed with a valid outside temperature again. There are only 3 sensors in a Leaf pack and they are in the back of the cells, i.e. away from the poles.

[mjc506]

See, way ahead of me

[MattsAwesomeStuff]

Lol, Damien compiling a binary and sending it to you on the roadside.

This is the nerd equivalent of when you see someone with a flat on the shoulder and pull over to help.

What a guy.

[johu]

I found that the PTC heater wouldn't stay on consistently and therefor didn't produce much heat either. So I checked. Now, I have fitted a microswitch to the heater flap servo, to enable the heater when the flap is in "full heating" position. But it turned out the servo drove back and forth so much that it disabled the micro switch for prolonged periods.
The servo has a simple pot inside for position feedback. These pots tend to wear out and the dust that gathers around the wiper will produce inconsistent resistance. Simple mans fix: WD40 The servo is now also much quieter as it doesn't move around anymore and the heater stays on, yay

[johu]

As always, the WD40 "fix" didn't last long. So after a long search I found sortof compatible feedback pots (the white one is new, green original).
https://de.aliexpress.com/item/4001252117889.html
When measuring the old pot with multimeter you could see it never really settle on a value. So now swapped out 3 of the servo pots and it is again quiet and hopefully the passenger side servo stops getting stuck in cold position.

Also reprogrammed the heater control. I had made it cut back heater power when voltage sags below 13.2V but at the point where it measures this seems to be the case. So now 12.8V is the derating point. Good enough to never run on the 12V battery alone.

[johu]

Completed another 1000 km round trip yesterday. No real surprises this time
One though: I charged at a supposed 100 kW charger. I thought it was a misprint, but indeed the first 1 minute or so it charged at 65 kW! That is 180A, I was rather worried that my 10mm² charge cables would melt but they held up. Will still replace them when the 40 kWh battery goes in.

[JaniK]

That was Chademo right? Is the Leaf BMS giving permission to charge over 50kW? I have had max. 47kW even on a 150kW charger with a stock 24kWh Leaf.

That speed is sweet though.

[johu]

Yes actually the LBC allows 70 kW when the battery is sufficiently warm and empty. Maybe the Leaf VCU then limits to 50 kW. Boring

Also checked my bills. Looks like I drove 300 km with 50 kWh. That is 166 Wh/km, summer is back

[JaniK]

Oh, so battery is designed to take more in. Nice to know.

Good consumption numbers.

[johu]

Now I just wanted to get a feel how negligent it is to run 170A over a 10mm² wire. Not easy to find out with the online calculators I found.
So using this calculator I found that 170A on 10mm² cable will drop 1.21V and generate 207 watts of heat along the roughly 2 m of cable length.
Finally, using this calculator
I found that after 60s the conductor (with an assumed mass of 400 g) will have heated by 80°C, assuming no cooling at all. Given an ambient temperature of about 20°C that is 100°C, already quite scary. But it doesn't stop there. We now continue charging at 40 kW for say 10 minutes. 78W of losses and 150°C temp rise after 5 minutes. We can hold it right there, it is not plausible. After multiple 43 kW quick charges I inspected the cable and it showed no sign of insulation melting. Which I'm sure would happen around 100°C.

So lets try this one: https://www.digikey.de/de/resources/con ... race-width
Yes it's meant for PCB trace width, but lets give it a try.
And now we see the main variable. The outer "traces" would heat up only by 16°C while on an inner layer they would heat by 80°C because convection is much worse. That is steady state though.
Repeating the calculation with 110A (a fairly common current from 0 to 50% SoC at a 50 kW charger) the inner layer would heat by 25°C. That sounds very plausible. So indeed at 170A I was pushing it quite hard. Maybe it reached close to 100°C. Will inspect the cables when I do the battery swap and replace them with thicker ones.

[johu]

A large truck pulled up today and brought my Norwegian 40 kWh battery

I was a bit shocked when I saw the large dent in the front but it turns out there is a lot of leeway between the enclosure and the cells.
Not such much in the rear, where rather small dents pressed down on the cells. I did bend the metal cover open a bit to see what is below and it seems to just have pushed the pouch seams sideways a bit. Should be ok.

Total pack voltage is 347V, one cell voltage I got ahold of was 3.613V and 347V/96=3.615V. Looks good.

I found they moved the main pack fuse away from the safety disconnect to a separate unit next to the relay box. Also it is 350A in place of 225A. Temp sensors are now mounted to the cell terminals in place of being stuck in the rear. All else seems unchanged.

Had to take a small break because the pack lives outside and a thunderstorm is approaching. Will report more later.

[johu]

So, got half of the cells out (the smaller bricks).
I connected stm32-car to the LBC and it just works as is, contrary to what "the internet" says.

Cell voltages a quite even within 13mV, better then my current pack. SoH=93% which I'm quite happy with.

[johu]

I have disassembled one of the lying down quarter packs and reassembled it as a straight brick. This time I did put labels on the tap wires to reconnect them in their original order.

[johu]

First day in the workshop today. We removed the 24 kWh battery and disassembled it. Unfortunately one pole broke but I think the electric connection is still ok.
First brick is in, we'll do the second brick tomorrow.

Also swapped the 10mm² cable for CHAdeMO charging for a 16 mm² one. The 10mm² one held up well but now it will have to carry full power for a longer time. Yes I know some will think this is ridiculously small but I couldn't see an issue with the small wire gauge over the last year and many quick charge sessions.

[EV_Builder]

Nice to read Johu keep them comeing!

[ZooKeeper]

Also swapped the 10mm² cable for CHAdeMO charging for a 16 mm² one. The 10mm² one held up well but now it will have to carry full power for a longer time. Yes I know some will think this is ridiculously small but I couldn't see an issue with the small wire gauge over the last year and many quick charge sessions.

I spent WAY too much time surfing for that answer. Seems to be that AC and DC are quite different in ampacity with the same cable???

From what I could find published, which was not much.... The ampacity is directly related to the total area (DC) and the combined strand diameters (AC) of the conductive material (not the assembly).

At least in the USA, the electric code for ampacity seems to presume a 100% duty cycle :O

If my Prius HV cables, that are nearly 4m long, are fused for 125A @ 240v DC, they should take 60A @ 480V for fast charging just fine I would think.

The only way I know to validate a cable's resistance is to measure the voltage drop thru it under load and do the math. NOT a task one should ever attempt!

[johu]

So, finally the battery is in

Once again it was super ugly because the bricks are so close together. There was already quite little space last time without pole covers, guess how little space there was with them. The final challenge was to limbo the nut with the narrowest possible extension between the bricks, tighten it, and hardest part, get the tool out again.

The BMS and current sensor are mounted much neater than last time but the BMS will need extra weather protection. Last time it sat on top of the front brick shielded of with some foam, now it sits on the side of the brick. Also I didn't need to deploy any hand crimped extensions.

The connection between the two bricks is a bit less neat, I attached two 10 cm long cables to the poles which need to be joined up after installing. Thats because the two UPPER poles are now connected and they are impossible to reach when mounted.

Did a bit of charging with poor mans charger and it took promisingly long to get from 30% to 50%

[ZooKeeper]

YEY, great to see this upgrade!

[johu]

Alrighty, cover is back on. It was a bit harder to fit the passenger side cover because the LBC is now also mounted next to the brick and almost touches the chassis. Replaced the connection between charge relay and drive relay with one of the left over bus bars.

Now charged to 93% and it seems the voltage curve has changed quite a bit. The pack was delivered at 35% and 3.515V. So I wonder what they will be at at 0%. At 50% I saw about 3.75V and now at 93% it is 4.16V. So it seems Nissan allow the full 4.2V on the 40 pack as opposed to about 4.1V on the 24 pack. So I think I will limit the SoC to 80% unless I plan a longer distance trip.

Also quite amazing is the allowed charge power. At 50% it was still 70 kW and at 93% it is still 7.5 kW.

[ZooKeeper]

Would you happen to have the Nissan part number off the Disconnect Switch? You have the same later design as I, my pack came w/o the switch, the local breakers do not have one and the parts folks are less than helpful w/o a VIN

Thanks!

[johu]

I only have the part number of the 24/30 kWh disconnect: 297C1-3NA0B . It is much the same, only the 24/30 one has the fuse integrated. The 40 model still has the mounting bracket for the fuse but no fuse
I have my old disconnect left over, send me a message if you want it. I'll probably keep the fuse though.

[johu]

First test drive today. 105 km SoC from 93% to 44%. 110 on cruise control, even took it to 180 km/h briefly So 200 km on one charge is comfortably possible.