Nissan Leaf Stack not getting revs underload Topic is solved

[vitecd]

Smoothly

It is just my feeling, the car didn't shake

My problems with Verter are : the car shakes, and not getting revs. The original assembly is just for reference, to be clear whether it is software or hardware. If the same setup is going smoothly with Leaf SW, then it is Verter SW or the settings.
Referring to the last logs, Verter is actually getting 6800 revs, but wobbling.
My current question, why Verter can drop the torque request? Why is it not linear?

[larsrengersen]

Likely because the Zombiverter is better at trying to protect your cells.
In the Leaf demo you are seeing a serious undervoltage error of 2,7V.

Since the lowest voltage before load was around 3,6V this indicates the pack is really shitty. Allowing the cells to go down that low molests your pack even more.
My suggestion would be: stop abusing this pack before things go wrong and get a better one.

[tom91]

the settings

Zombie is designed to work with 100s if not 1000s of combinations so it needs tuning for every setup.

why Verter can drop the torque request

UDCmin is being approached so it drops the torque requests. The fact the battery caves so hard and goes widely out of of balance on the cells means the pack is done for and not suited for your application. I am suprised the "leaf hardware" accepts this shitty of a battery.

[vitecd]

it drops the torque requests

Well, but why is it not limiting? I mean, after drops, it rises immediately, and fluctuates up and down continuously.
I've charged the Leaf and am going to explore some ideas. The battery is just for "bench" testing; I've another one on the shelf in cells that I can put together after I understand what to do in general.

Can anyone advise me on the base setting and something to play with, please?

-- // --

One more question (two)
Does "CAN mapping" works from the WEB interface? I've tried to push throttle position to the CAN to be able to log it, but in my case I can't set the "Parameter" for the mapping, the list is empty.
The other question is : does the "potnom" showing the pot's position? In my case it is always 0.

[Jacobsmess]

Potnom should change with the pedal position.
Throttle might be excluded from the can mapping on the OI board. I seem to remember there being a discussion about it on here for safety reasons, however this might be incorrect.
Why not swap the good cells for the weak ones you have and see if it addresses the issues that Tom91 and others are trying to help you and advising with?

[vitecd]

Why not swap the good cells for the weak ones

Because it is Nissan Leaf battery, 400kg box under the car, sealed and contain 96 cells inside. I can not see yet, what I'm limited to the battery voltage only. Originally I've started with another setup (boat) and there are no significant voltage drops. My current setup (Nissan Leaf) having the same problem and I wish to recognize it.

[J0hannes]

On your vehicle, the battery IS toast as previously mentioned. If you want to keep testing Zombieverter regardless, you need to adjust params to keep the setup happy, meaning you need ways to limit the max current draw to mitigate the voltage sag to levels the battery can handle.

My experience is on Zombieverter 2.22A official and 2.23 testing, just to tag this bit of information for anyone finding this post with search
Unlike the Leaf VCU, Zombieverter DOES NOT adjust the max torque request based on what Leaf BMS tells is allowed, so that's probably one of the reasons you can have a smooth accel and drive with Leaf VCU but not with Zombieverter.

On the latest params you uploaded, try to reduce throtramp (1 or lower) and put the throtramprpm to something close to or above the max rpm limit.
You can see on the plot that at 1800rpm, throtramp stops and then torque/speed spikes

Also test what adjusting throtmax between for example 50-70% does to the feel.

If you want to understand better how Leaf VCU controls the torque request, you might want to make few CAN logs so you can analyze for example how the Leaf VCU handles control when the battery is in such bad condition as yours is.

[vitecd]

Unlike the Leaf VCU, Zombieverter DOES NOT adjust the max torque request based on what Leaf BMS tells is allowed

I'll try to work around this. Thank you.

But, again, I came to this setup from the boat setup, having the same result but without the faulty battery. The battery in the boat is about 70%SOH

[larsrengersen]

A boat is a different game compared to a car.
In a car at some point you will start reducing throttle because you go fast enough already and get scared.
In a boat you can (and even need to) apply (full) throttle to retain your speed. Thus there is a higher current for a longer period of time. This puts a heavy burden on your pack.
Having said that: tuning a boat is even more delicate in finding the right parameters. Myself I don’t have any Zombiverter knowledge but what you need is:
- A BMS that calculates dynamic current limits (ideally based on SOC, temperature and time)
- A VCU that respects those limits

[J0hannes]

I'll try to work around this. Thank you.

But, again, I came to this setup from the boat setup, having the same result but without the faulty battery. The battery in the boat is about 70%SOH

I acknowledge that your boat setup has a better battery.
Hopefully you can appreciate the safety first approach the community is focused on when people give input on the issues you are facing. That said a battery issue on your sort of overrides the original issue if you can catch what I'm saying

Just to clarify, is your testbench Leaf also Gen2?
In first post you say that you are running zombie on pcb of your own design, care to share the schematic?

[tom91]

On the latest params you uploaded, try to reduce throtramp (1 or lower) and put the throtramprpm to something close to or above the max rpm limit.
You can see on the plot that at 1800rpm, throtramp stops and then torque/speed spikes

I had not spotted this yet, indeed this will cause very erratic behavior. Changing throtramprpm to lets say 12000 should dampen the shuddering ALOT.

@vitecd take your time to study the parameters of the zombie and when testing only change one thing at a time.

https://openinverter.org/wiki/Zombiever ... pot_Values

[vitecd]

Spent some time to play on the road.
Picture and a CAN from the same ride attached

[vitecd]

Few more scans adding.
I'll check the can logs, but looks like torq drops to 0 after some voltage decrees. Some more files attached.

[tom91]

Params?

[vitecd]

Another update.
I've matched parameters to Damien's "truck", and now it smoothed.
Going to check how is the weather on the river.

So, battery is not so bad for this subject.
Need to learn a lot.

--//--

If tonight, tomorrow no rain, will go to check the boat.

[vitecd]

Significant changes on the boat.
And... I'm 99% sure, all the problems related to the UDC, we've started today with about 76% SOC, and everything was great, but started vobbling in a few minutes, 1st thing I've tested the voltage and it was around 340, checked SOC 36%. Going back to the land.
The plan for tomorrow is to charge 100% and measure the lifetime on the water. Today, we achieved 50km/h on the water.
Screenshot from landing.
Video just before turning over for landing. SORRY for the video orientation.
In the first clip, you can hear it losing power and start jumping.

Going to learn the source code in depth.

[vitecd]

So, fully charged to 0, enough for 20 minutes on water )))