Interference: How to measure and fix it?

Nissan Leaf/e-NV200 drive stack topics
User avatar
janosch
Posts: 306
Joined: Tue Jun 30, 2020 9:23 am
Location: London, UK
Has thanked: 67 times
Been thanked: 54 times
Contact:

Re: Interference: How to measure and fix it?

Post by janosch »

Pete9008 wrote: Tue Nov 29, 2022 6:26 pm Any more details on this?
Hadn't actually plotted this before now, but I think Johannes is referring to the read cell voltages by the BMS, one half is more jittery than the other in a pack that we manually measured to be perfectly balanced at the time:

cell_voltages_columns_average_aligned.png
  • Presumably the jittery half is the front/are the further away cells.
  • It's not always the same cells that read high/low
  • both halves average out at almost exactly the same voltage
johu wrote: Tue Nov 29, 2022 6:06 pm A man from Germany who converted his Volvo had to split his pack in 3. 1/2 in the boot like yours with the stock tap harness, 1/4 fuel tank volume, 1/4 engine compartment. Obviously for the latter two he needed to extend tap wires also. The problem does not exist for him.
This is good knowledge and reassuring! I was worried I was the only person in the world with that setup/problem now.
johu wrote: Tue Nov 29, 2022 6:06 pm "old" sense chips, may be more robust?
I can quite easily test that by putting the 24kWh BMS in for a short test drive and see if that drops out when system under load.

Edit: Anyway, the cell balancing is a bit of a tangent. I will work through your comments and run experiments for the interference problem.
Attachments
cab01_during_chademo.txt
(2.24 KiB) Downloaded 46 times
cab01_stationary.txt
(2.24 KiB) Downloaded 50 times
User avatar
johu
Site Admin
Posts: 5681
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 153 times
Been thanked: 959 times
Contact:

Re: Interference: How to measure and fix it?

Post by johu »

Yes very typical sign that not the cell but the measurement is the issue: for adjacent cells one reads low the other one high by the same amount.
What's that "finger" in the first half during charging? Loose interconnect or high internal resistance of one cell? (Probably not)

EDIT: worth noting that the 40 kWh BMS not only has different sense chips but also smaller value balancing resistors. I think 40mA balancing current as opposed to 15 mA on the old ones, if I remember correctly. If you open up the interlock loop it will stop "balancing" - I think balancing is counterproductive with that sort of sense error and actually triggers the sense error because loading down the tap wire
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: Interference: How to measure and fix it?

Post by Pete9008 »

That plot looks like a good way of measuring whether changes are having a positive effect - make a change, rerun the plot and if the jitter is less the change has helped.

Assume stationary plot is with the inverter running but with the car stationary? If the inverter is off does the jitter disappear completely?
User avatar
janosch
Posts: 306
Joined: Tue Jun 30, 2020 9:23 am
Location: London, UK
Has thanked: 67 times
Been thanked: 54 times
Contact:

Re: Interference: How to measure and fix it?

Post by janosch »

God, what a great hive mind you guys are!
  • unplugging the inverter does not make the noise go away completely
  • not sure about that finger, it does not show in recent charts (re-wiring since)
  • going to try the 24kWh BMS now and see how measurements differ
Screenshot from 2022-12-01 11-19-09.png
See the plot with the inverter working hard, note the range as well:
  • MAXINT 8191mV in a whole block (A)
  • then vaguely reasonable readings for a bit (B) *
  • then from 200mV to 7000mV (C)
I wouldn't read too much into where the erratic measurements are at this stage, as the BMS is getting values that are so far out of bounds that it doesn't know what to do at all anymore.

Thats the main problem I am trying to solve, the 50mV jitter I can live with for the winter/until I have time for master/slave setup.

One hint: plugging and unplugging the BMS seemed to make a difference when Johannes and I were driving together, I have since made it worse again, just a little of acceleration will freak it out :)

* Edit: actually it goes down to 2500mV, so they are toast too

* Edit Edit:

Look at that, 24kWh BMS doesn't mind at all and has less jitter at standstill, both sides look the same! This might be good news for science, bad news for my project: "yes you can extend wires on 24kWh/30kWh pack, no you cannot do it on 40kWh pack". Driving around the block now.
Screenshot from 2022-12-01 12-01-36.png
Edit Edit Edit:
Look at that, under acceleration with 24kWh BMS one half sags as expected, but I get decent readings for the other half.
If I am disabling balancing anyway, maybe I use the 24kWh BMS, and with reliable current, temperature and most cell voltages I calculate my allowed dislim and chglim myself in the VCU... getting into BMS development there but could be a band-aid till I get to re-design the system and re-wire.

unless one half of the pack sags a lot harder than the other because of too thin HV bus? Pete? 35mm2 currently in the vehicle.
Strikes me as unlikely.
Attachments
Screenshot from 2022-12-01 12-20-32.png
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: Interference: How to measure and fix it?

Post by Pete9008 »

That looks a lot better :)

Like the idea about using the 24kWh BMS and doing the SOC calcs yourself.

Think its probably the comms between the individual bms ICs that is the problem, the 24kw and 40kw use very different systems. In particular the 24kw uses opto isolation across the disconnect (rather than the capacitive coupling on the 40kw) and that would make it a lot less sensitive to voltage differences between the two packs.

Edit - there is another thread asking about cable voltage drops which has got me thinking about it. Was going to do a few summs and will post the results later. I doubt it's causing the sag here though. How long are the cables?

Edit2 - do you know which pack is which on both the 24 and 40kw bmss?
User avatar
janosch
Posts: 306
Joined: Tue Jun 30, 2020 9:23 am
Location: London, UK
Has thanked: 67 times
Been thanked: 54 times
Contact:

Re: Interference: How to measure and fix it?

Post by janosch »

Just noticed that on 24kWh the lefthand side of the chart sags, whereas in 40kWh system the righthand side was jittery.
Maybe the old BMS reports the cells in reverse order in the CAN messages or this is a hint for something else ...
User avatar
johu
Site Admin
Posts: 5681
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 153 times
Been thanked: 959 times
Contact:

Re: Interference: How to measure and fix it?

Post by johu »

Almost looks like the right half isn't updating at all. Unlikely there to be 0 sag under acceleration. I remember this from my home made BMS. It had sanity checks (CRC) and would just stop updating when those failed. Usually happened under acceleration also.
Anyway, still an improvement.

Yes you could come up with your own charge curve, depending on cell voltage and temperature. Or Damien style, charge with full power until you hit 390V :D

EDIT: sag is always equal on all parts because the current is the same, says Kirchoff
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: Interference: How to measure and fix it?

Post by Pete9008 »

johu wrote: Thu Dec 01, 2022 12:50 pm Almost looks like the right half isn't updating at all. Unlikely there to be 0 sag under acceleration.
That would make a lot of sense!

Edit - although I seem to remember the 24kwh only has one way comms which means every message from every chip has to pass through all of them to get back to the master??

Edit2 - thinking about it if you are doing your own soc calcs could you then also run both the bms systems, half of each bms doing a pack?
User avatar
janosch
Posts: 306
Joined: Tue Jun 30, 2020 9:23 am
Location: London, UK
Has thanked: 67 times
Been thanked: 54 times
Contact:

Re: Interference: How to measure and fix it?

Post by janosch »

johu wrote: Thu Dec 01, 2022 12:50 pm Almost looks like the right half isn't updating at all.
Uuh, hadn't thought of that, raw data attached below, one half sags by 150mV, the other one 5 - 9mV, I think you are right, it might not update!
johu wrote: Thu Dec 01, 2022 12:50 pm Yes you could come up with your own charge curve, depending on cell voltage and temperature.
Yes, thats exactly my thinking, emulate the behaviour I saw before, but that is getting deeper into the calculations than I wanted to be - but I would have to do that anyway as soon as I switch to a third party BMS, no? Surely slaves just tell the master what to do and then the master is programmed by a "professional" (me or a vendor person with me on the telephone etc.)
johu wrote: Thu Dec 01, 2022 12:50 pm Or Damien style, charge with full power until you hit 390V :D
I don't know who that is, is that a previous forum member maybe?

johu wrote: Thu Dec 01, 2022 12:50 pm EDIT: sag is always equal on all parts because the current is the same, says Kirchoff
👍
Pete9008 wrote: Thu Dec 01, 2022 12:52 pm Edit2 - thinking about it if you are doing your own soc calcs could you then also run both the bms systems, half of each bms doing a pack?
The LeafBMS needs all connectors connected to give you cell voltages. I don't like that idea too much, adds complexity too and conflicting CAN messages with the same IDs.

I need to mull it over a bit, the question is, is it worth hacking something with the 24kWh BMS or is it quicker to just go Master/Slave thunderstruck or Orion.
Attachments
cab01_old_bms_stationary_no_inverter.txt
(2.24 KiB) Downloaded 54 times
cab01_old_bms_acceleration.txt
(2.24 KiB) Downloaded 54 times
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: Interference: How to measure and fix it?

Post by Pete9008 »

The other possibility is that the update rate is slow and some of those measurements are from before the run, some at the start and some at the end (or even after)? Could be worth doing a couple more runs with the 24kwh to see how consistent it is?
User avatar
janosch
Posts: 306
Joined: Tue Jun 30, 2020 9:23 am
Location: London, UK
Has thanked: 67 times
Been thanked: 54 times
Contact:

Re: Interference: How to measure and fix it?

Post by janosch »

Nearly went outside to test it, but they change bang on on cell 48, with 1 in 96 odds. Will check shortly though to be sure.

These are stupid questions now:
1) If one BMS gets good readings and the other doesn't the wires must be ok-ish, so should I shield the BMS unit itself? But then again it is half aluminium and plastic in the original Leaf.

2) Can I filter the input signals somehow?
We still have noise on the wires there, presumably it bothers one chip and not the other. Hm.
User avatar
EV_Builder
Posts: 1199
Joined: Tue Apr 28, 2020 3:50 pm
Location: The Netherlands
Has thanked: 16 times
Been thanked: 33 times
Contact:

Re: Interference: How to measure and fix it?

Post by EV_Builder »

So let me get this straight for a recap and reflection. On acceleration half pack sags and 2 cells go even further. It are measured values by the BMS. The communication is solid.

Dunno if you did already but:
I would tap the wires to BMS and measure during acceleration those extreme cells. Not that i think that the cells are doing it it's to check if the raw signals to BMS get disturbed or if the internal measurement gets disturbed. If you ask me it looks like current to the motor is drawn through the measurement wires. Yes not that much that they burn down , resistors are in play, but enough to disturb the measurement itself.

Does this make sense? If you know/measure a difference between both halves you either compare wiring and connections or BMS input circuitry etc.
Converting an Porsche Panamera
see http://www.wdrautomatisering.nl for bespoke BMS modules.
User avatar
janosch
Posts: 306
Joined: Tue Jun 30, 2020 9:23 am
Location: London, UK
Has thanked: 67 times
Been thanked: 54 times
Contact:

Re: Interference: How to measure and fix it?

Post by janosch »

EV_Builder wrote: Thu Dec 01, 2022 1:34 pm ... I would tap the wires to BMS and measure during acceleration those extreme cells ...
Yes you are spot on with your analysis, some notes:
- it is not two cells in particular, on the 40kWh BMS I just lose measurements from all cells, on the 24kWh measurement earlier two cells looked worse in one half, but I think thats a fluke.... although it might tell me something about individual wires
- we did tap into a random cell wire that was reading wrong as close to the BMS as possible, thats the shoddy photo of my scope in this post: (1)

I am mulling over some stuff now, I might override the BMS in the troublesome situations until I have a better solution.

I do get periodic checks of individual cell voltages whenever the vehicle is at standstill which is fine in city traffic, for motorway journeys I have to do some maths in the VCU and switch off/beep when I am getting close (assuming everything moves vaguely linear). Going to get inspired by/steal johannes WIP BMS code.

Edit: Linked this thread in the wiki for now (1)
User avatar
johu
Site Admin
Posts: 5681
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 153 times
Been thanked: 959 times
Contact:

Re: Interference: How to measure and fix it?

Post by johu »

janosch wrote: Thu Dec 01, 2022 1:20 pm 2) Can I filter the input signals somehow?
We still have noise on the wires there, presumably it bothers one chip and not the other. Hm.
I think there is some RC filtering in there, you could piggy-back some more capacity. That's a lot of soldering though, but maybe doing it for one sense chip could be worthwhile. I think the chips are also supplied by segments of HV and that's probably the part that disturbs comms like EVB is saying
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
Pete9008
Posts: 1801
Joined: Sun Apr 03, 2022 1:57 pm
Has thanked: 102 times
Been thanked: 347 times

Re: Interference: How to measure and fix it?

Post by Pete9008 »

janosch wrote: Thu Dec 01, 2022 1:20 pm 1) If one BMS gets good readings and the other doesn't the wires must be ok-ish, so should I shield the BMS unit itself? But then again it is half aluminium and plastic in the original Leaf.
The trouble is shielding only works for some types of noise. First you can split it into conducted or radiated, radiated can then be split into electric filed and magnetic field.

For conducted (which is what I think you have) foil or braid type screening will do very little. All you can do here is modify the circuit impedances to try and prevent the conducted voltage/current from flowing into/past the sensitive bits (this is the idea of the toroidal choke to increase the impedance of the signal taps at the noise frequency).

Similarly for radiated magnetic field shielding will do very little (unless you use mu metal). Simplest approach here is increase separation or change the relative orientation of cables. Twisted pair wiring is pretty good here as the twist causes the magnetic field to reverse for each twist and cancel out.

For radiated electric field noise foil or braid shielding is effective but only if it is continuous (it is surprising just how small a break you need to lose most of the screening effect). Again twisted pair works well here as does differential signalling (the interference is common to both so cancels out at the receiver).

The main problem with shielding here is that you have two sets of signals coming into the BMS, one from each pack, and any voltage seen across the HV power cable is also seen between these. these wire will always bring the noise into the BMS no matter how good the screening.
janosch wrote: Thu Dec 01, 2022 1:20 pm 2) Can I filter the input signals somehow?
We still have noise on the wires there, presumably it bothers one chip and not the other. Hm.
Yes but the trick is working out which signals it is that you have to filter. My guess is that it is the signals running on the PCB between the two BMS ICs each side of the pack split that are being affected. This comms link will essentially see the full noise voltage between the two packs. To reduce the noise voltage here you either need to reduce the amount of noise getting to the board in the first place or modify the circuit, neither of which are easy. To reduce the voltage you need to either reduce the voltage difference between the two packs (thicker, shorter HV power cable) or filter on all the remote cell tap wires (toroidal choke).

Note - when I talk about voltages here I'm referring to high frequency noise voltages (which could be the order of 10s-100s of volts) not DC voltage drops across the cable which are likely to be less than a volt.
User avatar
EV_Builder
Posts: 1199
Joined: Tue Apr 28, 2020 3:50 pm
Location: The Netherlands
Has thanked: 16 times
Been thanked: 33 times
Contact:

Re: Interference: How to measure and fix it?

Post by EV_Builder »

I think you need to measure on the bare PCB of the BMS.
We don't know the history of it I guess.

Did you measure current through the cell wires? While standing still and drive-in? My feeling is that then the current rises and the signal to bmw breaks down.
Converting an Porsche Panamera
see http://www.wdrautomatisering.nl for bespoke BMS modules.
User avatar
janosch
Posts: 306
Joined: Tue Jun 30, 2020 9:23 am
Location: London, UK
Has thanked: 67 times
Been thanked: 54 times
Contact:

Re: Interference: How to measure and fix it?

Post by janosch »

Ok, following the "own BMS/custom calculation" path now, and had a play with:

calculating discharge and charge limits

Comments for that in the thread over there please, I am not very experienced with Lithium and whats allowable.

Simultaneously we are re-routing some wires today and replacing the junction box, I might take further measurements at the BMS shortly.
Post Reply