Trying to ease my mind some re BMS etc. recalls

  • Thread starter Thread starter hwitten
  • Start date Start date
  • Replies Replies 34
  • Views Views 4K
My old dongle had an on off button which I used and left it plugged into the port all the time with no current draw. What make of dongle (link) do you have?
Thanks, that's good to know. I got this one.
https://www.amazon.ca/KONNWEI-Bluet...SJ3F2H4J18W&psc=1&refRID=6EQTZZ0VESJ3F2H4J18W
Received it the next day, so looks like our local amazon whse had them in stock.

I also spoke to Henrik Scheel (the current SoulEVspy developer) on email, and he was very helpful. There are indeed many fake ELM227 Konnwei KW902 v1.5 dongles on the market, and they may or may not work. To verify you have the correct one, you need to load down the ELM227 Identifier app. It will show what version you have. If it is green (supported) up to version 1.4, and the above ones (2.1 etc) are red (unsupported), then you have the correct one. Here is a link that also explains it.
http://cvtz50.info/en/elm327/

Anyway, the one I got is not fake, and does work. Henrik also explained that the voltage measurement precision is .02V, so mine is very normal.
 
I have that same dongle and (1) it's fine to leave it plugged in if you remember to switch it off, and (2) on mine the plastic button fell out and I now need to poke a pen in the resulting hole to control it. The button seems to be retained in place using only the label.
 
I also spoke to Henrik Scheel (the current SoulEVspy developer) on email, and he was very helpful. There are indeed many fake ELM227 Konnwei KW902 v1.5 dongles on the market, and they may or may not work. To verify you have the correct one, you need to load down the ELM227 Identifier app. It will show what version you have. If it is green (supported) up to version 1.4, and the above ones (2.1 etc) are red (unsupported), then you have the correct one. Here is a link that also explains it.

There is kind of a sad story to how all of this came about.

https://en.wikipedia.org/wiki/ELM327

The ELM327 is a PIC microcontroller that has been customized with ELM Electronics' proprietary code that implements the testing protocols. When ELM Electronics sold version 1.0 of its ELM327, it did not enable the copy protection feature of the PIC microcontroller. Consequently, anyone could buy a genuine ELM327, and read ELM's proprietary binary microcontroller software using a device programmer. With this software, pirates could trivially produce ELM327 clones by purchasing the same microcontroller chips and programming them with the copied code.[5][6] ELM327 copies were widely sold in devices claiming to contain an ELM327 device, and problems have been reported with the copies.[7] The problems reflect bugs that were present in ELM's version 1.0 microcode; those making the clones may continue to sell the old version.

Although these copies may contain the ELM327 v1.0 code, they may falsely report the version number as the current version provided by the genuine ELM327, and in some cases report an as-yet non-existent version.[8] Released software versions for the ELM327 are 1.0, 1.0a, 1.1, 1.2, 1.2a, 1.3, 1.3a, 1.4, 1.4b, 2.0, 2.1, 2.2 and 2.3 only.[9] The actual functions of these copies are nonetheless limited to the functions of the original ELM327 v1.0, with their inherent deficiencies.

It amuses me that there is now an app out there to detect the pirate versions.
 
Another screenshot of my cell voltages, SOC is at 78% now.. This time one is .02V less, again within the measurement accuracy variance. So I guess all is good.
upload_2020-10-26_17-48-36.webp
 
Your 94 is my 57, a lagger. Overall, I think it's just fine and I'd bet most Kona EVs would show a similar deviation. Also note that cell 94 could be 3.939 and still read 3.92 due to (digital) truncation.
 
Your 94 is my 57, a lagger. Overall, I think it's just fine and I'd bet most Kona EVs would show a similar deviation. Also note that cell 94 could be 3.939 and still read 3.92 due to (digital) truncation.
Although on the prev screenshot it was the same as the others. The developer said that .02 is the accuracy variance, and it is normal that at different times, different cells can show that discrepancy.

I would like to say thanks to you guys, though for putting me onto this dongle and app. It truly is a no brainer to use.
 
My car has had the service campaign 960 BMS update so I have every confidence that the BMS will do what it is supposed to do.

AC (7.2 kW) charged my car last night to 100% (Plunge pricing for my electricity last night so I was being paid to use electricity!)

Sceen shots from Torque Pro

View attachment 9682 View attachment 9683 View attachment 9684 View attachment 9685 View attachment 9686 View attachment 9687 View attachment 9688
Your voltage is nice and low.. Mine shows 4.16 volt at 100% but it has to be over 4.17 volts as my total pack voltage is 409 volts but I do have 38300 miles on my Kona.. The good thing is that my cell variation never exceeds 0.02 Volts..
 
That Dongle listed above is from AMAZON CANADA ... I did a search on AMAZON.COM (USA) and that specific one doesn't exist. Is there a USA version of that Dongle available? or one that anyone recommends that is reliable? Thanks
 
That Dongle listed above is from AMAZON CANADA ... I did a search on AMAZON.COM (USA) and that specific one doesn't exist. Is there a USA version of that Dongle available? or one that anyone recommends that is reliable? Thanks
I think this is the same one. Just test it when you receive it, and return it to Amazon if it doesn't work.
https://www.amazon.com/KONNWEI-Blue...dchild=1&keywords=kw902&qid=1603813633&sr=8-4
If it is stocked in your local amazon whse, then it is probably OK. It is the cheaper ones shipped out of china (that take longer to arrive) that you have to worry about.
 
Another screenshot of my cell voltages, SOC is at 78% now.. This time one is .02V less, again within the measurement accuracy variance. So I guess all is good.
View attachment 9707
Took another screenshot today, and probably the last I will post for a while, unless something drastically changes. Anyway this one shows all cell groups exactly the same now. Can't get any better than that. So with that I would say my mind is very much at ease now, thanks to SoulEVspy, and to those that recommended it. Certainly very quick and easy to use.
upload_2020-10-27_16-49-51.webp
 
I think this is the same one. Just test it when you receive it, and return it to Amazon if it doesn't work.
https://www.amazon.com/KONNWEI-Blue...dchild=1&keywords=kw902&qid=1603813633&sr=8-4
If it is stocked in your local amazon whse, then it is probably OK. It is the cheaper ones shipped out of china (that take longer to arrive) that you have to worry about.

It was in fact IN STOCK & I ordered it .. Due to arrive Saturday. Now I just need to figure out where to connect it .. lol Is it "obvious" or can someone post a pic of the location for the KONA 2020 ??? I assume the instructions will show how to download the data via Bluetooth or something? Thanks again.
 
It was in fact IN STOCK & I ordered it .. Due to arrive Saturday. Now I just need to figure out where to connect it .. lol Is it "obvious" or can someone post a pic of the location for the KONA 2020 ??? I assume the instructions will show how to download the data via Bluetooth or something? Thanks again.

Like nearly all vehicles AFAIK, the ODBII port is located under the dashboard to the left of the steering column. You should be able to reach it while sitting in the driver's seat, but it will be tricky until you know where it is and get a feel for it...
 
A couple of days ago I decided to level-2 charge to 100% and watch the latter stages of it
closely in terms of cell deviation. Normally the DC current going into the pack runs around
18 - 20 amps, which is a 0.1C rate for the tripled-up cellblocks. As it approached the top, the
OBC dropped current down by degrees and it was only 5A just before terminating. Once
finished, the "real SOC" was still well under the displayed 100% ...

08chg100.webp


but even at the top of charge, the high/low cell figures were never more than 0.02V apart.
This is just afterward, with my phone running my usual OBD2 "dashboard". I have *not* had
any BMS firmware updates so far, for a handful of reasons including that I don't trust any
of the local dealers to not screw it up. I can watch for hazardous conditions and limit
extremes on my own for now.

I misspoke about full-charge voltage in some other post, where I said it stayed under 4.1V.
Clearly it goes a little above that but frankly, I wouldn't start worrying until something rose
over 4.2. If I stay at displayed 80 or 90% for normal charges the cells will rarely crest 4V,
unless something goes really wrong.

I saw nothing that I could identify as active balancing during the top taper, pack temps
remained about the same as ambient, and the charger and VCU/motor assembly were the
usual warm as the charging system runs the coolant loop through all of that when working, due
to the 10%-ish inefficiency producing about 700 watts that need to go somewhere. There
was no warmth at all in the hoses connected at the front of the pack.

I found that enough electronics power up during charging that even with the car "off", I could
talk to my bluetooth OBD2 dongle -- from cozily inside the house, yet, and watch the min/max
figures slowly progress upward. The cell *numbers* for each also did about what they always
do, settle on 1 and 84 with occasional blips to something else on one scan cycle. Per the other
Torque thread, I'd love to learn how those figures come up for other OBD2-jockey owners.

_H*
 
I've had no BMS updates either and I've never witnessed cell balancing take place, at least I'm assuming that's the case because it always jumps suddenly from 98% to 100% on all the half dozen times I've both charged that high and kept an eye on it. I'll assume also that charging current goes to zero at 98% such that an assessment of cell-group voltage can take place without being disturbed by charging current ripple.
I'd have to guess that somewhat more than 0.02 V imbalance is required to kick off the process and mine has always been in that range, 15,700 km to date. Cell 57 has been the only lagger at max 0.02 under, but if the values are truncated then technically the difference could be pushing 0.04.

At 100% SoC(disp) I'd normally see 94.5% SoC(bms) so 93 shows plenty of room, noting that you can add a single decimal place to your displayed value. We know that others have seen 95.5% or slightly higher.

Yes, the OBD port is responsive during traction charging, Run and Utility modes. And, almost certainly during aux battery saver events but I haven't verified that. TP can be set to start logging automatically when it sees the port wake up, which can be useful in some cases.
 
Back
Top