20-30 Cell Smart ANT BMS unable to read one of the cells when standing by.

[Marco Milano]

Battery is all wired to my BMS, but I am getting a “Cell verification error” on cell 16 and alarm beep.
This happens only when BMS is turned on and is “standing by”.
The error and alarm beep go away when “Auto balance” is enabled or charging is started.
The wire connecting the BMS to cell 16 has been replaced, but the problem is still there.
Cell 16 gives the correct 3.78V value when checked with a multimeter.
Did anyone have this problem before? How was it solved?

 

[amberwolf]

Looks like it's reading the cell just fine, as it showing you a voltage. This voltage is very different from the others, and about a tenth of a volt from the lowest cell, which usually indicates a problem (cell is different in characteristics from the rest, lower capacity / higher resistance / etc).

The BMS is probably designed to alert you to this problem (it's just not a clear message), so you can replace the problematic cells or otherwise repair the pack.

 

[Marco Milano]

Looks like it's reading the cell just fine, as it showing you a voltage. This voltage is very different from the others, and about a tenth of a volt from the lowest cell, which usually indicates a problem (cell is different in characteristics from the rest, lower capacity / higher resistance / etc).

The BMS is probably designed to alert you to this problem (it's just not a clear message), so you can replace the problematic cells or otherwise repair the pack.

I did replace the cell but it is still doing the same with cell number 16. I’ll look into the BMS parameters. There might be one that monitors voltage difference between cells and trips the alarm. I’ll also try to charge the battery and balance the cells as much as possible.
I sent an enquiry to the AliExpress vendor I bought the BMS from, but they did not answer yet.
Thank you for your attention and help.

 

[DaLanMan]

Have you checked the connection?

If the wire going from that cel to the bms goes across a multi-port junction (so the wires are ganged into a single plug) you have minimum 4 points of failure per wire in that junction. I would go over that carefully with magnification if you have it, look for a dry soldier joint or slipped wire-pin connection.

that could be an alternative reason for the odd behaviour. If the wire=good, the board=good and the cel=god but you are getting intermittent response issues. Last thing to check is for all contact points being stable.

 

[amberwolf]

If the reading is incorrect on that cell number in the BMS, regardless of the actual votlage of the cell connected there, then the BMS channel may be defective or damaged.

If the readings of the BMS on all channels are identical to an external meter's reading of the same points, then the BMS is detecting a real problem.

If disconnecting the cell from the BMS (but leaving it connected in the pack itself) changes the voltage reading on the cell, then the BMS channel probably has some defect or failure allowing voltage leakage from some other source that is affecting the reading.

 

[Marco Milano]

Have you checked the connection?

If the wire going from that cel to the bms goes across a multi-port junction (so the wires are ganged into a single plug) you have minimum 4 points of failure per wire in that junction. I would go over that carefully with magnification if you have it, look for a dry soldier joint or slipped wire-pin connection.

that could be an alternative reason for the odd behaviour. If the wire=good, the board=good and the cel=god but you are getting intermittent response issues. Last thing to check is for all contact points being stable.

I thought the wire might be problem, but it is not since I replaced it with a new one. The “sensing” wire and crimped terminals show continuity and zero resistance checked with a multimeter.
I think the problem is unfortunately in the BMS itself.
Debating by myself if I should get another one took a few days.
Another BMS is on the way.

 

[Marco Milano]

If the reading is incorrect on that cell number in the BMS, regardless of the actual votlage of the cell connected there, then the BMS channel may be defective or damaged.

If the readings of the BMS on all channels are identical to an external meter's reading of the same points, then the BMS is detecting a real problem.

If disconnecting the cell from the BMS (but leaving it connected in the pack itself) changes the voltage reading on the cell, then the BMS channel probably has some defect or failure allowing voltage leakage from some other source that is affecting the reading.

I think this is the problem, that BMS channel is not working right. The only way I can check this is by replacing the BMS with another one that is in the mail right now. Luckily they are on sale so it is slightly (very slightly) cheaper than I originally paid for it.
I’ll report back on how replacing it works out.
Thank you again to everyone who helped.

 

[amberwolf]

I think this is the problem, that BMS channel is not working right. The only way I can check this is by replacing the BMS with another one that is in the mail right now.

Are your cells reconfigurable, or welded together?

If the former, and you can move a cell (group) from one series position to another, you can take cells from a position that shows the correct voltage, and move them to the position that does not, and vice-versa. If the voltage problem moves to the new position, it's your cells. If not, it's your BMS (or connections to the cells from it).

 

[Marco Milano]

In order to avoid welding cells I used 26 pieces of 46160 31Ah li-ion cylindrical cells which are bolted.
I chose to go this route to be able to play around with the cells in case problems like the one I am having showed up. Will eventually use some epoxy resin to secure the cells once everything is working right.

 

[DaLanMan]

My luck with epoxy is pretty bad. The high heat stuff I got is pretty brittle, and the units I make that are not bike related get banged up a bit, I have found bits of epoxy sand loose in the battery enclosure a few times. Now I stick to Barley paper and shrink wrap, although I want to do some testing with Silica schmoo (to quote Big Clive). Just been busy with other priorities lately so no progress yet on that test.

 

[Marco Milano]

I do not have a lot of experience with high heat epoxy resin. Used it a couple of times to fix metal objects like a motor oil pan and a muffler. Repair was only temporary though, until I received new parts. The epoxy did hold up for a few weeks. Cannot tell how much longer it would last though.
I will eventually experiment tack-gluing the cells in accessible points to allow removal of the drops of glue for future cell maintenance.

 

[DaLanMan]

avoid hot glue, there is high-temp but it has a nasty habit of out-gassing when heated (I noticed it at about 55c starting to get stinky) I have no idea what it is that it outgassing in high temp hot-glue, but as a Bio Chemist I will assure you. It is not good for you.

High temp Silica gel so far besides it being a royal pain in the arse to clean off (If I have to use it, I wrap the cels in thin tissue and then schmoo over those, so if I do need to take it back apart, it will come out) I have yet to regret going the extra inches to make it easy to un-shmoo things. The very last time I got lazy.. I had to pull and entire cel apart to fish out a dropp bit of wire that found the small breather hole I leave in the shmoo so there is room for expansion... That was a long weekend. Un gooping 91 cels....

I fully respect every one has their own preference. I saw Rossman (the mac repair guy in NY who burnt his bike up) showing off his "awesome" (TM) Build skills, which appears to be "bolt it up, then grab a couple cases of black high heat shmoo and start smearing" It gave me the wobblies, he completely incased every electrical component and connector, including the ground bus..... ::shivers;::

I just found that by making enough mistakes I learned how to hate myself a little less when it comes repair time.