Old BMS General Thread

Threads relating to the BMS system begun by Peter Perkins

Moderators: GregsGarage, retepsnikrep

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

Re: Old BMS General Thread

Postby GregsGarage » Sun May 20, 2012 12:27 pm

Further thoughts on using command 7 to help diagnose your problem. If you send command 7 and all the slaves flash their led in turn then your problem is the master bus, slave bus is working o.k. If the slaves don't flash at all then your problem is the slave bus and/or slaves themselves. Remember the slave bus is daisy chained, any break will stop command 7 at that slave. The master bus is parallel connected so shouldn't be affected by a problem at any particular slave, unless it gets shorted which would bring down the entire master bus but shouldn't do any long term damage to any components.
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Re: Old BMS General Thread

Postby retepsnikrep » Mon May 21, 2012 4:46 am

That all sounds very sensible.

Are any alarms being triggered?
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

isaac_alaska1
Posts: 100
Joined: Sun Apr 03, 2011 7:41 pm

Re: Old BMS General Thread

Postby isaac_alaska1 » Mon May 21, 2012 5:03 am

the mechanic is being very slow about getting back to me, so i don't know anything right now other than that picture. he didn't mention that the screen was doing anything like flashing alarms, so i'm guessing that somehow the number of alarms was reset to zero. i don't have access to a working system right now (the only completed one is on the snowmachine) to see what actually happens if cells is set to zero.

i did send another email and told him how to use command 7, but i haven't heard back from him other than the first email he sent.

i'll keep you posted and let u know if it gets fixed :)

isaac_alaska1
Posts: 100
Joined: Sun Apr 03, 2011 7:41 pm

Re: Old BMS General Thread

Postby isaac_alaska1 » Mon May 21, 2012 7:27 pm

received this from the mechanic today:

The only light on was the one that flashes once per second until I changed
the setting from zero to 48. It was showing 0 cells I went into the menu
and reset it to 48, then pressed the center button I then had an alarm go
off the right LED came on and the second to the right came on then off.A
white screen came on and read "cell 29 serial". I turned it off and turned
it on again and now the second to the right is coming on again and the
white screen is reading "cell 29 serial" again.


the leds are in this order:

cell high, cell low, toggle, watchdog error.

it sounds like somehow the number of cells was reset to zero like we suspected, but now there's something wrong with cell 29, as well. maybe the cold is affecting it and has pulled it down far enough that the cell voltage isn't high enough to run the slave pic on cell 29. that's been one of the lower voltage cells most of the time anyway. i'll have him try sending command one and read the cell voltages and see if anything comes from that.

*edit*

is there another way to scroll through the cell voltages? i don't remember if there's a separate menu listing for reporting cell voltages or if it's done through command 1. i just do it automatically without thinking about it and can't remember all of the buttons that get pressed =/

*edit again*

i looked through the code and found the cell voltages listing. i guess i've never used command 1....what does it do?

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Re: Old BMS General Thread

Postby retepsnikrep » Tue May 22, 2012 7:08 am

If the number of cells is now right and it all worked OK before and you are now getting cell 29 serial error then yes could be the cell group is dead/very low <2v

Command one tells the slave to send the cell voltage to the master. It does not display it.

Going into the menu1 and pressing up should display the cell voltages but may fail due to the serial error.

1) You need to physically test the voltage on that cell 29.
2) Check that cell voltage is present at the cell pic power pins 1 & 8
3) Check the load is not stuck on even if the led is off. (Failed transistor)
4) Replace your bad cells :(
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

isaac_alaska1
Posts: 100
Joined: Sun Apr 03, 2011 7:41 pm

Re: Old BMS General Thread

Postby isaac_alaska1 » Tue May 22, 2012 12:58 pm

So it's actually cell 26 that's having a problem, rather than cell 29.

the mechanic sent me pictures of the cell voltages today. everything looks good (if a bit low) up until cell 25. Cell 26 and beyond all have 1.75 listed as the cell voltage. I had thought that if there was a problem with the comms then the cell voltages wouldn't even be listed for cells that hadn't reported a cell voltage? I may be wrong though. does anyone remember/ can test it?

the mechanic said that the toggle LED is staying on constant, and then the watchdog alarm comes on after a while.
EVEN if there's a legitimate comms error or low cell, the master software SHOULD keep looping, and the program shouldn't freeze, correct? or does it get pulled out of the loop and return to "main" without finishing the cycle?

i'll have him manually test the voltages and see what comes up. there's a plane going up tomorrow and i'm sending him two new master pics and 5 slave pics so if it's a pic problem then hopefully it can get sorted. i've attached a picture that shows where the cells stop reporting accurate voltages.
Attachments
IMG_0089.JPG
IMG_0089.JPG (95.71 KiB) Viewed 8105 times

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Re: Old BMS General Thread

Postby retepsnikrep » Tue May 22, 2012 1:15 pm

Cell 26 bad slave or flat battery that prevents the slave bus signal being passed to next slave anyway.
So everything above 26 if garbage.

You need to diagnose cell/slave 26

I don't want to be a forteller of doom but are you using liPo? :shock: Be careful.

You have made a lot of changes to the software so i have no idea what it's doing.
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

Re: Old BMS General Thread

Postby GregsGarage » Tue May 22, 2012 1:32 pm

issac, I disconected one of my slave boards and got a serial comms error as expected. When I get the cell voltages up it reports the voltage from the program loop before I disconnected the slaves, that is all cell voltages look normal. When I cycle the power to the bms, again I get the serial comms error, but now when I look at the cell voltage, all cells after my disconnected slave show 1.75 volts on the master display. This is to be expected because the slaves subtract 1.75 volts before sending the voltage data and the master adds the 1.75 volts back in to the reading. So 1.75 + 0.0 from the slave gives you a reading of 1.75 volts. The slaves can only report from 1.75 volts to 4.30 (?) volts. The limitation is due to slave voltage data only being 1 byte long from each slave.

As far as the discrepancy between only getting readings up to cell 26 and a serial comms error from cell 29, if all the cell voltages are low your comms are going to be erratic and the cold weather is probably not helping. You really need to check pack voltage and more importantly individual cell voltages with a volt meter. The low cell voltages are worrying, the cold weather shouldn't affect resting cell voltage as far as I know. Did something get left on, ignition, DC-DC converter etc. that could have drained the pack? On my Matiz I have a contactor in the middle of the pack that prevents anything from accidentally being left on that could drain the pack. Ignition off and the contactor is open, for charging I also have to close the contactor.

Make sure whoever is checking these voltages does so in a safe manner, high voltage is dangerous.
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191

isaac_alaska1
Posts: 100
Joined: Sun Apr 03, 2011 7:41 pm

Re: Old BMS General Thread

Postby isaac_alaska1 » Tue May 22, 2012 1:47 pm

The mention of cell 29 was actually a typo, it was a problem with cell 26 all along. The cells have seemed like they have stayed balanced pretty well and 26 shouldn't be low enough that it would be almost a volt lower than the rest.

what is the lowest voltage that the pics will work on? the data sheet says 2 volts but i don't know what point the logic stops working.

i'm sending him an email and new pic controllers for the slaves, i've seen a pic quit working one time before, i'm not sure what happened to it or what had caused it to quit.

Thanks for the information! it's extremely useful.

btw did you notice if the toggle LED continued to toggle after one slave board was disconnected? the mechanic made it sound like the it was triggering a watchdog error, but that may have only been when he was reading cell voltages.

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

Re: Old BMS General Thread

Postby GregsGarage » Tue May 22, 2012 3:44 pm

For reference, I am running version 0.29 of the master software. My led does toggle between alarms, but not during an alarm. While you are in the menu the led isn't toggled, so if you stay in a menu for more than a minute the watchdog will reset.

I tested the slave pics down to 2 volts and they work fine, I seem to remember that you could go even a bit lower. You will probably lose reliable bus communication before the pic stops working. I can get my cells down to 2.3 volts under load and still get cell readings. Below that and I start to get serial comm errors as well as the expected low cell warnings. Really you need to check actual cell voltages with a meter before starting to change pics.
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191


Return to “BMS thread”

Who is online

Users browsing this forum: No registered users and 3 guests