[Ben321]

I have been working on installing an aftermarket kit on a Prius to makes it into a plug gable hybrid. I have as input a 16s LiFEPO4 battery.

I have a problem however with the BMS that come with the kit. Cells #12/16 and cell #13/16 both read incorrect voltage (cell 13 reads .2V too high compared to the multi meter reading and cell 12 reads .2V too low). I am worried that the BMS will then "burn" these 2 cells by overcharging cell 12 and over-discharge cell 13.

I was wondering if there is any way I could re-adjust the BMS reading so it properly detect the correct voltage. Otherwise, would it be possible to maybe insert a low resistance on one or many balancing cable to compensate for this reading ?

I could also just buy a cheap BMS circuit without monitor but then I have no idea how I could trust such circuit without feedback.

The kit in question is an Enginer kit (bought a used one)

 

[john61ct]

Critical piece of kit, I would just replace it with one that works.

 

[frodus]

How are you measuring the voltage on the cell? At the cell? or at the Enginer BMS?

What kind of multimeter are you using? Is it calibrated? How accurate is it and on what scale? i.e. if it's 0-600VDC and 1% accurate, it could technically be off 1V. If the scale is 0-10V and it's 1%, it could be off +/- 0.1V.

 

[Ben321]

I measure the Voltage directly at the cell. For every cells except 12 and 13 , the reading on the multimeter is just about .01V difference (less than 1% difference). For these 2 cells the difference is .20V. If the difference was the same for all 16 cells then I could consider the multimeter to be faulty, but in that case I cannot rule that mutimeter out.

This is a very simple multimeter, I use a resolution of 10mV +/- 5% (20V scale)
https://www.pyleaudio.com/sku/PDMT28/Digital-Backlit-LCD-Multimeter,-AC,-DC,-Volt,-Current,-Resistance,-Transistor,-And-Range-W-Protective-Rubber-Case

 

[Duncan]

Hi
Are you saying that all of the cells are within 0.01v and that the BMS thinks that two are 0.2v different
Or that two cells are 0.2v different?

 

[Ben321]

Hi
Are you saying that all of the cells are within 0.01v and that the BMS thinks that two are 0.2v different
Or that two cells are 0.2v different?

All cells are about 3.3V eacH when checking with Multimeter

BMS LCD screen thoug shows cell #12 at 3.1V and cell #13 at 3.5V

 

[Moltenmetal]

Return the defective unit to whoever sold it to you and replace it with one which is working. There's no easy way to recalibrate a defective individual BMS channel, and no obvious reason why a couple channels would be so out of range. Agree it's a critical piece of safety equipment and needs to be replaced.

 

[Ben321]

These are used equipment I bought over 1 year ago. Returning them is not possible, I either need to repair or buy new part.

Ok so now I need to find a new BMS unit that would be compatible with my 90AH Lifepo4 48V 16s kit. Google search usually return me some result that seems for eBike kit (probably much lower than 90AH).

I did try to contact the original company but their quote makes no sense. Over 2000$ just for the BMS.... this is ridiculous.

I did saw a Youtube a video of someone who replaced ths Enginer BMS with suposedly a better quality one. I would like to do the exact same thing and was wondering if the community on this forum could help me figure out a trustworthy source for that kind of equipment.

Regards

 

[frodus]

That's a cheap multi-meter without calibration. From the manual for it, it's +/- 0.5%. On the 20V setting (which you'd be using), it could be off 0.1V from the REAL battery voltage. Comparing the same meter readings between cells won't tell you anything. It could be off 0.1V and you'd never know it. You can't compare a meter to itself. It's good for getting a rough idea of the voltage of a device, but still 0.5%.

As you're seeing, the BMS not likely a calibrated piece of equipment either.


How long are the BMS to Battery cables? Are they all the same exact length? Are all the crimps good on the ends? It's possible there's some resistance and the voltage reading at the BMS is wrong (Although it shouldn't measure high unless there's noise).

 

[Ben321]

I did all wire the same length on purpose. They are all 70cm long (the top section of the battery is about 60cm x 40cm, 70cm was pretty much the shorter i could get to have so room to place the BMS).

I am thinking of opening the casing of the BMS itself and see if there is not some dirty or something between the connector of cell 12 and 13.

I have also read on forum some success story of electric car with absolutely no BMS. Maybe I could just manually bottom ballance the pack every X number of charge... but i might be too lazy for that.

 

[dcb]

I have also read on forum some success story of electric car with absolutely no BMS.

IMHO, at a MINIMUM you should at least monitor half pack voltage vs the other half (without causing an imbalance in the process) and terminate charge or take action accordingly. Maybe I should quit talking about it and make a kit.

 

[Moltenmetal]

frodus, a multimeter like that should read the same reading (+/- 0.1V) on the same voltage every single time. Its accuracy may not be very good, but it should be very repeatable.

As to a replacement BMS: the Ah rating of the batteries doesn't matter. All a BMS needs to be is an alarm trip module on high and low voltage on every cell. Cell balancing at the end of charge has its strengths and weaknesses and absolutely is NOT essential if you're willing to spend a little time and effort doing a basic top-balance yourself if and when it's necessary.

The cheapest, reportedly reliable vehicle-type BMS available right now seems to be the one from ZEVA- I have no experience with it personally so cannot comment.

Other OEM units intended for e-bikes might be able to be made to work, but I personally have no experience with these either.

At the very bottom end is using something like a CellLog8, which is a cell monitor capable of watching up to 8 cells with alarming which you can use to shut off a charger and possibly to give an under-voltage alarm/interlock- not sure if they offer two separate alarms or not. A bit of searching here of old threads will find people who have done this, who you can PM for more details. You'd need two of them for your 16 cells. They're very cheap, but there's some worry about them causing some imbalance over time.

 

[frodus]

frodus, a multimeter like that should read the same reading (+/- 0.1V) on the same voltage every single time. Its accuracy may not be very good, but it should be very repeatable.

Correct, but he's also comparing the voltage on the multimeter to that of the BMS, with a different accuracy. I'm just saying, it's not going to be exactly correct.

As far as the difference between cells, you're right, the accuracy doesn't change when the multimeter looks at diff cells. So if it's off 0.2V between 2 cells, then those cells are likely off 0.2V from eachother.

BUT, the way that BMS's are usually done, is that each cell gets it's own voltage sense, so it's like having one multimeter per cell, not one that gets switched between each cell. So with resistors, and other components and their tolerances, the sensed voltages might vary more between cells, esp if the tolerances of the components aren't tight (>5% resistors for example).

 

[Ben321]

So I finnally decided to unplug the BMS and open it up.... and holly cow ! This thing got seriously damaged !

First of all my intuition was right, there was totally something wrong on the wire between cell 13 and cell 12. Here is a commented picture:

Text on the previous picture: We can see here that there are a total of 16 fuses. Each of them is covered by some red shrink-wrap. Fuse #13 has no shrink wrap ! So it was either changed or something ! (13.... superstition is not logical but funny nevertheless )

So i first though, Ok well maybe it is a question of making sure all fuses are connected tightly, and or replace all of them to be sure they have the same resistance. Using the multimeter i started checking the resistance of all of these and they all seem fine (0.06 ohm, which is the same reading than connecting the two electrodes of my multimeter together, so yes you are right this is a cheap one).

However i decided to also take a look under the board... here is the horror show:

Text on the previous picture:
Now when I take a look underneath the BSM board, first of all the casing is full of "marks" and small black "spots". It look like they were causes by electric arcs.
When checking the green board (not very visible on the picture) there is a symetry between the small dots on the metal casing and some black graphite like burn dots on the green board. This corroborate the theory about electrci arcs. This BMS sure did too a lot of beating !
Finally, there is a part that seems to have seriously overheaten. On the back on the casing there is a cigarette like burn. There is a total of 16 of these parts on the board, so this must be one per BMS wire.

And here is a close-up of the part that I could "try" to replace myself.

So now the dillema: i can try to fix that BMS myself. But if that thing took such a beating chances are it cannot really handle the full discharge load and stress when the system is running.

So I am open to suggestions for a simple system that will not damage the battery.

 

[dcb]

good call on inspecting it! I wonder if a battery connection was opened and it reversed through the bms (and it were not designed for it).

here is a very simple balance monitor, appropriate caveats, doesn't control anything, just watch it like a gas gauge for imbalance. don't know if that enginer system works without an official bms or not or what the signals are.

http://www.diyelectriccar.com/forum...-pack-balance-monitor-batt-bridge-165953.html

there are a few enginer hacks out there too ie
https://www.youtube.com/watch?v=J9SqGUH7rGc

 

[Duncan]

Hi Ben

I just use a "Batt Bridge"
http://www.evdl.org/pages/battbridge.html

This will tell you immediately if something is out of wack on your battery - then do regular checks with something like a CellLog
(Regular is six months for me)

 

[Ben321]

Hi Ben

I just use a "Batt Bridge"
http://www.evdl.org/pages/battbridge.html

This will tell you immediately if something is out of wack on your battery - then do regular checks with something like a CellLog
(Regular is six months for me)

Thank you very much Duncan. I might use this as a short term solution (while I am still motivated to balance my cell manually everyday)

I have another question, this time about Lifepo4 Charger. I think my charger cut-off voltage is too high. It is a 58.8V charger. I have heard it is possible to lower the cut off voltage. I would like to bring it down to 55V or something. I dont see anything special on the charger itself, do I need to open-up the case to access some potentiometer ? And if so, how am I sure I get the exact good cutoff ?

 

[Duncan]

Hi Ben
I'm not an electronics guy and I know that I would NOT be able to identify the bit to change to drop the voltage

But there is a quick and dirty fix
Diodes effectively drop the voltage by about 0.7 volts - so a couple of diodes in series with your charger will drop the voltage at the battery while retaining the voltage at the charger

This is what I did to drop the 14v of my DC-DC (laptop power supply) to about 13v so I could use it to supply my 4S headway auxiliary battery without trying to keep them at too high a voltage

Anybody who actually knows about electronics want to chime in here?

 

[Ben321]

Hi Ben
I'm not an electronics guy and I know that I would NOT be able to identify the bit to change to drop the voltage

But there is a quick and dirty fix
Diodes effectively drop the voltage by about 0.7 volts - so a couple of diodes in series with your charger will drop the voltage at the battery while retaining the voltage at the charger

This is what I did to drop the 14v of my DC-DC (laptop power supply) to about 13v so I could use it to supply my 4S headway auxiliary battery without trying to keep them at too high a voltage

Anybody who actually knows about electronics want to chime in here?

That sounds interesting but that is a 15amp charger. Are you sure a diode can survive 15AMP ?

 

[Duncan]

Yep - just got to buy the right one

http://nz.mouser.com/ProductDetail/...=sGAEpiMZZMtoHjESLttvkpHLBuj34cQg8luRuXKkJI4=

This one is $4 - 20 amps continuous, 150 amps max - 150 v max

I am NOT recommending this diode - and this is a NZ site - I bet its a lot cheaper in the USA - I simply don't know enough! - just read the specs and you will know as much as I do!
This is just the first one that came up!!

But you can get nice diodes more than capable for just a few dollars