Flying ADC 16 channel BMS 96S test

[maiks]

Hi all,

I have a Leaf current sensor (Hall effect type) with three wires: GND, Voltage, and Signal (current output). I'm trying to connect it to the Flying ADC 16 BMS, but I'm a bit confused about the proper wiring.

I've been reading through the documentation here:
https://openinverter.org/wiki/16-cell_BMS

In the Pinout section, it says:

“5V, CUR+, CUR-, GND. A single-ended current sensor is just connected to CUR+.”

That line threw me off. Since my sensor has only 3 wires, and 5V, CUR+, CUR- and GND are available on the BMS, I'm not sure how exactly I should hook this up.

Could someone please guide me on the correct way to wire this Leaf sensor to the BMS?

Thanks in advance!

[johu]

GND to GND, "Voltage" to 5V, "Signal" to CUR+
Parameter idcmode to AdcSingle

I think the Leaf sensor outputs an asymmetric signal. So at 0A it sits at 3.2V or so. Can be adjusted with idcofs

[maiks]

GND to GND, "Voltage" to 5V, "Signal" to CUR+
Parameter idcmode to AdcSingle

I think the Leaf sensor outputs an asymmetric signal. So at 0A it sits at 3.2V or so. Can be adjusted with idcofs

Thanks a lot

[PetersonOctavius]

i manually balanced my modules.
so now, should replace the two boards?
i want to update the working boards to the latest firmware...kinda scary cause all 6 where measuring fine on .20B ...when i went to .23B the 2 boards went bad.

[johu]

Don't know what happened on your side. Also updated my 2.0 modules after you reported and they are fine.
Maybe you can build a tester, too? 16 resistors in series. like 270 Ohms or so. Then run your battery voltage (or better a lab supply) over all of them and tap off to the BMS. There shouldn't be anymore current draw between the BMS running and not running. If there is, something is fishy

[PetersonOctavius]

my boards are always on idle.
i CAN-mapped idc to the ISA shunt like mario did, but no luck, idc was jumping all over the place.
so i CAN-tx idc from the VCU, then i CAN-rx idc on the bms with my own id, and it worked.

now bms is getting ISA current values, but via the VCU.

so how do i get the bms to balance?

bms25B-params-april19.json

(766 Bytes) Downloaded 665 times

[Proton]

@Johu
The new version boards that you sent work perfect. Now I have 6 boards that work I like the "Additive balancing option.
I upgraded the 3 old boards to 0.25 version and the 3 new ones are at 0.23B. Can I update the software while they are running or I have to do it one by one.
Also do I need to unplug the cables from the batteries before the software upgrade or I can leave them connected to the batteries?

DO you want to fix the old ones or you are going now with the new design. If you want I can send them back to you.

[Jacobsmess]

Just a thought (provided this hasn't already been addressed with hardware changes) but is there a way of stopping hardware failure when updating with HV still connected? Or even just a warning for those of us running older boards?

Or is it just a matter of updating over CAN?

[johu]

Sorry missed this.
In theory over CAN it should update flawlessly as the boot loader pulls down the mux control pins. In the latest revision hardware even swd update should be possible because I fitted pull down resistors.
In practise this has always worked on my test systems. But there needs to be more testing before we can be sure it works for everybody.
I'm working on a system with 128 cells for more test data

[johu]

so how do i get the bms to balance?

Set ubalance to 3700 or so, balance to both and idlewait to 60. Then 60s after seeing no current you should see balancing activity in uXstt

[Proton]

I'm working on a system with 128 cells for more test data

Are you going to have a screen for monitoring? I really do not like the wireless connection and I also have to find a way to display it on a screen.

[johu]

Are you going to have screen for monitoring? I really do not like the wireless connection and I also have to find a way to display it on a screen.

They are all connected to a BeagleBone, so Linux. I'm sure I'll hack something up in python and javascript

[PetersonOctavius]

Set ubalance to 3700 or so, balance to both and idlewait to 60. Then 60s after seeing no current you should see balancing activity in uXstt

balancing activity in uXstt?
i did not understand that part.

[Jacobsmess]

I've got a possibly FET related issue,
On 3 of my boards I have odd voltage readings.
U5-U8

U2-U3

U2, U6, U10-U11

Would the first port of call be to check the voltage between the drain and source on all the FETs?
On a board with correct voltage readings I get around 0.6V in diode mode on a few of the FETs but I've not checked them all.
I did swap the headers and may have damaged the solder pad in doing so but I still get continuity from the header pin to the resistor so assume all is OK there.

[Jacobsmess]

Red probe on the single pin, black probe on the lower left one (or upper right for the mirrored FETs).

your image shows the black on the lower right leg, testing mine I get nothing with red on top centre pin and black the lower left so I assume this is incorrect (as written rather than the picture). I've just tested all combinations and only one combination gives around 0.6 on most FETs with Q1 and Q5 giving 0.7V in the diode mode, I assume this is ok as it matches the 3 working boards.

I've also checked all the resistors and they all match apart from

R3 7.9kohm
R12 8.35kohm
R4 7.89kohm
R8 4.8kohm
R5 10.4kohm
R16 12.1kohm

But again, these match on the working boards so I'm a little stuck.

I tried a jumper from the connector pin to the leg of the first resistor on one board and this didn't seem to help...

Any other options? I've reflashed the hex and the binary, neither has worked.

I also have 2 boards that are failing to power up, I can flash using the STLink and the VCDG pins, and I get LED action but when I connect the GND, 12V, enable and CANH/L I dont get anything...

Any advice would be greatly welcome.

[Proton]

Just a thought (provided this hasn't already been addressed with hardware changes) but is there a way of stopping hardware failure when updating with HV still connected? Or even just a warning for those of us running older boards?

Or is it just a matter of updating over CAN?

I just updated the new v2.3 boards with the High Voltage connected. no issues.

[maiks]

I recently received six ADC16 BMS boards and an ESP32 board from the OpenInverter shop. For testing, I connected one BMS board following the instructions on the wiki:

ENAL_IN and 12V connected to a 12V power supply
GND to GND
BMS CANL to ESP32 CANL
BMS CANH to ESP32 CANH

I successfully connected to the ESP32 via Wi-Fi, accessed the web interface, and updated the Node ID to 10. However, I’m encountering the following error:

"Communication problem between ESP and STM"

Could you please assist me in resolving this issue?

Thank you!

[johu]

By default there is no termination resistor. Close the solder jumper JP1 on the wifi module to enable one.

[maiks]

By default there is no termination resistor. Close the solder jumper JP1 on the wifi module to enable one.

it has worked .thank you

[skr]

I hope we can get this stupid bug ironed out as the BMS works super well otherwise

I have bootleg copies V2.3 (BOM same or same spec substitutes from LCSC).

Out of 10 boards I assembled managed to test 5 so far, latest build fw (with sweep pattern commit).
12S test battery per module, 36S total, with 100/300mA polyfuse in front of each cell.
Had two problematic boards so far, both times in middle of 3 board chain (13S-24S).
Got 2 muxshort errrors so far, not sure what happened with the first one, could not quickly find any dead fets on it, just took another board.
Second time I was measuring first cell voltage for calibration on phoenix screw terminal with a battery operated scopemeter when all channel polyfuses went pyro on me.
also killed V10 Q74 and V12 Q87 while measuring V1.

I maybe lack the physics 101 required, but maybe some sort of capacitive magic is going on here, maybe not enough dead time and gates still have charge?
I took the next board, but not sure if random muxshort is solved with the latest rev.

[PetersonOctavius]

i had 6 boards working just fine. after upgrading to .25.25B two of them randomly got the muxshort.
so far not sure why the random muxshort. i upgraded all the other boards, they are working fine. some i unplugged the battery before updating, some i left the batteries connected. all where fine.

i found 5 fets not giving the 0.5 when i did the diode test with a multimeter. i removed them from the on board and put it on the other one.
couldnt test it cause my 5v regulator got damaged. (dumb of me for swapping 12+ and ground).
is there a replacement regulator of mouser maybe?

[skr]

lm2734 (LM2734YMK/NOPB) or other PNs in same package could maybe work. Please compare LM2734 12V->5V bom and the above BMS switcher bom if actually compatible, pins seem to be compatbile.