BQ76940: BQ4050 reading fail and wrong working

David Kim

Prodigy 50 points

Part Number: BQ76940
Other Parts Discussed in Thread: BQ76930, TIDA-01093, TIDA-010030, ISO1541

Hello,

I designed BMS for 26S with TIDA-01093. But I have diffierence results as followings,

down part of bq76940, all of cell volt between each VC is correct when regsrc is not used.

but when regsrc is used, VC15-VC14 : 2.10V, VC14-VC12:-0.698V, VC11-VC10B:-0.75V.

when regsrc is not used and all of cell volt(VC) are correct, reading value is different from input voltage.

Here are results form I2C reading

N[1], Cell Voltage = 3439 mV
N[2], Cell Voltage = 3449 mV
N[3], Cell Voltage = 3449 mV
N[4], Cell Voltage = 3448 mV
N[5], Cell Voltage = 3422 mV
N[6], Cell Voltage = 3379 mV
N[7], Cell Voltage = 3465 mV
N[8], Cell Voltage = 3450 mV
N[9], Cell Voltage = 45 mV
N[10], Cell Voltage = 3412 mV
N[11], Cell Voltage = 119 mV
N[12], Cell Voltage = 1919 mV
N[13], Cell Voltage = 3429 mV

I discharged cell1, cell13. cell1 channel is correct, cell13 channel is not correct because value of cell13 is not reduced when I discharged.

I attached my schematic which is reffered from TIDA-01093.

bms_bps.pdf

over 2 years ago

0 David Kim over 2 years ago

Prodigy 50 points

Also I have problem with regsrc(25V) and regout. without load from regout, sometime regout is 0V when I connect 13 cells to bq76940. sometime regout is 3.3V.

what's wrong with bq76940? is there any reason?

0 Matt Sunna over 2 years ago in reply to David Kim

TI__Guru* 78803 points

Hi David,

The REGSRC circuit looks correct on the upper device, but it does not look correct on the lower device. For example, REGSRC does not have the required capacitor connected. I also see some issues on the cell connections. See this video around 7:30 for some important points: https://training.ti.com/how-create-schematic-bq76920-bq76930-and-bq76940 I think some of your connections are not correct.

Best regards,

Matt

0 David Kim over 2 years ago in reply to Matt Sunna

Prodigy 50 points

I attahced modified files from 15cells with bq76940.

Is it ok for 13cells?

From TIDA-01093, all of the regsrc are connected capacitors.

do I have to connect VC5x label to VC5x pin and VC10x label to VC10x pin? They already connected each other via diode(VC5 pin to VC5x pin, VC10 to VX10x pin).

they really make me confuse. I also add my modified schematic. please help me to be done.

5466.BQ76940 for 13 cells.pdf

bms_bps_v1.1.pdf

0 Matt Sunna over 2 years ago in reply to David Kim

TI__Guru* 78803 points

David,

The top device has the correct REGSRC circuit:

The bottom device does not. REGSRC should be supplied from the source follower circuit and have a 1uF capacitor:

On the upper device you have the circuit below. Does VC5X_U connect to the VC5X pin somewhere? Does VC13X_U connect to the VC10X pin somewhere? You need to have 3 of these for the BQ76940 (see the BQ76940 datasheet example) since it has 3 cell groups (VC5X, VC10X, BAT). TIDA-01093 uses BQ76930 which only has 2 cell groups.

Regards,

Matt

0 David Kim over 2 years ago in reply to Matt Sunna

Prodigy 50 points

Dear Matt,

I did modified pcb and connect as your comment. But bq76940 chip didn't work for REGOUT and commnucated with I2C. I think the chip is broken.

Therefore, I attached modified circuit.

And I tested BQ7694003DBTR before, now I tested BQ7694006BTR.

Q1. Is there any difference between BQ7694003DBTR and BQ7694006DBTR except CRC function?

Q2. is it possible to use BQ7694006DBTR instead of BQ7694003DBTR?

I just followed TIDA-01093

you mentioned I need to connect 1uF cap

Please help me out to be done. I couldn't do anything without your help.

Could you modify my attachment?

0 David Kim over 2 years ago in reply to David Kim

Prodigy 50 points

Dear Matt,

I found the reference TIDA-010030 with bq76940 for 13 cells. they didn't use diode between VC5 and VC5X, between VC10 and VC10X.

But There are diode between VC5 and VC5X, VX10 and VC10X on TIDA-01093.

which one is correct between TIDA-01093 and TIDA-010030?

Best regards,

0 Siddharth Sundar over 2 years ago in reply to David Kim

TI__Prodigy 250 points

Hi David,

Unfortunately, Matt is on vacation currently. He will be able to respond to your queries by march 31.

Thanks,Sid

0 Matt Sunna over 2 years ago in reply to Siddharth Sundar

TI__Guru* 78803 points

Hi David,

The diodes are required. And they are present on TIDA-010030 on page 3.

Best regards,

Matt

0 David Kim over 2 years ago in reply to Matt Sunna

Prodigy 50 points

Dear Matt,

And I tested BQ7694003DBTR before, now I tested BQ7694006BTR.

Q1. Is there any difference between BQ7694003DBTR and BQ7694006DBTR except CRC function?

Q2. is it possible to use BQ7694006DBTR instead of BQ7694003DBTR?

0 Matt Sunna over 2 years ago in reply to David Kim

TI__Guru* 78803 points

As the datasheet shows, these two devices have a different I2C address and the CRC is enabled on the 03 but not on the 06. It should be possible to use both if changes are made to the MCU code to account for these two things.

Regards,

Matt

0 David Kim over 2 years ago in reply to Matt Sunna

Prodigy 50 points

Dear Matt,

Thanks for your quick reply even though you are on vacation.

Happy vacation lol.

0 David Kim over 2 years ago in reply to David Kim

Prodigy 50 points

Dear Matt.

I had failed to impelemt bms with bq76940 for 26s1p. Now I modified all from documentation. Please check connection of my design because I already faced troble with PCB.

Thankse-bike_bms.pdf

0 Matt Sunna over 2 years ago in reply to David Kim

TI__Guru* 78803 points

Hi David,

This looks much better. See my notes below:

R200R should be 0 Ohms (it is hard to see if this is correct in schematic)
Looks like C91 and C28 are the same - only use one
Looks like C97 and C69 are the same - only use one

Best regards,

Matt

0 David Kim over 2 years ago in reply to Matt Sunna

Prodigy 50 points

Dear Matt,

I modified 26 cells design with your recommandation and TIDA-010030 and then checked several time.

TIDA-010030 used N channel fet for cell balancing but I used P channel fet because TIDA-01093 used P channel fet. But I had same result before.

I attached modifed schematic.

e-bike_bms_v2.pdf

I read cell voltage with block reading as followings,

N[1], Cell Voltage = 2656 mV hex [1b03]
N[2], Cell Voltage = 2755 mV hex [1c16]
N[3], Cell Voltage = 2983 mV hex [1e18]
N[4], Cell Voltage = 2876 mV hex [1d11]
N[5], Cell Voltage = 2957 mV hex [1e18]
N[6], Cell Voltage = 3191 mV hex [20a2]
N[7], Cell Voltage = 3260 mV hex [21a5]
N[8], Cell Voltage = 3269 mV hex [21a5]
N[9], Cell Voltage = 42 mV hex [42]
N[10], Cell Voltage = 3270 mV hex [21a5]
N[11], Cell Voltage = 3218 mV hex [20a2]
N[12], Cell Voltage = 992 mV hex [97d]
N[13], Cell Voltage = 40 mV hex [42]
Batt Voltage = 31434 mV

I read N[9] cell voltage as following,

Cell Voltage = 42 mV

Therefore, there is no error in code for reading data from bq76940.

Best regards,

David

0 Matt Sunna over 2 years ago in reply to David Kim

TI__Guru* 78803 points

Hi David,

Did you try to read N[14] and N[15]?

N[9] and N[13] are shorted in the schematic, so you may need to adjust which registers you read the cell voltages from.

Best regards,

Matt

0 David Kim over 2 years ago in reply to Matt Sunna

Prodigy 50 points

Dear Matt,

Thansk for your advice for cell reading.

I had curious problem with bq76940.

When I activate lower bq76940 with TS1 for using REGOUT for CPU power source, cell 12, cell13, cell25 and cell26 are measured -0.7V.

But I used external 3.3V for CPU, lower bq76940 works well.

upper bq76940 had same problem like lower bq76940, when I used REGOUT which is used pull-up. If I used external 3.3V power souce, it's ok.

I had another problem with lower bq76940.

D19 anode measured 21.3V from ground but D19 cathode mesured 44.3V.

D20 anode measured 22.3V from ground but D20 cathode mesured 21.5V.

D21 anode measured 16.92V from ground but D21 cathode mesured 16.8V.

Are bq76940 broken?

Best regards,

David

0 Matt Sunna over 2 years ago in reply to David Kim

TI__Guru* 78803 points

Hi David,

This behavior is strange. How much current is REGOUT driving? Is it within the datasheet limits? Is it being used to drive any other circuit besides the CPU?

Best regards,

Matt

0 David Kim over 2 years ago in reply to Matt Sunna

Prodigy 50 points

Hello Matt,

The design is almost same as TIDA-01093. CPU usally used 3mA. If CPU start to acess bq76940, VC14 and V15 are measured -0.7V with multimeter.

upper bq76930 is same phenomenon as lower bq76940. upper side used REGOUT for drive of ISO1541 which used 3.5mA maximum.

ISO1541 is used in TIDA-01093. when I used external 3.3V for upper and low side, VC14 and VC15 works well.

I had another problem with lower bq76940. I had same proble with two pcb as folloing even though I used external 3.3V.

D19 anode measured 21.3V from ground but D19 cathode mesured 44.3V.

D20 anode measured 22.3V from ground but D20 cathode mesured 21.5V.

D21 anode measured 16.92V from ground but D21 cathode mesured 16.8V.

Are bq76940 broken?

Best regards,

David

0 Matt Sunna over 2 years ago in reply to David Kim

TI__Guru* 78803 points

Hi David,

Can you try to replace one of the BQ76940 devices to see if the problem remains?

Matt

0 David Kim over 2 years ago in reply to Matt Sunna

Prodigy 50 points

Dear matt,

I replaced one of the BQ76940 device. lower bq76940 is solved. But I still have problem VC14 and VC15 when I used REGOUT.

I tested REGOUT of lower bq76940 with electric load and multimeter for using external 3.3V. If I used 2mA~10mA it had same phenomenon which has problem VC14 and VC15.

Finally, I found my fault from circuit, REGSRC connected with D19 anode. After connection to Cell13, It works well.

Best regards,

David

0 Matt Sunna over 2 years ago in reply to David Kim

TI__Guru* 78803 points

Hi David,

Glad to hear this problem is solved!

Matt

0 David Kim over 2 years ago in reply to Matt Sunna

Prodigy 50 points

Dear Matt,

Is It same functionality to perform cell balancing circuit between p_channel and n_channel as following picture?

n channel fet p channel fet

If I want to perform cell balance for Cell0-Cell 1, VC1–VC0 is enabled by setting [CB1], Therefore, Q1(n channel fet) is trun on, Q17B(p channel fet) is turn on.

Best regards,

David

0 Matt Sunna over 2 years ago in reply to David Kim

TI__Guru* 78803 points

Hi David,

Yes, it should be the same register setting in both cases. The paragraph above Figure 5 and Figure 6 in this application report explain this in more detail: https://www.ti.com/lit/an/slua749a/slua749a.pdf

Best regards,

Matt

0 David Kim over 2 years ago in reply to Matt Sunna

Prodigy 50 points

Dear Matt,

Thanks for your reply.

I couldn't understand the charge detect function. I measured the voltage as following picture between charger connection and disconnection.

I connected common probe of multimeter to BAT-.

2.9MOhm is meausred between BAT- and BAT-. DISCHG and CHG FET are turned off. REFERENCE_TIDUC43A mentioned that when charger is connected CHG_DET can be 2.8V.

I think after detecting CHG_DET, MCU turns on CHG FET. Also after detecting LOAD_RM_DET, MCU turns on DCHG FET.

After CHG and DCHG FET are turned on, if I connect LOAD or Charger. That will be cause of spark and demage.

Best regards,

David