L99BM114: Wiring clarification for Unmounted Cells (5S configuration) - Section 6.8

[_gabrielfbdiniz_]

Hello everyone,

We are currently designing a decentralized Battery Management System (BMS) node for our Formula SAE electric car. We selected the L99BM114 due to its robustness in automotive environments and high measurement accuracy.

Our architecture requires this specific node to monitor a 5-series (5S) cell module. Since the L99BM114 supports up to 14 series cells, we will have 9 unused channels.

While reviewing Section 6.8 of the datasheet (Unmounted cells), the exact hardware implementation for our case felt a bit ambiguous. We want to ensure our schematic doesn't trigger false internal diagnostics, such as Cell Open Wire or Balancing Short/Open.

Could anyone clarify the recommended wiring for the unused pins in a 5S configuration? Specifically:

1. Voltage Sense Pins (C6 to C14): Should all these unused inputs be shorted together and tied to the highest active potential (C5)?

2. Balancing Pins (S and B) & Filtering: How exactly should we handle the external filtering resistors (RLPF) and the S/B balancing pins for these 9 unused channels? Should they be left floating, or is there a specific pull-up/pull-down tie-off required?

If there is any reference schematic or application note that covers a reduced cell count configuration for this IC, that would be incredibly helpful.

Thanks in advance for the support!

[Max VIZZINI]

Hi,

You need to connect all sensing and balancing pins need to be shorted in series or the IC will not properly work.

The first 2 and the last 2 cells HAVE to be connected, and minimum number is 4 cells. 

I attach below a project we did for 6 cells battery.

Best Regards,

AutoDevKit Team

[Max VIZZINI · Accepted solution]

Hi,

This could be for 5 cells:

When building your new board, please follow the yellow box highlighted in the datasheet, all sensing and balancing have to be short-circuited at pcb level.

Best Regards,

AutoDevKit Team

[_gabrielfbdiniz_]

Hi Max,

Thank you so much for the detailed explanation and the diagrams! This is a lifesaver.

Since we are designing our own custom PCB for the Formula SAE team rather than using a general-purpose 14S evaluation board, I want to make absolutely sure we implement the "yellow box" guidance correctly at the schematic level.

I've updated our KiCad schematic based on your feedback:

1. We split the active cells into a top block (C14 down to C11) and a bottom block (C2 down to C0).

2. The physical connection between the top and bottom blocks is made by routing the C11 input to the C2 input.

3. All the unused sensing and balancing pins (C10 down to S3) are shorted together and tied to the C2 node through a 1kΩ resistor.

Could you please take a quick look at the attached image and confirm if this is the correct way to handle the hardware for a custom 5S board?

Thanks again for the excellent support!

[SRomeo]

Hello Gabriel,
The scheme you shared differs from the suggested reference.
You may take a look to the application example of L9963E datasheet (page 133)
Another reference may be the AEK-POW-BMSCCTX (page 4)
I add the pic for comparison:

• FBx positioning differs from reference

Placing Ferrites (FBx) immediately after the filtering Capacity protect from EMI the battery and the corresponding Cx IC route.
With your configuration, in case this protection disrupts, the corresponding Cx line won't be fully decoupled. (Moreover, if manual balancing is enabled, this may also land to unexpected current paths)

There is also some ambiguity on the labels. You are using the same name for Cx pins and battery pins. Make sure to fix that. Moreover:

• Your Cx connection is wrong.
  To mantain electical continuity you actually need to short every Unused Cx pin. 
  (To mantain 100% compatibility with our reference design 100ohm coupling resistor should be added) 
• Connect every unused Sx and Bx pin to C2 (the higher connected Cx of your second block) via a 1K resistor.
  This is done to minimize noise that floating pins may perceive. 

Make sure to report clearly on your schematic these infos.

Best regards,
Simone
AEK_Team

[_gabrielfbdiniz_]

You are absolutely right! Good catch on those details.

I have updated the schematic based on your suggestions: I changed the net labels to 'Battery X' to avoid any confusion, added the 100Ω resistors to the unused Cx pins, and corrected the positioning of the ferrite beads.

I've attached the final version of the corrections below. Thank you so much for the tips and for taking the time to review this so carefully!

Best regards,
Gabriel

[kl1]

Hello, my apologies if this should be a different post, but I'm in a very similar situation. I have a 5s pack, but I currently have the STEVAL-BMS114TX evaluation board for the L99BM114. Since the chip is on an eval board, there isn't any way to short-circuit the sensing and balancing pins at the PCB level. Could I please ask for your advice on what to do in this situation? 

Thank you very much. 

[Max VIZZINI]

Hi,

If 'PCB level' means on the STEVAL-BMS114TX the answer is no.

We are going to release a new board with control for flexible number of cells (6 or 10 or 14). In this board you have dedicated drivers for the different configuration.

Target is to have it in store by end of June, lookout for AEK-POW-BMS63EM.

Regards,

AutoDevKit Team

[kl1]

Hello, 

I appreciate the quick response. The manual for the STEVAL-BMS114TX says that it can support between 4 to 14 cells, so in the case of 5s cells, I'm wondering what to do with the unused cell inputs?

Thank you.

[Max VIZZINI]

Short them in series. See schematic in this same post.

AeK Team