AEK_POW_BMSNOTX (L9963E) cell voltage measurement

[hno-wires]

Evaluating the L9963E BCC chip for a BMS design, using the AEK_POW_BMSNOTX eval board together with AEK-MCU-C4MLIT1 (spc58ec) running the example BMS application as reference but seeing some odd results in the cell voltage readings.

All readings are a bit low compared to actual cell voltage, but cell1 and cell14 are off by a fairly large amount.

What could be causing this?

cell1 & cell14 reading off by 0.5v
cell2 to cell13 reading off by about 0.1v

Cell1 reading seems to improve if the HotPlug selection jumper is mounted. But not cell14.

[SRomeo · Accepted solution]

Hi Henrik,
I'm assuming you are using the AEK BMSHOLDER mockup, running on your C4MLIT1 one of the demo application available on AutoDevKit studio.

Usually, when one of our BMS board perceives a consistent shift, the device is actually missing the ground reference.
Please follow this procedure:

 1) Disconnect the BMS holder connector
 2) Check both your BMS boards and place each Jumper JP1 in the 1/2 position:

  3) Reconnect the BMS holder connector.

Note: Sometimes, even with the jumpers, this problem happens. That's because the BMS connector is not inserted correctly. Try to reconnect it by pushing along the borders, uniformly, as in the picture:

To check if the measurement are right I suggest to compare voltages as in the picture:

If you have any doubt, please take a look to the schematic.
You can find it on the UserManual

Best regards,
AEK_Team

[hno-wires]

Yes I am using the BMSHOLDER during evaluation/test.

Measuring cell0-cell14 pads on shows the exact same voltages as on the batteries. And the reported value is consistently 0.10-0.11V below the separately measured battery voltage.

The cell1/cell14 confusion was a combination of missing the hotplug selection jumper and a measuring error. With the jumper mounted the reading of both cells is in line with the other cells.

The connector is properly seated as far as I can tell. 

[SRomeo]

Hi there,

If you look at the scheme present on the UM you should notice the presence of 100 Ohm resistors between Battery contacts and L9963E/F's Cx pins. Using RLPF is the proper way to connect battery cells to the AFE.
You may find the application scheme on L9963E device datasheet. 

You may check this VR4 voltages. If this is not zero, the voltage perceived by every Cx pin is reduced by a little bit. But the device handles it. Take a look to L9963E datasheet to understand more

Best regards,
Simone
AEK_Team 

[hno-wires]

I am reasonably familiar with the schematics, and those RLPF resistors should not impact the voltage reading. There is no meaningful current on the voltage reading pins Cx, only the balancing pins Sx/Bx_x.

Confirmed by measuring over the ESD capacitors. Measuring cell C13 gives 3.90V, measuring over the ESD capacitor between C12-C13 pins (component C17 on the eval board) gives 3.89V. But the reading from L9963E is 3.78V. And the same pattern for all cells. And the same offset with or without balancing active for the cell.

Just noticed the a similar offset is also seen for VTREF. Measuring on the board gives 4.96V but the reported value is 4.81V.