Post Go back to editing

LTC6811 reading wrong voltages

I have a custom PCB with an LTC6811 and 2 LTC3300 and connected 12 18650 cells.

The measurement at the LTC6811 input pins is very different from the readings of the ADC registers. 

Both the voltages actually measured (with a digital voltmeter) and read out from the LTC6811 are stable.

i am using RC filters at the inputs with 100Ohm/10nF as suggested and measuring voltages directly at the LTC input pins.

5V are supplied from a L5017MR buck converter located 3cm away from the LTC6811.

Measured calculated LTC_reading
C12 46,01 3,86 3,83
C11 42,15 3,86 3,823
C10 38,29 3,86 3,826
C9 34,43 3,79 3,758
C8 30,64 3,75 4,012
C7 26,89 3,83 3,232
C6 23,06 3,83 4,373
C5 19,23 3,85 3,248
C4 15,38 3,86 4,113
C3 11,52 3,821 5,727
C2 7,699 3,844 0,901
C1 3,855 3,855 3,825
46,01 44,668

I tried with different sampling speeds but the result remains unchanged. What can be the reason ?

The two LTC3300 balancers are communicating via GPIO and both read status 0x000F which seems ok to me.

I will try with fixed resistors and a constant power supply tomorrow but i doubt the result will be much better/different.

Any advice welcome !

  • Some more info:

    I set VUV: 2.576, VOV: 4.192 and all cells are reported as UV=FALSE, OV=TRUE (0xAAA)

    VD measured = 3.1684V
    VA measured = 4.3578V (a little too low)
    Sum of cells: 45.448V

    Status shows MUXFAIL=1 which i do not know how to interprete ("set to 1 if any channel decoder fails")
    Why should the decoder fail and why does this lead to erroneous readings ?

    after issuing DIAG cmd MUXFAIL = 0 but the readings are unchanged no matter if all cells or just a pair of cells is read.

    Is it possible that i destroyed the chip when powering on one cell after the other starting with lowest cell ? I would not know HOW to properly start a device other than that.

  • The config reads back as 0xFE A1 60 10 00 00 configuring 2,576V UV and 4,192V OV respectively

    and the OV flags are set - this all makes no sense to me !

  • Can you use a multimeter to measure the voltage at the C pins and compare that to the voltage measured at the 18650 cells (also with the multimeter)?  If you have zener diodes on your PCB it is important to make sure they are not unintentionally conducting a small amount of current lowering the voltage on the C pins.  This can cause low readings similar to what you're seeing.

    Looking at the "LTC_reading" column of the table would suggest that the OV and UV flags should be set. Cell 2 is reading lower than the UV threshold while cell 3 is reading higher than OV threshold.  The C2 pin or cell 2 connector wire is likely not making contact which could cause this.  This can be checked with the ADOW command, or continuity checked manually with a multimeter without the cells connected.  C5 and C7 may also have this problem but with partial contact.  Is there flux still on the board around the IC (C pins or filter resistors/capacitors)?

  • Thank you for guiding me in the right direction. I found some soldering issues and have implemented ADOW command according to the manual. Due to the fact that a capacitor network is installed (as suggested for the LTC3300) this algorithm had to be changed a little but unconnected cells are now being detected. Still i have some deviations but I will further investigate the reasons and try to understand. Is the precision of the RC filter 100Ohm/10nF affecting the measurement?

  • Hello again. I had a short also between one C input and the neighboring balance output. Now all readings seem to be fine. At the beginning I was afraid that the PCB was wrong or maybe the LTC6811 broken but finally everything seems to be fine. Opposite to the reference design I am using a doubled sided PCB only - i hope this is not a major issue when the LTC3300s will be active in a next step. Thank you for your assistance !

  • Hi,

    Thanks for the Response,

    I'm getting 0xFF only when i'm reading the Cell Voltages.

    Can you please suggest me the Debugging steps to Resolve the cell reading issue.

    Note: Able to Communicate the LTC6813(Write Configuration and Read Configuration Working fine).

    Please suggest me the Required Command Sequence(I'm giving option 1,2,3,and 4).

    Please find the attachments for your Reference.

    If you are Convenient to help this issue please ping me on Whatsapp(+91 8523851441)-So that we'll communicate directly.

    Thanks in Advance.

  • Are you sure your PEC calculation is correct ? Can you provide the frames sent/received when reading configuration pls.

  • Hi Vienna Tom,

    Thanks for the Response,

    If you are Convenient to communicate over the WhatsApp, please ping me to this WhatsApp number(+91 8523851441)-So that we'll communicate directly.

    Please find the Attachments for Configuration Frames(both TX and Rx).

    Write Configuration::

    Read Configuration::

    Start Cell Voltage Conversion::

    Poll ADC Conversion::

    Thanks for your Valuable time and Patience.

  • Hmm, with the last sequence you are trying to poll conversion status (PLADC). I am currently not at my development system but I can try to help you more deeply later this week. Oh a look at the datasheet tells differently. SDO is tied low while the conversion is taking place. So you would get a nonzero value when the conversion is completed and can then continue to read the register-groups of your cells. I would not use this method but do it as described in my next posting.