Post Go back to editing

LTC6813 responding with 0xFF for everything

I'm using the LTC6813 on a custom built PCB, and have the reference Linduino software that I got from here running on an Arduino Uno.

The VREG bus is at 5.25V and then turns off after 2s if no command is received(as expected), however, the VREF1 and VREF2 pins are at 0V.

When the evaluation software requests information from the device of any sort, the SDO is held high (0xFF) during the data part of the transmission, but a PEC of 0x66, 0x4C is transmitted as well. 

When I send the command "1" to the Linduino software (Write and Read Configuration: 1), I get this as the output - 

Written Configuration:
CFGA IC 1, 0xE4, 0x0D, 0x28, 0xA0, 0x01, 0x00, Calculated PEC: 0xAD, 0x6A
CFGB IC 1, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, Calculated PEC: 0xC2, 0x12

Received Configuration
CFGA IC 1, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, Received PEC: 0x66, 0x4C
CFGB IC 1, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, Received PEC: 0x66, 0x4C

Reading cell voltages in a loop (command 11) gives me -

Cells, 6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,6.5535,

I am unsure what is going on since obviously the chip is responding to commands with messages that have a valid PEC value. 

Thanks in advance, I'm happy to provide any more information that you'd like, including schematics of the board to find any hardware issues, if there are any. 

Chai

  • Hi Chai, 

    As per the data provided , we see that ADC conversion is not happening as Vref1 & Vref 2 are 0V, They should be in the range as mentioned in the datasheet page no:5.

    If you can share your schematic we can try to understand the issue.

  • There are 2 LTC6813 chips on the board one to measure cell voltage and the other to measure the cell temperatures. The schematic symbols of the chip used to measure cell voltages are named as Cxxx, symbols for chips measuring temperature are labelled Txxx. The TKH45 resistor and the temperature measuring LTC6813 have not been soldered yet. The V+ net is directly connected to the highest cell. Everything else is as below. The BJTs are part of a dual BJT package(PHPT610030NK). DTEN is tied to VREG.

    J3 ties ISOMD to GND and J5 is connected to the Arduino, the transformers are not soldered as well as R7(the resistor connected to CSB). R3 and C10 are soldered to reduce noise on the SCK line that we observed when they hadn’t been placed.

    Voltage measurements when command 11 is sent over Serial to an Arduino Uno running the DC2350_6813 sample sketch mentioned in the original post-

    VREG (Capacitor C3) is at 5.25V. However, every 2 seconds, the voltage drops to 4.4V before quickly recovering to 5.25V.

    Capacitors  C1 and C2 have 0V across them at all times with no noise or glitches even during startup.

    Current measurements –

    With SPI inputs floating, the current hovers around 2mA, VREG is at 5.2V and the WDT doesn’t time out. No noise is observed by a scope on any of the SPI lines, all are at 0V. The voltage drops every 2s are more pronounced, dropping to under 2.5V, but are very inconsistent.

    When the board is powered up while connected to an Arduino that is already powered on, but with no commands sent to the chip, there is a 0.3mA spike when power is initially supplied which falls to 25uA for 2s and then to <0.1uA(multimeter cannot measure any lower) after that.

    When the chip is active but not measuring voltage, current draw is around 0.1mA. During measurement, current is at around 23uA and with spikes to 650uA during measurements. The average current when measuring every 1s is 100uA.

    Chai

  • I also found this elsewhere on the forum where someone is having the same issue on the LTC6804 and LTC6811 parts

    ez.analog.com/.../ltc6804-2-and-ltc6811-2-reference-voltage-1-and-2-at-0v

  • Hi, 

    1. VREF1 can be observed when ADC Conversion is happening. It can be measured using DSO.

    2. From the Schematics i cannot see a pull up resistor on SDO( MISO) line. So if there is no pull up then communication will not happen.

    Please confirm for the above 2 points.

    Phani

  • Hi,

    1. I am unsure what a DSO is. I was measuring the voltage across the filter cap with an oscilloscope and could not see any voltage.

    2. The communication is working correctly, I am receiving a valid PEC. However, the data frame is just repeated 0xFF. The pull up is enabled internally on the microcontroller.

    UPDATE: We have working communications occasionally with no obvious change in configuration

    Chai

  • Hello Chai - 

    You are missing a step. You have configured the device and are reading a results register, but before you read the register you need to tell the LTC6813 what you want to measure - and initiate a measurement - by sending an ADCV command. So, the sequence looks like this: 

    1.  WRCFGA (Write Config)

    2. RDCFGA ( Read Config )

    3.  ADCV (starts measurement- you should see the supply current on Vreg bump up to about 15 mA, and the Vref1 should go to approximately 3.15 V, and Vref2 will go to about 3.0V) 

    4.  RDCVA, RDCVB, etc. (read results of measurement) 

    Regards, 

    John Smith

  • Hi Chai,

    Please find the latest sketch book code for LTC6813. Try using this code.

    I hope, this should resolve your problem. 

    6813_SKETCHBOOK_Code.zip

  • Hi, 

    I apologize for the delayed response, I had some hardware issues with the design that I had to contend with first.

    On step 3 of the above sequence, where the VREFx voltages are supposed to come up, nothing happens. The VREF voltages stay at 0V and the ADC measured values are completely arbitrary which is likely a result of an incorrect VREF voltage. I am unsure what might be causing this issue

    Thanks

    Chai

  • I used the code in the sketchbook above and this did not rectify the issue. 

    However, I believe I have identified the issue with the board, but am completely lost for a solution. I used a redesigned board with some minor hardware changes including an LED that indicated power status as well as one to indicate the WDT status. These two resulted in my observation below. 

    When the LTC6813 sketchbook is on command 10 or 11(pull voltage data in a loop), the power LED turns on when VREG goes to 5.3V and stays there for 2 seconds. The small dip in voltage I was observing before at the end of 2 seconds was now a drop to 1.4V, approximately the forward voltage of the indicator LED, likely due to the much higher current draw from LED compared to only the LTC6813 before. Moreover, the WDT pin goes into high impedance and the LED connected to it briefly turns on as well. This would indicate that the WDT expires and the chip enters STANDBY, but right after it does so, a communication signal from the sketchbook code causes it to turn on again. 

    I can only attribute this to the chip perhaps not recognizing the command from the sketchbook, and thus, deeming it an invalid message, allows the WDT to expire. I am not sure why this is happening or how I may rectify the issue, but I have not been able to communicate with any of the 5 LTC6813 chips I have bought so far and am starting to believe that perhaps I received a faulty early batch and am suffering the consequences. Will I have to buy a new batch of chips to test or is there perhaps some bug in the sketchbook itself or the setup of the code before compiling it that I have missed?

    Hoping I receive a quick reply since time is running out for me on this project and I must move on to further testing ASAP

    Thanks and Regards

    Chai

  • Hi,

    I am having the exact same problem as you. Did you solve it in the end?