Hello foxBMS-team,

for some projects i searched for an open-source-BMS and found this awesome work. After getting familiar with the docs and the toolchain i tried to build the application, but encounteres some problems. After following the setup descripted in the docs i can't build the application, just getting an error message:

Build failed Traceback (most recent call last): File "E:\01_Repos\07_FoxBMS\tools\waf3-2.0.22-1241519b19b496207abef1f72bbf61c2\waflib\Task.py", line 180, in process ret=self.run() File "E:\01_Repos\07_FoxBMS\tools\waf-tools\f_ti_arm_cgt.py", line 1172, in run self.generator.bld.fatal( File "E:\01_Repos\07_FoxBMS\tools\waf3-2.0.22-1241519b19b496207abef1f72bbf61c2\waflib\Context.py", line 261, in fatal raise self.errors.ConfigurationError(msg,ex=ex) waflib.Errors.ConfigurationError: Extracted version from git repo (1.1.0) does not match version defined in waf (1.1.1).

after going back to the version 1.1.0 i'm able to build the binaries. Probably there is a small typo in some configs or scripts for WAF?

Kind regards.

Hi there,

thanks for this project. Very well structured.

Ever wondered why ADC measurements are off? And can get negative? The ADC equation in the TI TRM is formulated backwards (adc counts as function of input voltage). It has been implemented like this in adc.c. That is wrong. The function needs to be inverted first, so:

static float ADC_ConvertToVoltage(u16 adcCounts)
{
    /** For details to equation see Equation 32 in Technical Reference Manual SPNU563A - March 2018 page 852 */
    float result_mV = (((adcCounts + 0.5f) * (ADC_VREFHIGH_mV - ADC_VREFLOW_mV)) / ADC_CONV_FACTOR_12BIT) + ADC_VREFLOW_mV;

    return result_mV;
}

I have also removed the cast in the call to ADC_ConvertToVoltage and renamed that function appropriately.

What is your policy on pull requests? Do you like reports like this or would you rather have a PR?

Best

I have the following observation:

Based on the datasheet of IC PCF2131TF, pin 12 is BAR(INTB), and pin 13 is BAR(INTA). But in the schematic 'rtc.SchDoc' present in the TI TMS570LC4357 Master v1.1.1, the label assignment is interchanged. Is this correct?

Hello, there is an issue in lines 76 & 77 of the can_cbs_tx_voltage.c file, particularly regarding the start bits. Cell voltage 2 has the start bit for cell voltage 3 and cell voltage 3 overwrites cell voltage 0. The correct start bits to correctly line up with the dbc file are 11, 30, 33, 52.

Many components in the schematic of the master board TI TMS570LC4357 Master v1.1.1 are present in the PCB layout but not in the BOM.

These missing components are cross-marked in the schematics for example: In the buck-boost-converter.SchDoc

Packovercurrent is being checked for Cell Current Limit Violation at line 335 in <soa.c>. Wouldn't the pack current be much greater than Cell Current Limit. Also there exists a function for Pack current limit violation however was not able to find the usage for it.

suspect, there are two mistakes in the balancing control driver of the mxm driver. both are located in: mxm_17852.c

1st: line 408: pState->batteryCmdBuffer.msb &= ((uint8_t)1u << 7u); /* set CBRESTART bit */ because the Bit CBRESTART is active high, it should be set. actually the highByte of MXM_REG_BALSWCTRL is getting zeroed out by the command and the timeout-restart-bit is not set. also see the datasheet, page 310 and page 311 for CBRESTART -> pState->batteryCmdBuffer.msb |= 1<<7;

2nd: line 468: pState->batteryCmdBuffer.lsb = 0x0EU; -> enable balancing switches for a minimum amount of time. i would suggest to change to 0xEEu or even 0xFEu to bring the balancing duty to 93,75% or even its maximum if possible between measurements (not sure, EE works fine for me).

after this two changes the balancing driven by the max17852 works. the alternating balancing scheme between even and odd cells could be seen as well (works).