LMT84: LMT84 interfaced with MSP432P4011 running TI-RTOS configured through sysconfig

Shreyas Kulkarni26

Part Number: LMT84
Other Parts Discussed in Thread: SYSCONFIG, CC2640R2F, CC2640, CC2650

Hello sensors forum,

I'm facing some difficulty in getting the LMT84 temp. sensor to work properly with MSP432P4011 running TI-RTOS, with peripherals configured through sysconfig. I have asked a related question in the MSP forum but it hasnt been resolved so posting it in the sensors forum too.

My previous post -

A summary - I have a custom PCB with LMT84 connected to the MSP432 ADC pin (P5.4/ADC14.A1). I'm using internal VDD (+3.3V) reference to do a single ended measurement. I have configured the ADC using sysconfig tool, which does not allow me to control the sample/hold time etc.

My circuit schematic is as below -

My issue -
The LMT 84 is outputting a steady 888 to 910 mV output (measured using DMM) corresponding to the correct room temperature found by (1035 - Vout) / 5.5. But my MSP432 ADC reads the same very inconsistently. I get readings all over the place including up to 1000 mV. The average reading is about 50-60 mV higher which completely throws off the temperature reading. This is repeatable across multiple identical PCBs we have manufactured (machine assembled).

I understand that there are some considerations about the RST resistor I have set to 0 ohms in the above schematic. The LMT84 datasheet recommends a 3k resistor if the loading capacitance is more than 1100 pF. However I looked at the MSP432 ADC spec and it seems that the loading capacitance on the ADC is < 15 pF (Table 5.27 on page 70 of MSP432P4011 datasheet), which means my total loading is < 1100 pF. My understanding is that I should not be needing RST to be 3k.
However I did try with RST = 3k and got somewhat better results, but not quite to what I was expecting. There is still a large offset and some variance in the measurements.

We have used the same circuit with a CC2640R2F which has a similar SAR ADC. In that design, I was able to configure the ADC's sample and hold time (this was before sysconfig) and that worked reasonably well.

Is there some obvious flaw in the approach we are taking? The multimeter measurements match well, it's just an issue with the ADC being able to read those accurately.
Would appreciate some insight on this!

-Shreyas

over 3 years ago

0 Ren Schackmann over 3 years ago

TI__Mastermind 35855 points

Hi Shreyas,

The way you've drawn the schematic suggests you've placed 100pF near the LMT84. I agree with your assessment of the series resistor, but the situation may change as the ADC begins to sample. It may be the ADC that needs capacitors to stabilize the voltage at its input. How long is your wire/trace from the LMT84 to ADC?

Does the ADC have a recommendation for external capacitors? Do they appear in application circuits for the ADC? If you've had success configuring sample time before, why don't you implement that now?

thanks,

ren

0 Shreyas Kulkarni26 over 3 years ago in reply to Ren Schackmann

Intellectual 290 points

Thanks Ryan for the response.

That's an interesting point you bring up. The trace connecting the cap to the ADC is about 2.3 cm. I did not find any external cap recommendations, but perhaps I can bump up the CFT cap from 100 pF to somewhere around 2-3 nF, along with the 3k recommended resistor and see if the performance improves.

Regarding the sampling time, in the other MCU, (TI's CC2640) there was an explicit ADC configuration script that allowed us to change the sampling time. However with the MSP432, they have provided a tool called sysconfig which is GUI based and I'm unable to configure some of these aspects through it.

0 Josh Wyatt over 3 years ago in reply to Shreyas Kulkarni26

TI__Guru* 90906 points

Shreyas -

Please see this SLA, it may help you out:

https://dev.ti.com/tirex/explore/node?node=ACiPjIrJJP9mC8aGF1ePdw__z-lQYNj__LATEST

0 Shreyas Kulkarni26 over 3 years ago in reply to Josh Wyatt

Intellectual 290 points

Thanks Josh, appreciate the response.

However, not quite. I'm looking to use the ADC rather than the ADCBuf driver. Do you know which file I can play around for modifying the ADC sampling duration?

Previous to sysconfig, I did this for CC2650 in CC2650DK_7ID.c file as follows-

   {
    //Application Specific ADC
       .adcDIO = Board_DIO29_ANALOG,                                        /* This ADC is specific Application board */
       .adcCompBInput = ADC_COMPB_IN_AUXIO2,
       .refSource = ADCCC26XX_FIXED_REFERENCE,
       .samplingDuration = ADCCC26XX_SAMPLING_DURATION_42P6_US,
       .inputScalingEnabled = true,
       .triggerSource = ADCCC26XX_TRIGGER_MANUAL
   }

I cant find an equivalent file or a way to do this through sysconfig for MSP432P4011

-Shreyas

0 Josh Wyatt over 3 years ago in reply to Shreyas Kulkarni26

TI__Guru* 90906 points

Shreyas -

Here in the lab https://dev.ti.com/tirex/explore/node?node=ACiPjIrJJP9mC8aGF1ePdw__z-lQYNj__LATEST

it shows that the sampling rate can be set in:

adcbufdifferential.c , located in the MSP432SDK (for example i have v4.2, so path is: C:\ti\simplelink_msp432e4_sdk_4_20_00_12\examples\rtos\MSP_EXP432E401Y\drivers\adcbufdifferential

Here is the area to look for:

/* Set up an ADCBuf peripheral in ADCBuf_RECURRENCE_MODE_CONTINUOUS */

ADCBuf_Params_init(&adcBufParams);
adcBufParams.callbackFxn = adcBufCallback;
adcBufParams.recurrenceMode = ADCBuf_RECURRENCE_MODE_CONTINUOUS;
adcBufParams.returnMode = ADCBuf_RETURN_MODE_CALLBACK;
adcBufParams.samplingFrequency = 200;
adcBuf = ADCBuf_open(CONFIG_ADCBUF_0, &adcBufParams);

0 Shreyas Kulkarni26 over 3 years ago in reply to Josh Wyatt

Intellectual 290 points

Thanks Josh.

Just wondering if ADCBuf and the regular ADC are the same peripheral? I am not using the ADCBuf, but just he regular ADC driver.

0 Josh Wyatt over 3 years ago in reply to Shreyas Kulkarni26

TI__Guru* 90906 points

Shreyas -

Found you have another post on this topic, for this MCU, that is in progress. Makes more sense here to stick to MSP Forum and SDK

https://e2e.ti.com/support/microcontrollers/msp430/f/166/t/944543?tisearch=e2e-sitesearch&keymatch=ADC%20MSP432

Moving responsibility to MSP Forum.

0 Shreyas Kulkarni26 over 3 years ago in reply to Shreyas Kulkarni26

Intellectual 290 points

I swapped out the capacitor with a 4.7 nF and now I'm getting stable and quite accurate measurements from the ADC. I guess the sourcing circuit wasn't strong enough to charge the internal ADC capacitance without fluctuating, hence leading to unstable measurements.

Sensors

Sensors forum

LMT84: LMT84 interfaced with MSP432P4011 running TI-RTOS configured through sysconfig