CC2340R5: CC2340R5 Errata Revision D ADC_09

Hi Jay,

Can you clarify "deviation of temperature"?

The accuracy of the of the ADC temperature diode is +/- 3 C if that's what you're asking.

Best regards,

Bun

Hi Bun Kuch,

Thank you for the reply.

"deviation of temperature":

From CC2340R5 Errata Revision D ADC_09, " The magnitude of the jump is 64 LSBs higher or lower than the expected ADC output when ADC is used in 12-bit resolution setting."  

Suppose "64 LSBs jump"  doesn't happen, after calling TempDiodeGetTemp and caculating, we get 20 centigrade

If  "64 LSBs jump" happens, what could be the temperature we get?

"The accuracy of the of the ADC temperature diode is +/- 3 C if that's what you're asking."

I knew this from the CC2340R5 datasheet pdf.

Another 2 questions

From CC2340R5 Errata Revision D ADC_09, 

Q2)  This can be done by ?

HWREG(ADC_BASE + ADC_O_DEBUG1) |= 0x600;

Q3)  Every time we measure the temperature, ADC.DEBUG1:CTRL[10:9] bits should be set high before calling TempDiodeGetTemp(), right?

Hi Jay,

1. A 64 LSB jump would implied +/- 2.27C variation, which is calculated by taking the temperate range from -40 to 105 C giving us 145.

145/(2^12), where 12 is the number of bits in the ADC giving us 0.035C per quantized step. 0.035C * 64 LSB = 2.27C

2. I'll have to refer to our software engineer to help answer this question

3. same as #2

Best regards,

Bun

Hello Jay,

2)

Just to confirm:
ADC_BASE is 0x40050000 

ADC_0_DEBUG1 is 0x00000E20

I am a little unsure where we got 0x600 as the "high" value, any extra context here?

3) It seems like this setting is set once. 

Thanks,
Alex F

Hi Alex,

Thank you for the reply.

2) why 0x600 is used? It's from the CC2340R5 Errata Revision D ADC_09 part 

I supposed “hight” means the bit9 and bit10 should be 1. so the value is 0x600.  

Is this correct?

3) In my application, between two temperature measurements, MCU would enter deep deep sleep mode.

So Every time before temperature measurement, ADC.DEBUG1:CTRL[10:9]  should be set, right?

Hi Bun,

Thank you for the reply.

1) For the hardware part, I think it's clear. But I'd like to ask further for the software part.

In function int32_t TempDiodeGetTemp(void), for every temperature measurement, TempDiodeGetTemp(void) would read 12-bit ADC 10 times, 2 dummy read, 4 for high side of the diode(afterwards, get an average value) and 4 for low side of the diode(afterwards, get an average value) . 

If the best case for ADC is +/-2.27 Centigrade, Could I expect the variation from the application side is +/-2.27/4 ?

If you could help find a SDK engineer for this question, that'll be good.

Thank you

Jay

Hello Jay,

Jay JI said:

2) why 0x600 is used? It's from the CC2340R5 Errata Revision D ADC_09 part 

-Thank you for the clarification, this makes sense now, and I also got 0x600, your HWREG write here should work! 

Jay JI said:

So Every time before temperature measurement, ADC.DEBUG1:CTRL[10:9]  should be set, right?

-Lets try it like this for the first test, as long as the device does not completely reset the value should stick, but there should be no harm in setting (or checking) every time. 

Jay JI said:

If the best case for ADC is +/-2.27 Centigrade, Could I expect the variation from the application side is +/-2.27/4 ?

-As you stated I may need to reach out to the team that developed this function, however we would have 4 measurements with a +/- 2.27 variation for each individual measurement I assume. 

Thanks,
Alex F

Alex Fager said:

-As you stated I may need to reach out to the team that developed this function, however we would have 4 measurements with a +/- 2.27 variation for each individual measurement I assume. 

Thank you for helping. Looking forward to the further information.

Hello Jay,

I have reached out to the internal team for more information about this function in regard to your question! Currently awaiting feedback.

Thanks,
Alex F

Hello Jay,

I have been connected with the Indvidual familiar with the function, currently still awaiting their response to your question. 

Thanks,
Alex F

Hello Jay,

As an update the internal is looking into your question, seems like a second order polynomial correction is applied software wise to this function, going to get the data-set measurements next.

Thanks,
Alex 

Hi Alex,

Sorry for pushing, could I ask if there're updates for this question?

Thanks

Jay

Hi Alex,

Could I ask if any updates?

Best regards

Jay

Hello Jay,

Sorry for the delay, I have reached back out to the team to see if we can get that data to you. 

Thanks,
Alex F