QD2/4: how much reads the internal temperature sensor @ 5V, 115°C ???

[fat_wombat]

Hello everyone!

Being quite happy with the QD2 (which controls a demanding automotive application in my design), I have one last question.

In order to protect the electronics, the QD2 must shut down, as soon as the PCB surface temperature exeeds 115°C. Some +-18°C tolerancies are allowed.

The internal temp sensor seems to be a perfect match for this task, however I cannot find any data (graphs, tables) saying how much would the ADC read at Vdd = 5V and 115°C.

I am out of flash and cannot apply the calibration procedure described in AN3031. And I don't really need it, because I don't want to measure the whole temp. range, but just need a single point (ADC reg. content), corresponding to the 115°C, if the ADC reading goes below that threshold, I'd just shut down.

Unfortunately, the charts in AN3031 are  for Vdd up to 3.6V and a simple extrapolation won't work (already checked this, I guess the proper ADC value is somewhere about 200 for S9S08QD2 and ~100°C, which by no means matches the data from AN3031).

And I don't really want to spend my time in the thermal chamber, heating the µC up to 115° and trying to get the corresponding ADC reading  thru the BDM interface. 

Does anybody have an idea, where can I get those data?

Many thanks!

[bigmac · Accepted solution]

Hello,

Using the data given in paragraph 8.1.1.5 of the data sheet, the nominal voltage at 25 deg C is 1.3894 volts, with a tolerance of +/- 18 deg at this voltage level.  This corresponds with two diodes connected in series.  With a slope of 0.0033 volt per deg C (or 300 deg C per volt) the voltage should reduce by 0.25 volts for 100 deg C, and by 0.3 volts for 115 deg C.

For a regulated supply voltage of 5.0 volt, and 10-bit ADC resolution, the ADC reading should be nominally 233 for 100 deg C, and 224 for 115 deg C.  This was calculated using the formula -

ADC reading = ((25-temp)/300 + 1.3894)/5 * 1024

Regards,

Mac 

[fat_wombat]

Hello Mac,

as always, your advice is highly appreciated :smileyhappy:

Thank you very much for your explanation, now I know the logic behind the whole thing!

I was confused by the datasheet, saying the slope is 0.0033 but then, one row below, just 0.0017; I guess there's an error in that DS.

But now I know for sure (because I've tested it), that you're right with the 0.0033 slope, they obviously changed to 2 pn junctions in series.

Thank you once again, it is working now fine!