Hi,
I used thermal.c to read temperature. But the environment temperature and thermal sensor have range of 20-30 degree. Is it correct?
What is the temperature I can reference. And what is thermal sensor error range?
Thanks!
Solved! Go to Solution.
Hi, Jon
If you are using 95C max part, and your board can easily up to > 80C under normal usecase, then I think you should try to optimize power consumption, either from software or physical method to cool down your platform.
Yes, if the accuracy is within +/- 10C, then you should set the critical point threshold to 85C, and the hot point threshold should lower than critical, suggesting < 80C.
Can you tell me whick version of BSP you are using, in the latest BSP release 4.0.0, we used universal equation for thermal sensor instead of calibration data in Fuse, the error is supposed to be < 10 C.
Has Freescale published any documentation stating the accuracy of the internal iMX6 temperature sensor, after applying the calibration fuse data? We need to make sure we are not operating at an unsafe or unstable temperature.
The temperature measurement accuracy is not specified in the data sheet. The latest product, software, and fusing should be within +/-10 degreesC of the actual.
I'm using the consumer grade (95C max) part, so does Freescale then recommend we set the thermal reset temperature down to 85C to guarantee stability? Ie. don't ever let it get hotter than a reported temp of 85C? I'd like to get the accuracy better because some of our parts are reporting they're in the 80C range.
Hi, Jon
If you are using 95C max part, and your board can easily up to > 80C under normal usecase, then I think you should try to optimize power consumption, either from software or physical method to cool down your platform.
Yes, if the accuracy is within +/- 10C, then you should set the critical point threshold to 85C, and the hot point threshold should lower than critical, suggesting < 80C.
From dgd:
Please refer to AN4579 (rev.0, 11/2010) "i.MX 6 Series Thermal Management Guidelines", which shows the calibration procedure.
http://cache.freescale.com/files/32bit/doc/app_note/AN4579.pdf