Yes. Just the other night I did see the highest temp reach 25.0 and even 25.1 and the 6th temp bar never came on... but in the past I'd seen it click on at 25.0 and off again at 24.9 so I believe the hysteresis (or lack thereof) is correct. I've been displaying temp to 0.1 degrees C now for about a week and it's certainly more entertaining than just using the car's 1.0 precision. There is a +/-0.1 error if you repeatedly sample but you can certainly discern temp changes of 0.2 degrees C or more. With the increased screen real estate on CANary you could certainly do a lot with this. Due to my limited screen space I'm just displaying Tmax (which is almost always T1 or T2, the front stack) as well as the delta from Tmax to Tmin just to get a sense of the spread..garygid wrote:GregH,
So, it looks like the raw temp data will indicate
a precision of very near a tenth of a degree C,
even if we do not know much yet about the accuracy.
Akso, it seems that tell Pack temperature Bar display
operates on the highest temperature, and there is
essentially no hysteresis.
So, using the raw values, we might try to display the
extended-precision temperatures, like "C20. 3".
In the CANary, scrolling plots of the 4 high-precision
temperatures on one graph would be very interesting.
Note I've been editing the post a few back to update the observed transition points for temperature conversion. I look forward to seeing data from folks in warmer or cooler climates so I can tweak my software for optimal precision... On my group 4 page I show the raw data, the car's temp (Celsius 1.0 precision), and my calculated temp (Celsius 0.1 precision) such as:
473 24 24.4
476 34 24.1
493 22 22.4
491 22 22.6
(my current battery temps)
And my main page shows 24.4-2.0