I have 2xSXT, one powered by 15v and one 24v and they shows 10.4v and 17.2v. Also temp is wrong 25 and 28. RouterOS 5.1.
ive the same problem
is the temperature and voltage error very big for you too, meno?
I know that the SXT sensor is not very accurate, and should be used only to monitor changes, not measure accurate reading. We have improved this in the next hardware update.
But if the difference is very big, there could be a miscalibration issue when the unit was manufactured.
Nice, broken units again.
yes normis Ihad used ubnt throghcable level 1 with 24v power supply and the SXT reads these values.
its not a big deal ..
I bought some SXTs this month (July) and still have same the problem (voltage ~16V instead of 24V).
Meno, how long is the cable? The longer the cable, the bigger the voltage loss. This is normal.
We’ve got the very same problem with our SXTs.
Even on the desk with a short and directly connected Ethernet-Cable
and 24V power supply I see only about 16V in System-Health.
In scenarios with 150m of PoE ethernet cabling (1x repeated inbetween with a PoE-Modded RB) we only see up to 4 Volts lost. But the SXT constantly shows much too less voltage.
why it could be different, see mine.
is it normal?
just a note - voltage measurements are done behind some protections chains and that could lower displayed voltage, as actual voltage is lower in place where it is measured compared to voltage on jack of PSU.
Yeah, I could understand this if it would have been about 0.7 to 1.2 volts droppped for masurement.
But eigth volts xdropped for measurement?
or 12v… 
this must be a calibration issue, please send the serial number of this device (8 or 12v difference) to support, we will check production logs.
Mine shows a temperature of 44C while it is in a romm that only hast 19C … so what temperature does the sensor show? CPU?
yes, of course the CPU, not room temperature.
Well this bug scared the crap out of me this morning. I have 2x SXT-5hND units on the same new solar powered site. One is reading 10.3V, and another is reading 9.8V. I expect that both of them are receiving in the order of 13.4V. I’m going to have to put a remote serial port in to monitor the charge controller now that I know these boards won’t give me an accurate power reading.
if you can couple precise voltage reading with reading from the SXT you can use the delta pretty reliably if surrounding conditions do not change much.
Also, there is around 2V voltage drop on PoE protection chains + drop on cable.
I’m guessing this doesn’t just apply to the SXT’s? We have a solar site where the solar charger accurately shows the battery voltage about 4V “higher” than an RB450G and 2V “lower” than an RB411AH. Very strange considering the RB411AH is PoE across 50m of cat5 as I would have thought the voltage would drop not increase 
Exact same problem here… Anyone resolved the issue? Thx
for older RouterBOARD product this value could be off by some value, if voltage is changing and other parameters does not, then error delta is the same. Newer boards have different scheme and that has higher precision.