I am contemplating taking the APM 2.5 apart and painting the inside black to ensure the barometer is not going to sense light. Any reason I shouldn't do this?
If the barometer is sensitive to light why would the case be transparent? I don't think I care about seeing the status LED's.
Replies
It can also very well be that there is something wrong with the Barometric pressure sensor on the APM, that has been in production.
Seeing the video post here, seems like the behaviour is very weird (mark the word very). Though, my speculation that is there something wrong with the internal electronics of the sensor or soldered connections, may not be true. However,
It is a god damn Air Pressure Sensor not a Light Detecting sensor.
Replying to subscribe to this discussion.
My hex did this today in loiter mode sun went behind a cloud and it climbed by about 2m when the sun came out the hex came back down and it did this several times everything else was stable.
I flew my arducopter in direct sunlight and althold wasn't stable, this was the log after:
Here i made a quick test about baro sensitivity to light. So i guess it is very important to properly cover baro so there would be no differences in light intensity on sensor itself. While testing it was overcast day and as you can see even by casting a little shadow when covering baro by hand, it makes a 2 meter error.
Hmmmm, I had an odd experience the other day that I couldn't explain with my SSS running Arduplane 2.74b. It was the third flight of the day and the sun's position had changed throughout the day. When I opened the canopy exposing the APM 2.5 (enclosed in it's case) in order to plug in the battery, the sun was shining directly on the board. This is the first time that day that it has seen direct sunlight on it at startup and as soon as power was applied the ailerons went to a full right position while the APM was in Manual mode, and the elevator went to full up position. After about three seconds the throttle started slowly increasing and I immediately pulled the power. I tried powering the board two more times with the same result as before. I realized the sunlight was on the board. I shaded the board with my hand while I applied power the third time and voila everything worked fine! After the APM had initialized there didn't seem to be a problem.
Is this related to the same problem you are discussing here with the barometer? Can someone explain to me how it affects the APM initialization?
Regards,
Nathaniel ~KD2DEY
I will test the barometer out of the case in a few hours because I am going to spray paint the inside of the APM 2.5 case black.
Permalink Reply by Randy 15 minutes ago
Dave,
You're right. That actually looks like baro noise. You have the foam on the baro of course? The other possibility is that light is hitting the baro at just the right angle. We've discovered in the past month or so that strangely the baro can jump up to 17m if a bright light hits it at just the right angle. The solution is to put some black electrical tape (or something) over the case where the baro is (you can likely see where the baro is because it's covered in pink foam).
Reading the data sheet for the MS5611-01BA03 Barometric Pressure Sensor, I find no reference to light sensitivity. Is this an observed characteristic or information about piezo-resistive sensor barometer products in general (I can find no current information on internet except one reference that indicates most manufactures have eliminated this issue for sensors with semiconductor junctions. No indication the MS5611 even has pn or np junctions. I thought the foam was to reduce errors caused by turbulence (wind, props, flight speed) Just curious.
I used black magic marker on the inside and you can see the leds still it tints it furher enough to help block the light
-
1
-
2
of 2 Next