Does the PX4 need a case or cover for the barometer, like the APM does? Been trying to get my Y6 to fly in Altitude Hold and all it does is bounce up and down. Vibration is not the issue. The APM needs a cover, so does the PX4 need one too? If it does, where is it? Thanks
It just needs a piece of foam over it sir. No need for a casing.
By foam, i mean those really spongy type.
Hi,
I am currently having trouble with PX4 in Alt Hold modes and it strongly seems to be related to prop-wash back-pressure resulting in auto-throttle correction bouncy pulses. 
Switch into Alt Hold mode from stabilize and copter immediately goes from nice steady hum to loud - quiet - loud oscillation while moving up and down over about a 3 second period in a 6"+ range.
Nothing dramatic, but it appears that the props pressurize the baro which thinks it is going down and so it increases the throttle (which both increases prop back pressure and the copter also now senses it is going up, so the baro now responds by decreasing the throttle. 
You can see and hear this while the copter is flying in Alt Hold: Whrrr-UP – Quiet-Down – Whrrr-UP and so on. Switch to stabilize and nice constant easy hold prop sound. 
I have a piece of medium foam heavily compressed sandwiched over the baro between the PX4FMU and PX4IO board and this has only a small improvement.
My Loiter is not working well, but with constant throttle oscillation for the Alt Hold part it isn’t worth trouble shooting that until the Alt Hold part is working better (and is probably related).
Stabilize is working excellent, would rate a 9 out of 10.
Bottom line is un-encased PX4 can definitely be a problem and foam may not be an adequate solution.
Probably depends on characteristics of copter and whether prop down wash can affect local air pressure where PX4 is.
I also have an APM that does not have this problem at all: 
On the APM the baro jitter on the log is about 1/10 to 1/5 whole division on the PX4 it is about 1 so 5 to 10 times as bad.
My Accels are all in the range of < 1/10 G in fact < 1/20 G thanks to Kyosho Zeal Gel mounting.
I have tried a partial top cover and this did improve the situation on my Hoverthings Flip, but still not adequate.
I really think that the PX4 needs a small enclosure to let these pressure pulses even out.
Best Regards,
Gary
[quote=“gary”]Hi,
I am currently having trouble with PX4 in Alt Hold modes and it strongly seems to be related to prop-wash back-pressure resulting in auto-throttle correction bouncy pulses.
Switch into Alt Hold mode from stabilize and copter immediately goes from nice steady hum to loud - quiet - loud oscillation while moving up and down over about a 3 second period in a 6"+ range.
Nothing dramatic, but it appears that the props pressurize the baro which thinks it is going down and so it increases the throttle (which both increases prop back pressure and the copter also now senses it is going up, so the baro now responds by decreasing the throttle.
You can see and hear this while the copter is flying in Alt Hold: Whrrr-UP – Quiet-Down – Whrrr-UP and so on. Switch to stabilize and nice constant easy hold prop sound.
I have a piece of medium foam heavily compressed sandwiched over the baro between the PX4FMU and PX4IO board and this has only a small improvement.
My Loiter is not working well, but with constant throttle oscillation for the Alt Hold part it isn’t worth trouble shooting that until the Alt Hold part is working better (and is probably related).
Stabilize is working excellent, would rate a 9 out of 10.
Bottom line is un-encased PX4 can definitely be a problem and foam may not be an adequate solution.
Probably depends on characteristics of copter and whether prop down wash can affect local air pressure where PX4 is.
I also have an APM that does not have this problem at all:
On the APM the baro jitter on the log is about 1/10 to 1/5 whole division on the PX4 it is about 1 so 5 to 10 times as bad.
My Accels are all in the range of < 1/10 G in fact < 1/20 G thanks to Kyosho Zeal Gel mounting.
I have tried a partial top cover and this did improve the situation on my Hoverthings Flip, but still not adequate.
I really think that the PX4 needs a small enclosure to let these pressure pulses even out.
Best Regards,
Gary[/quote]
wow looks like a case may be needed. I am assuming that you have tuned the PID settings for the alt_hold.
Its interesting that you are having problems with the loiter mode. I, too, am having problems with the loiter mode. Once i flick it into loiter mode, the quad just drifts to the left really quick but the flight happened in really windy weather. will try once more in calm weather.
Ive since gone a bit further in analysis.
On my Flamewheel F330 also with a PX4 the problem still exists but is much less, and on my Flamewheel F450 with an APM the jitter is barely evident at all.
But it is all caused by prop tip proximity to central frame plate corners.
On the Flip, the 10" prop tip ends are actually over the very edge of the copters central frame plate, permitting prop down wash to directly impinge on the frame plate itself resulting in considerable pressure pulses as the tips pass by.
On the F330 there is about an inch of clearance of the 8" props from the frame plate, and the induced pressure pulses are much less.
And on the F450 there are about 1 1/2 inches of clearance from my 11" props and the pressure pulses are negligible.
My conclusion is that this is really only a problem for copters with little central frame plate clearance from the prop tips.
It can be alleviated by either using smaller diameter props or by building an enclosure for the PX4 and venting it on or near the top.
It should not be as pronounced an effect for the encased APM or Pixhawk, because the case acts as a pressure equalizing box.
But if you have “bouncy copter syndrome” look at your prop tip clearance from your copters central frame plate first.
Best Regards,
Gary
Thanks for the insight. Where is the barometer located? Is it the silver square just next to the large chip in the center of the board?
Thanks,
Richard
No, that is the clock oscilator, the baro is the silver rectangle with 2 small holes, near the JP2 label.
Update on bouncy copter syndrome.
I was wrong 
I had a PX4FMU with a genuinely bad baro.
It actually displays about plus and minus 1 meter of jitter when it is sitting absolutely still with a USB only hookup on my table - no motors or props spinning.
Put in a new PX4FMU and problem went away, alt hold now rock solid, baro jitter very small and am using a piece of foam over baro but no enclosure or housing.
A nice thing about the PX4 is that the baro is at the edge of the PX4FMU board with the PX4IO board directly opposite and you can just tuck a piece of foam so that it is squished between the boards at the edge.
The bad thing is that it is half under the PX4IO’s little 5 volt linear power supply so you have to be careful tucking the foam in.
Apparently prop-wash back pressure was not a significant factor.
Best Regards,
Gary
I have the exact problem…
But mine occurs at random times… I designed and 3D printed a case, and have the pink sponge my old APM use to have, but still… my altitude drops to 0 at random times…
Did you ask for warranty to 3DR??? or did you buy another??
Cheers!!
Gary,
Have you tried without the foam between the boards? I’m just wondering if the foam makes any real difference.