I have tried everything I can think of so far but cannot get this IRIS to maintain altitude in forward flight. This includes auto, alt hold and loiter mode. I have attached a log. You can see the auto attempts at about 6, 9 and 12 minutes. As soon as auto mode is engaged and the forward flight starts the altitude starts dropping. If you let it go, it will fly itself into the ground. The throttle out gets really noisy too. This PH is my first auto flight computer so I’m still in the learning curve on the logs.
You will probably notice some other compass related issues in the log as well. My Ublox is currently being shipped back to 3DR for replacement but I think the fact that the Ublox is pressing down on the Pixhawk I2C connector when the cover is on my have something to do with it (see attached pic). I’m not really asking for help with that, just the alt hold control.
After over 70 flights my Iris is really only good for hovering. I am willing to accept the learning curve but this is a bit frustrating. All of my issues with this copter are random except for the alt control. It’s always there. Any feedback is appreciated.
as I’m facing similar issues, I did a lot of tests around this.
Here is in a nutshell how far I’ve gotten:
The altitude loss seems to be related to the barometer failing which is indicated by the BarAlt diverging dramatically from the GPSAlt.
My initial though was that wind (either from gusts or because of IRIS moving) could be causing this. So I placed IRIS on the ground and used a hairdryer (with cold air setting!!!) and blew air onto IRIS and into IRIS (vents and back) to see the effect on the barometer. Even if you funnel the air directly into the back and create an overpressure inside the hull, the perceived altitude by IRIS drops only by less than 2 meters and not as much as the 10-15 meter that I have been observing.
Next I noticed that the barometer failing was related to high vibrations along the z-axis, which in return was related to oscilation on the Channel Out (motors). See this post here for more details. What was causing the z-axis vibrations was the big mystery at this point.
Next I looked into the problem of the uBlox GPS touching the Pixhawk. The same problem that you are also facing. The results of these test can be found in this post. After isolating the Pixhawk from the GPS by raising the hull and thereby the GPS, the sporadic z-axis vibrations go away.
At this point it looks to me like the Pixhawk touching the GPS creates a feedback loop for vibrations, which leads to oscillations. With that we need to wait until 3DR comes up with a fix for this design mishap, unless you want to fix it yourself and see if it makes any difference.
Thanx a million for the input. Yesterday I bought some foam sheets of varying thickness to make a spacer gasket that lifts the cover enough to keep them from touching while also absorbing some vibration. Of course I also picked up longer bolts to accommodate the spacer height. It’s odd that the 3DR folks that read the log I posted on this thread didn’t see any issues except for “magnetic compass interference”. They said it controlled altitude fine which is hilarious.
Thats when I gave up on them and posted here. I am anxious to test the gasket but I sent my Ublox back to 3dr this week for a replacement. I was having random issues with poor GPS performance like DOP > 3 all day. Some days it never gets below 3 and when I would have good GPS signal it could momentarily drop to zero sats, cause a GPS failure and then go back to normal. I suspect this could have also been caused by the vibration loops.
I fully agree with your concept of the feedback of the vibration. This could randomly cause numerous issues like I am having. The Iris has been a pure disappointment so far if perceived as an RTF package. I bought the Iris to avoid the struggles of a raw build and tune. I’m learning a lot but disappointed to say the least. I have about 80 flights on it and only maybe 5 have been bug free with the exception of just hovering for battery testing.
Thanx again, might be a couple of weeks before I get my Ublox replacement.
My idea about the foam gasket sucked. I failed miserably in making it. There just isn’t enough width of usable surface where the cover contacts the surface.
Looks like I will be using foam spacers in the ares where there is enough room. This doesn’t provide a long term solution but I could at least test. After about 5 lights I can confirm what others already have - “the contacting of the Pixhawk to the Ublox causes various issues that render the IRIS to a hover toy and even that can go bad”.
It’s been proven by at least one user to eliminate the excessive accZ vibrations(which seem to be causing bad alt hold performance) not to mention the issues that can be induced into the compass and GPS with the vibration loops that are going on.
Does anyone know where to get the gel pads for mounting the Pixhawk? If they are thinner than the foam that the Pixhawk is mounted on it would help a lot by lowering the installed height of the Pixhawk.
Contact help@3drobotics.com to get a set of thinner mounting foam. It seems that the foam supplier supplied a batch of foam pads which were too thick to 3DR, hence the reduced distance between PH and GPS.
Thanx for the feedback. I have contacted them and sent the pic of the damaged I2C connector caused by the Ublox pressing/rubbing against the PH. They offered no solutions. They are however replacing my GPS due to random, very poor HDOP and SAT numbers. Even when the GPS signal is good it drops out completey (zero satellites) randomly. I’m installing a mast and getting the GPS away from the PH.
You have to specifically refer to the foam. It was just a few days ago discovered that the foam manufacturer sent a bad batch - too thick foam - to 3DR which led to the reduced distance between the Pixhawk and the GPS. Craig wrote about this in another thread and said that people who have this issue should contact help@3drobotics.com and specifically request a new set of (thinner) foam, referring to the forum.