Has anyone noticed their IRIS losing altitude while moving or turning in alt hold mode? If I leave it to hover in place then it does a decent job, not moving more than a couple feet up for down depending on wind. However as soon as I start/stop moving I will see altitude drops as much as 20 feet sometimes. Anyone have idea what my issue might be?
You need to be a member of diydrones to add comments!
Replies
Check out this thread in the Ardupilot forum. Looks like there is a solution worth testing. I will as soon as I get my replacement Ublox back.
I just don't see how they could have not known the GPS was pressing against the the PIxhawk unless whomever was making the covers changed the mold slightly causing some batches of covers to have different heights/profiles.
mp
I completely agree with your comment in the other thread around the Iris being a pure disappointment so far with regards to being an RTF package. Unfortunately I can't write that there, as this kind of honesty will lead to account termination.
I also bought the IRIS hoping that I could fly with it and that I don't have to build a Quad from scratch. While I also learned a lot along the way for which I'm partially grateful that initial hope completely fell short. The IRIS has just way to many hardware bugs (design mishaps) that I'll trust it for any serious flying.
While I did get in a few decent flights, the unpredictable issues spoil these few moments. And looking at them close I always discovered some design issues, ranging from batteries with an unusual discharge curve, incorrectly configured failsafes, insufficient motor mount screws, insufficiently mounted electric components and now the flight controller vibration issue.
I don't feel comfortable taking something flying for real mission (other than hovering at low height over a wide open field) if I can't trust the device.
I'm almost at the point where I want to go methodical through all the components that could fail, investigate potential issues and fix it. But then isn't that what 3DR should have done in the first place?
And it would bring me even further away from "just flying" and closer to spending more time on fixing.
I'm getting really frustrated with the IRIS. One day its bad hdop another day other weird stuff.
In the attached log you can clearly see that every time I goto auto to execute the mission, the altitude starts dropping below the desired altitude. At the end of this flight the copter just went buggy. I could hear what sound like motors cutting in and out. I was trying different modes and was really having trouble controlling it. When it really freaks out like this I started getting messages from the GCS that there is no data.
I had another mission right before this one that wasn't that bad. Just when you think the Iris is starting to act right it wigs out.
I'm not great at the logs yet but I think my compass is acting up as well. I saw if flying sideways at waypoints today.
Log is attached
Thanx,
mp
73.BIN
I think I'm on to something that would explain random, erratic behavior of the IRIS that I am experiencing. While putting the flash card back in the Pixhawk I started looking close at the connectors and found what you will see in the attached photo. The I2C connector on the Pixhawk is in really bad shape with exposed metal on 1 or more wires. I have never disconnected any of these. I noticed it when I first received shipment but never did put on strong enough glasses to see it clearly.
If the ublox GPS cable goes to the GPS input on the Pixhawk then the I2C cable must be for the compass? I'm not sure, what do you guys think? The connector is definitely at a low integrity level.
mp
pixhawki2cconn.jpg
That indeed doesn't look good.
I noticed that the uBlox GPS module attached to the top of the hull actually touches the cables on the top of the Pixhawk. In particular one of the uBlox plugs touches the I2C connector on the Pixhawk, exactly where you are experiencing issues. As the Pixhawk flight controller vibrates during flights, the uBlox will rub against the connector and cables and most likely damage them.
To me that looks like a design issue by 3DR.
I placed a piece of paper on top of the uBlox, so that the connector can't rub against the Pixhawk. I might replace that with some fabric or a piece of foam.
Also, if the uBox GPS touches the Pixhawk, vibrations from the frame will travel via the frame and uBlox onto the Pixhawk.
So would shorts or disconnects on the IC2 buss cause these types of issues:
1) random poor GPS performance like low sat count and low hdop
2) random loss of GPS signal when 12 sats are being seen. I've seen in the logs where the GPS sat count would drop to zero then go good over a few second period. Also get the GPS alarm from Pixhawk and verbal from MP.
3) Randomly losing altitude and heading the wrong way in a auto mission.
4) Randomly start toilet bowling in loiter like the compass is wigging out.
I can't seem to get more than 2 reliable flights in a row. Just when it all looks going things start going bad again with no physical or config changes. I'm going to see if 3DR will send me another I2C cable. Even if it's not the cause, it needs to be replaced.
Doesn anyone know what is on the I2C buss besides the compass?
You can follow the I2C cable to that little distribution board to the left side of the Pixhawk controller. If you follow the cables out of that board you'll see that they lead to:
- uBlox GPS module
- ESC power monitor
- LED light in back of IRIS
Some of the issues that you are reporting could be affected by a poor GPS / compass module or a poor connection between this module and the Pixhawk. Swapping the cable might be a good first step, but it might not solve it completely.
In my opinion the position of the uBlox module so close to the Pixhawk is a design issue (http://ardupilot.com/forum/viewtopic.php?f=48&t=7509&sid=b3...).
You might want to ask 3DR what they recommend you to do for fixing this long term. Just replacing the cable is not enough. The new cable will be rubbed blank just the same way as your current one without any additional fix.
Thanx for the input. They sent me a Fedex label to ship my GPS back and they will ship me a new one. I gotta make sure they send me a new I2C cable as well. Hopefully this will fix the issues.
mp
Good to hear that you're getting a replacement. Don't forget to put some foam between the new uBlox and the Pixhawk. Otherwise you'll soon run into the same issue again.
"Connecting " PIXHAWK with the UBLOX (even by foam), will affect vibration damping. Check your logs before and after.. I think its better to move the PIXHAWK backwards, even if this requires a different location of the 3DR radio.
Some inspiration: http://www.rcgroups.com/forums/showpost.php?p=26970863&postcoun...
@Mark: compassmot is not required. But I used plastic screws instead of the magnetic ones to fix the GPS.
http://www.rcgroups.com/forums/showpost.php?p=27425378&postcoun...
-
1
-
2
-
3
of 3 Next