Any Melexis gyro problems?

[maxvtol]

After many successful flights of my quad tiltrotor, my yaw gyro quit working, a Melexis MLX90609 300deg/sec from Sparkfun (http://www.sparkfun.com/commerce/product_info.php?products_id=8372). Thank goodness it quit working while my machine was on the ground.

Has anyone had problems with this gyro before I order another one?

I've only had a minor crash before the problem started, and I did hook up one gyro wrong when I first made my mixer, but I can't remember which one and there was no smoke, it just didn't work. That was well over a month ago with many flights in between. I'm also using a voltage divider to get the gyro voltage from a max of 4.5v to <3v for connection to my ADC, using a 10k and 15k resistors.

A video of my crash below (I had hooked up the FMA Co Pilot sensor and was testing it when the BEC failed), maybe someone could comment on the severity of their crashes and any problems after. This one only resulted in one prop, 4 servo splines, motor shaft and a 3s 4250mAh battery, about $150 after it was all said and done (not counting this gyro).

Thanks for your help!
Crash Video (http://www.youtube.com/watch?v=GLMoKYmFsLY)

[Jack Crossfire]

The main problems with gyros:

Vmax on the A/D converter is not stable.
A/D converter doesn't boot.
Vibration breaks solder joints between the gyro & processor.
Vibration breaks umbilical wires.
Exposure to inverted/excessive voltages always destroys them regardless of "many flights in between".
Never had a problem with sparkfun's reflowing.

[maxvtol]

Thanks Jack,

Didn't see any visible signs of problems on the chip using magnifying glass, probably internal problems. I'll chalk this one up to experience, another one is on the way, hope to be flying by Thanksgiving.

[wadiprawita]

Dear maxvtol,

I also have same problem with mlx90609. I build 2 IMU, using mlx90609 (3 axis) and MMA7260q. One of them is installed in airplane for flight test, no problem until now. One of them is used to refine the fusion algorithm on my desk, powered from laboratory DC power supply (and regulated on board using lm2940). Strangely the one of gyros (yaw axis) from the IMU on the desk suddenly stops functioning. During zero rate the output is 3.8v (should be 2.5 v). Is this also your symptoms ?

For your information I bought 9 mlx90609 gyros from sparkfun, and I had this experience with 2 break out board (I'm pretty sure the problem is not with sparkfun, but maybe with my soldering, to hot maybe ? or other issue : regulator, etc).

Regards

-doni-

[maxvtol]

wadiprawita,
I don't recall the voltage, but it did go to one extreme, I believe it was .5v. It did it the first time about 5 minutes after I started one of my flights. The next time it did it within a couple of minutes. And of course when I took it off to bench test it, it took over an hour. I have a new one now, but haven't installed it yet.

At first I thought it may have been one of the gyros I hooked up wrong when I built the board, but after your experience, I'm not sure. Please keep me posted on any problems, and I'll do the same. My other 2 gyros seem to be working fine.

Mine was also from Sparkfun, but just the single gyro board.

[wadiprawita]

maxvtol,

I still have spare gyros. I will try to examine the board within the next 3 days, I will replace the gyro if I have to. I will inform you if I find something new.

Regards

-doni-

[wadiprawita]

Well,

after examining the IMU, I'm a little worry, yesterday the system with works fine for about 5 minutes. I turned off the system and restart, the symptoms showed again. This morning I run the system for more than 1 hour and the gyro still works fine.

Regards

-doni-

[maxvtol]

mmm.. sounds like I better do an extended bench test on mine when I replace it.

Thanks, keep me posted on your progress.

[nik314]

Dear maxvtol,
Pay attention, the OUTAR of the Melexis gyro has a limited load - it should be less than 200K (15+10=25K in your case). Unfortunately this value is not mentioned in their very short public datasheet on the website. But they've promised to put an updated version of the datasheet soon.
Nevertheless, it's unlikely 25K load could cause the fault you had.
Regards,
Nik

[nik314]

Sorry, I meant "more than 200K", of course...
Nik

[maxvtol]

Thanks for the update, but if I'm using lower ohm resistors, wouldn't that cause more load?

[nik314]

I think it could add some nonlinearity but as mentioned it's unlikely it could cause a crash...
Regards, Nik