Stuck Z Gyro - Can I fix it? - Discussions

Hi,

I have an APM with a Foxtrot IMU and my Z gyro has stopped working for the second time. The first time I sent it back to Jordi and he returned it saying all was good - nothing wrong with it. Well, that was a bit embarrasing - I though I was going mad, but now it has stuck again!

When I run a simple test program that outputs all the ADC channels from the ADS7844 ADC chip, all of the signals are working, except channel 0 - the Z gyro channel, which stays at 1762-3 bits.

I haven't (yet) attacked it with a multimeter, mainly because I assembled my IMU on the bottom (as an oilpan) so getting to it whilst it is powered up is a problem.

Can I fix this? Is it possible to reflow the solder joints on the gyro chip? Is 1762 ADC bits a clue to where the fault lies? (i.e. open circuit somewhere, or stuffed ISZ500?) Can I replace the gyro? Anyone done this?

It's at least 6 months old and has never flown yet, apart from on a FedEx jet - bench plaything only, so far.

Thanks in advance for help.

Andrew

Andrew, You can easily power the shield by its' self. There are +5V & GND just down from the USB plug. Every thing is labeled on the back side of the APmega board. You can't see it because your AP is inside out. :0 Just kidding. The oilpan is an analog device mostly and works just fine even if there's no one to listen. Download the eagal files and reader if you haven't all ready. The Z voltage goes into pin 1 of the AD converter chip, which also connects to a nice solder pad that is for the RC filter enabling. I would solder a little stub to connect a scope lead or analog meter. I can only twist my wrist about 180 deg at a time and that doesn't give me enough time to read the flickering digits before the motion stops. Your eye can see the analog display in real time. The reading should be 0 - 3.3 VDC at +/- 500 deg/sec. So no motion would be ~ 1.65 V +/-. The converter is 12 bit so the max value would be 4096 and you getting 1762 ? Zero would be 2048 in the digital world. Good luck,

The ADC chip is right in the middle of the board and the Z( #1) pin is the top left.

Replies

Andrew Rabbitt June 8, 2011 at 3:42am

Hmmm... 1.426V on all points between the ISZ500 and the ADC when stationary. 1.426V / 3.314V * 4095 = 1762 ADC bits

I think the ADC is working well, but why isn't the ISZ500? 3.314V is being supplied to the closest point I can measure to it. Is it stuffed? Can it be replaced? Should I buy a chinese APM copy instead?
Greg Fletcher June 7, 2011 at 8:01pm

Andrew, You can easily power the shield by its' self. There are +5V & GND just down from the USB plug. Every thing is labeled on the back side of the APmega board. You can't see it because your AP is inside out. :0 Just kidding. The oilpan is an analog device mostly and works just fine even if there's no one to listen. Download the eagal files and reader if you haven't all ready. The Z voltage goes into pin 1 of the AD converter chip, which also connects to a nice solder pad that is for the RC filter enabling. I would solder a little stub to connect a scope lead or analog meter. I can only twist my wrist about 180 deg at a time and that doesn't give me enough time to read the flickering digits before the motion stops. Your eye can see the analog display in real time. The reading should be 0 - 3.3 VDC at +/- 500 deg/sec. So no motion would be ~ 1.65 V +/-. The converter is 12 bit so the max value would be 4096 and you getting 1762 ? Zero would be 2048 in the digital world. Good luck,

The ADC chip is right in the middle of the board and the Z( #1) pin is the top left.

This reply was deleted.