Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Problems using accelerometers/gyroscope

Status
Not open for further replies.

Hairybear

New Member
Hey all,

First post on here so firstly hello to all, already found the chat link very helpful.

I'm building a circuit 'tilt sensor' which displays 20 LEDs driven by a pair of LM3914s. I'm using a freescale MMA7361 stamp which has a selectable +/- 1.5 and 6 g output.

Here lies the problem.

I'm only interested in measuring +/- 20 degrees from the horizontal. Reading the data sheet, it shows there is a 800mV change, per 'g' and 1g is equivelent to 90 degree's. This would mean 800mV/90deg would give me my mV change per degree moved, ie 8.8mV per degree moved/displaced.
When I put the output through my oscilloscope, the trace is unstable. The trace p-p is about 30mV therefore as 1 degree should be circa 9mV, I'm unable to display any degree of acuracy beyond 3-4 degrees.

Any help, advice welcome...maybe you've had the same problem...maybe its noise or are these 7361 chips that bad?? The Chip has conditioning so presume its not that...

Over to you!
 
Last edited:
You are set to 1.5G???
I do not know the "stamp". The MMA7361 shows a capacitor on each output. You might add a RC low pass filter and try to reduce the noise. (average over time) The part has a bandwidth of 300 to 400hz. If you average and only have a bandwidth of only 40hz it should help the noise.
 
Have you got capacitors on the outputs?

From the data sheet:-

"4. Use a 3.3nF capacitor on the outputs of the accelerometer to minimize clock noise (from the switched capacitor filter circuit)."

You've also made an error in your assumptions, although not one that would make much difference in your application. The output is 800 mV per g, and from flat to vertical is a change of 0 - 1g. However, the output will not change linearly, the formula is:-

Output in volts = 0.8 sin (A) where A is the angle from the horizontal. So the output at 20 degrees is 273 mV. The sensitivity near the horizontal is 14 mV / degree. There is little change over +/- 20 degrees but it drops to about 10 mV / degree at a tilt of 45 degrees, 7 mV / degree at a tilt of 60 degrees, and 0 mV per degree when vertical.
 
Thanks for the replies. I think i've sorted it, or at least identied the problem. I'm using the 1.5g ouput however when I move the sensor, the trace on the scope a mixture of static and dynamic errors. Therefore, when i hold the sensor in my hand the vibrations from my hand 'shaking' were causing dynamic changes what looked like a 'noisy' signal, infact its just the accelerometer being too sensitive. I held the sensor in a clamp and moved it, thus damping out the vibrations (dynamic motion from my shaky hands) and the output was smooth, linear and as per the spec (within reason). Problem is, this doesnt help me, as I'm only interested in displaying an angle, ie static output, therefore if the sensor is on a vibrating vehicle for example it will be displaying the rate of acceleration due to vibrations, rather than 'at rest' static angle.

Hmm....back to drawing board.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top