Sugestions on how to make a precision inclinometer

Hi. I need to build a precision inclinometer to mount on a truck to check if it is within certain inclinations values when it is parked. I was thinking on using a 3 axis digital accelerometer as MMA8451q using low-noise with oversample mode and a digital low pass filter. The accel is 14 bits precision. I need a 0.1 degree precision from -45 to +45 degrees. Is that a good idea? Any good sugestions?

thank you!

Sink0 said:

Hi. I need to build a precision inclinometer to mount on a truck to check if it is within certain inclinations values when it is parked.

Click to expand...

That depends on the purpose of the exercise ..

If you need only one inclinometer to perform on task – go and buy an off-the-shelf unit and don’t bother with R&D ..

Example: **broken link removed**

If you think of many units in many applications then maybe developing an inclinometer by yourself would make sense ..

:wink:
IanP

This may sound trivial but in practice it's not.
0.1 degree precision with a range of +/-45 degrees sounds easy, I recommend you do some calculations.

At 0.0 degrees you have a force of cos(0.0) = 1g, at 0.1 degrees you have a force of cos(0.1) = 0.999998477g
At 45.0 degrees you have a force of cos(45.0) = 0.707106781g, at 44.9 degrees you have a force of cos(44.9) = 0.708339838g
At 90.0 degrees you have a force of cos(90.0) = 0g, at 89.9 degrees you have a force of cos(89.9) = 0.001745328g

In the most sensitive mode, +/-2g you get 4096 counts per g.

For 0.1 degrees you will have the following number of counts at the different orientations.
At 0 degrees: (1 - 0.999998477) * 4096 = 0.006238208
At 45 degrees: (0.708339838 - 0.707106781) * 4096 = 5.050601472
At 90 degrees: (0.001745328 - 0) * 4096 = 7.148863488

This looks good, for a perfect accelerometer.
The sad truth is that you have (for MMA8451q) a typical offset of +/-30mg or +/-122.88 counts. No problem to handle with calibration if it's constant. It's not.
In my experience with accelerometers from ADI you have a lot of offset drift, both over time and temperature.

Maybe it's no big deal. Maybe you don't really need 0.1 degree precision.

We decided it was not worth the trouble and decided to use this inclinometer: **broken link removed**

Good luck!

@IanP

Yes that is for a product, with not big production, but still a product, so price got some impact here.

@skogsjanne

Thats one of the reasons i am asking here. Offset is a big problem. How much did you pay for that module. It seems to have two hidden accelerometer and a PIC MCU. Same i wil be using (but with a STM32F100 device).

The 0.1 degree requirement is due to competidors specifications, but is a meanless spec. The inclinometer is mounted on a truck and calibrated in loco. So for sure the truck is not going to be placed on a perfectly plane, so an offset will exist anyway. The inclinometer is for safety purposes, and to not allow the truck to park under certains inclinations scenarios.

Any sugestion on how to deal with the offset problem? In special the temperature one? Any other device that seems to be more stable?

Actually i am considering that MMA8451q is 13 bits and not 14, even using oversampling and low noise mode, due to noise.

Any sugestion is apreciated.

Have you checked this type tilt-sensors , for using in truck .
Dual-axis, operating temperature -40...+85 Celsius,
I made pre-design for my client with these few years ago

An industry leader in design & development of tilt measurement systems, our tilt sensors are the hallmark of accuracy,
repeatability, and reliability. Products include TrueTILT™ metal & glass, single and dual axis, electrolytic tilt sensors,
inclinometers, inclination sensors signal conditioning electronics, packaged tilt sensing assemblies, tilt switches,
and on/off switches.

**broken link removed**

**broken link removed**

they have mini signal conditioner boards too
**broken link removed**

I remember that prices for
0729-1751-99
Mini Signal Conditioner Assemby with 0717-4318-99 Dual Axis Sensor and SPI Interface

price was about $50 when ordered < 10 pcs three years ago.

**broken link removed**

Other manufacturers...

**broken link removed**

**broken link removed**

**broken link removed**

@Sink0 Sorry but I don't know the price. It was much more expensive than the ADI parts at that time but that's only expected. Send them an email and ask.