Need sensor idea

CrazyIgels · Joined: 26 Oct 2007 Posts: 79

Hello All!

I need sensor idea:
This sensor need for measurement of tilt angle of solid granite flat base used in high precision optical system making CNC.

Measurement range: 6 degrees total +3 up from zero level and - 3 down to zero level
Measurement axis: one axis (i have plan to use many independent sensors in all parts of CNS machine)

Measurement precision: one arc minute

Measurement angle variation speed: relatively slow, something like 1 deg. per sec

Can i use MEMS aceleration sensor for this? may be you can tell me sensor model / part name?

Or this task impossible to solve with MEMS?

As main sensor chip i think use something like Mega8 or Mega 8535 and BASOM Smile

Thanks all.
Sorry my bad english

PS
So, i sorry if this is wrong place for my message - move it to right place :)

(BASCOM-AVR version : 2075 , Latest : 2.0.7.6 )

AdrianJ · Joined: 16 Jan 2006 Posts: 2483 Location: Queensland

Sorry for the long delay, I only just noticed this.

MEMS acceleration sensors will suffer from temperature and aging drift down in this range.

We use the Invensense MPU6050 for meauring G loads in the 2g range, and angles around 0.1 degree. These have 16 bit resolution on their internal ADCs, as well as built in temp sensors. You might be able to get that level of precision with something like them.

I would be looking for tilt angle sensors designed to do this in the range and accuracy you want. Dont know of any specifically, but I am sure they exist. Might be expensive.

snipsnip · Joined: 10 Feb 2014 Posts: 74 Location: Melbourne

Another option - you can get 5000+ count encoders pretty cheap <$400 ish , Calibrating the zero pulse would be a bitch though.
But it would yield a reliable and repeatable measurement, but you'll have to cycle the tilt through the zero encoder pulse after mounting to the slab.

I've used spirit levels many times to set up machinery - good ones with 0.01mm over 1m accuracy. No way can I see a mems type devise getting anywhere close to the accuracy of old school methods. Easy enough for Bascom/avr to read the bubble position but only good for a few arcmins (-.05 to +.05mm/M) range - but no calibration required. Very Happy