lil forces

DMKMM
Posts: 103
Joined: Thu Apr 17, 2008 12:12 pm
Location: Lincoln, NE

lil forces

Postby DMKMM » Wed Jun 04, 2008 8:22 pm

Hello again...

im having trouble finding a very small 3 axis accelerometer, i've looked in to the IMU06 but its a little large for my needs. as an example; the type of stuff im looking to do is like take the lil sensor and put it like on a head light to see how much vibration is seeing, things like that... i would also like it to be relatively easy to hook up to a DL2 and be able to record.

any ideas? links would be awesome.

-thank

ZEDohSIX
Posts: 40
Joined: Sat Nov 19, 2005 3:50 am

Postby ZEDohSIX » Wed Jun 04, 2008 9:01 pm

Try one of these...

http://www.sparkfun.com/commerce/produc ... cts_id=849

The hookup is straight wire to the 5V/ground and analog inputs. I used another of these modules and housed it the plastic case that came with some SDram I bought. If you want it watertight, either pot it with the proper stuff or otherwise just fill the case up with Araldite or other 2 part epoxy glue.

Of course calibration and signal processing are then up to you and you have to either manually compensate for or put up with the drifts and temperature effects that will occur.

tristancliffe
Posts: 220
Joined: Thu Mar 15, 2007 3:41 pm
Location: Norwich, UK
Contact:

Postby tristancliffe » Wed Jun 04, 2008 9:57 pm

If you are actively trying to measure a vibration then you'll need to be sampling the data at a high frequency - 1000Hz, maybe more. If you're just trying to see how wobbly something is then you might be okay. But bear in mind my rigid car, with the engine off, not moving, inside out of the wing still has quite large fluctuations on the accelerometers I'd questions it's ability to actually be useful...

But I'm know to be wrong most of the time, so you'll probably be fine now I've talked up problems.
Monoposto 2000 - Reynard 883 Toyota

faraday
Posts: 267
Joined: Sun Feb 25, 2007 1:18 am

Postby faraday » Thu Jun 05, 2008 5:06 am

Some more detail please Tristan...
Are you speaking of the internal DL1 accelerometers?
These may be of the same family as recommended above. The 3-axis version quoted has a lower frequency response than the 2-axis units, but even its slowest axis (z) is over five times the DL1 logging rate.
The sensor will not be the limiting factor. They are quite low noise (i.e. can resolve low "g"s)

Give us a picture of your stationary noise please Tristan.
:cry: You may have a hardware problem...
I assume you are talkind about wing and not wind :oops:
But, you are quite correct...

The DL1 is not designed or built to measure vibration, with its internal or external sensors. The problem is its sampling rate, but as Tristan hopes, it will probably be fine for what you want to do.
I doubt that you want to capture the waveform of the vibration or put accurate quantitative values on it.
Bear in mind that the mass, etc of your sensor...
You should have no problems with most automotive structures with the ADXL322 at 100 Hz. The amplitude could be a bigger problem than the frequency if we are talking about shoddy engineering.

My only practical experience is on a go kart trying to achieve a satisfactory mounting for the DL1, using the DL1 itself as the sensor.
If you are talking about headlights mounted on a post, a la clubman, you can probably see the shake, so the frequency is low. Structures on the kart were not obviously resonant, but the DL1 told a differnt (believable) story.
The amplitude of the wobble was such that the acceleration at the change of direction was well over 2g. The waveform was clearly clipped at the limit of the internal sensors when the vibration was started by a simple whack.
On track, however, the forcing vibration from engine and road irregularities, etc, produced a much dirtier waveform that the 100 Hz sampling could not capture properly, even though I'm confident that the fundamental frequency of the structure was < 10Hz.
:?
Think about what frequencies exist in the car's systems that might excite the components of interest. An engine at 6,000 RPM is an obvious starting point... 100 Hz...

You could try to measure the stiffness of mountings. A spring balance and a tape measure or vernier caliper may be sufficient to get a ball park figure, then look at the mass, guess at its distribution and do some calcs on MOI, etc. The maths shouldn't be too hard.
Before you start, however, look at the maths to calculate the parameters of interest to see what accuracy you need in the measurements. Might just mean you need a dial gauge, however.
:wink:


Return to “Sensor-related questions”

Who is online

Users browsing this forum: Bing [Bot] and 27 guests