Wednesday, August 12, 2009

Time to NERD IT UP

The accelerometer arrived. Things haven’t been this exciting around the house since Scully and Mulder kissed.


It’s a model X6-1A, from a little company called Gulf Coast Data Concepts. They are very reasonably priced, and should, I think, be standard equipment in the better high-school science labs across the country. I was an Art History major, and if I have one of these in my lab, you physicists have no excuses.

First order of business was to shove Chuckette into an Ergo Baby carrier, with her little pink headband holding the accelerometer onto her face, then run around and do a few jumping jacks, walk some stairs, etc. There wasn’t a lot of control in this trial run. It was mostly to practice working the accelerometer. During the test run, Chuckette was actually all the way inside the carrier (Her face wasn’t poking out as above) so the accelerometer was more snugly affixed to her face by the friction of the carrier itself. Here’s a peek at some G-force data. I did a few jumping jacks, lightly jogged about 40 feet, walked down then up a sidewalk that had some steps, then walked up the stairs in our home to the computer. Here’s what the raw G-force plot looks like for that. The X-axis is in red, and you’ll note it has a baseline of 1, because, well, that’s what gravity general gives us as a baseline.

ergo test run g-force plot

The scale isn’t so good on that beta software, so I loaded the raw data into excel and calculated the G-forces myself, and plotted them again.

Ergo Baby RMS

I plotted only the “RMS,”a blended calculation of all 3 axes, defined as such: sqrt(x^2+y^2=z^2). You can see that some of the jumping up and down, as well as a the brief jog, generated G-forces in excess of 4.5 Gs. (Those are the clustered spikes just before the halfway point.) You can see from the results thereafter that a brisk walk on the sidewalk, punctuated by walking down some long concrete steps, resulted in plenty of 2 G forces. Just walking up the smaller steps in our house (the last 1/4 of the graph, roughly) resulted in 1.5 to 1.75 G forces.

You will also note that the two graphs don’t really match up along the horizontal axis. That’s because the first one comes from the little built-in java applet provided by the accelerometer makers, and it properly plots the time. This isn’t as easy as it sounds, as the accelerometer doesn’t just run constantly (although I could set it to.) Instead, it records intermittently, only in response to changes of > 0.1 G, or once every 3 seconds, whichever is sooner. Therefore, as it fills out rows of data, it may plot several consecutive rows that are 3 seconds apart, then suddenly plot 100 rows within a few seconds. (When it is recording, it records 20 times per second at its slowest pace.

My second graph is from excel, and just gives the readings from the time value equal horizontal spacing, not accounting for the time issue.

This graph, which is simpler, is probably better at illustrating the problem. You can see that the first column of data, “Time,” plots equidistant along the X axis, regardless of the value of the cell. If things were set up the right way to properly account for time (if these were, say, seconds) those values would stretch a lot further toward the right side of the graph, and the slopes on the plot would get more gradual.

effed up graph illustration

I need to obtain a little more Excel Fu to get these right. Search engines, here I come.

Update: Here’s the proper graph. Needed just to use a “scatter” plot in excel.

Ergo Baby RMS correct

I’ll do more baselining soon.

No comments:

Post a Comment