Go Down

Topic: ADXL362 signal-to-noise ratio way too low (Read 792 times) previous topic - next topic

Genoil

Hi,

I wired up a Sparkfun ADXL362 breakout exactly as described in the tutorial by AD, and it basically works as in returning values, but the data I get back seems almost random. When I lay it down on the table I would expect values fluctuating around 0, but nothing like this:
Code: [Select]

XDATA = -32 YDATA = 792 ZDATA = 448
XDATA = -289 YDATA = 272 ZDATA = 688
XDATA = -799 YDATA = 524 ZDATA = 688
XDATA = -285 YDATA = 272 ZDATA = 1196
XDATA = -32 YDATA = 272 ZDATA = 1204
XDATA = 224 YDATA = 524 ZDATA = 1188
XDATA = -285 YDATA = 280 ZDATA = 1204
XDATA = -285 YDATA = 780 ZDATA = 448
XDATA = -32 YDATA = 512 ZDATA = 688
XDATA = -32 YDATA = 272 ZDATA = 700
XDATA = 224 YDATA = 272 ZDATA = 688
XDATA = 224 YDATA = 280 ZDATA = 700

In the default settings, 1 == 0.001G, so it sometimes fluctuates about 0.3G, which is 1/6th of the total range (+/-2G), when it's not even moving?

I must either be doing something wrong, or the unit is broken, or I simply have to accept this level of noise. Does anyone else have experience with this part? I'm using an 16Mhz Arduino Pro Mini 168 @ 3.3V.

pito

Provided the table is not placed in a moving train :) the results are not good. The noise should be something about +/- 5 max at the lowest digit when built on a solder-less breadboard when laying still.
Rotating the sensors in Earth's gravity you have to see +/- 1 G clearly (when still).

robtillaart

First thing I would check is the wiring - optional rebuild it - bad contacts is one of the most occurring "breadboard bugs"

How long are your cables?
Rob Tillaart

Nederlandse sectie - http://arduino.cc/forum/index.php/board,77.0.html -
(Please do not PM for private consultancy)

michinyon

You should not "expect to see values fluctuating about zero".

You should expect to see values corresponding to a vector of 1g  pointing towards
or away from the centre of the earth.

Go Up