View Full Version : Using the ADXL202 instead of the Memsic 2125 Dual-axis Accelerometer

Russ Ferguson
07-26-2005, 10:40 PM
I would like to have a schematic drawing showing how to connect the ADXL202 to the BS2, and some sample code.

The "BASIC Stamps Archive" contains several threads about using the ADXL202, and even indicated that Parallax sold the part at one time. If anyone has copies of the circuits and the code that was developed at that time I would appreciate getting a copies of it.


Ken Gracey
07-26-2005, 11:24 PM
Hello Russ,

About four years ago we sold the ADXL202EB evaluation kit. Our package of web downloads is attached. I used this part on my R/C airplane as an in-flight level controller.

Ken Gracey
Parallax, Inc.

Russ Ferguson
07-27-2005, 12:05 AM
Thank you for the info.

Tanner Electronics received a small quantity of the parts in 14-Lead CERPAK (not the evaluation board). They will need to be mounted on a surface mount adapter board to be used in the way that the MEMSIC parts are used.

Given the attachments that you provided and the information on the evaluation board from:
it will be possible to use the IC with the BS2.


Russ Ferguson
08-09-2005, 02:15 AM
In examining the documentation for the MEMSIC MXD2125GL and the Analog Devices ADXL202JQC
I noticed a difference between the formulas that each spec gives for acceleration.

MEMSIC Spec sheet for MXD2125 Rev D page 5 Figure 3
A(g) = (T1/T2 - 0.5) / 20%

Analog Devices Spec sheet for ADXL202 Rev B page 6 Figure 12
A(g) = (T1/T2 - 0.5) / 12.5%

The Parallax documentation uses the formula from the Analog Devices Spec sheet
"Memsic 2125 Accelerometer Demo Kit (#28017)"
page 2 Figure 2
A(g) = (T1/T2 - 0.5) / 12.5%

Is this a typo on the MEMSIC spec sheet?

Ryan Clarke
08-10-2005, 10:48 PM
Hi Russ, sorry this wasn't answered sooner.

The value of 12.5% is correct (per message from Andy):

Either the datasheet is incorrect or it covers multiple chips. The pulses that come from the thing have about a 10 ms cycle time, and when they are horizontal, the duty cycle of the pulses is 50%. Taking one axis from vertical to horizontal, to vertical the other way results in a variation of +/- 12% in the duty cycle. At room temperature, this tilt sequence will give you high times that go from 3.75 to 5 to 6.25 ms (assuming exactly 10 ms cycle time). The duty cycle varies sinusoidally with tilt, so inverse trig will give the correct angle. Tracy Allen (www.emesystems.com) published a technique for doing that with the BASIC Stamp, and Jon Williams incorporated it into the example code in our datasheet.

Hope this helps.


Ryan Clarke

Parallax Tech Support
rclarke@parallax.com (mailto:rclarke@parallax.com)

Russ Ferguson
08-15-2005, 06:38 AM

Thank you for looking into this and posting this response.