ADXL335 output signal

Hi, i'm using ADXL335 accelerometer to measure some angular acceleration of my robotic system. I'm using avantech DAQ board, and i log the output of the sensor through Simulink's real time workshop . the problem is, i'm getting a sinusoidal signal from the accelerometer, as compare to the constant 1.5V for a fixed position (0g). Can any1 help me on this problem? thanks

Schematic?

Is gravity working correctly?

Probably bad or non-existant decoupling, add an 100uF and a 100nF caps to your voltage supply lines, near the sensor.

Common ground?

the problem is, i'm getting a sinusoidal signal from the accelerometer

How would you know that without using a scope or graphing the output?

If you did that then you should have mentioned the amplitude. While the ADXL335 is one of the "low noise" accelerometers you will still see what appears to be a sine wave on the output of each pin if you look for a low value AC signal....

Yes the earth is constantly moving, yes the ADXL335 still detects/produces noise -- and maybe you were seeing the earth ringing because of the earthquakes off Japan - or California.

So no circuit, no really useful info...

for the accelerometer, i've logged the output through simulink's scope function block. it's a direct connection from power supply to the accelerometer then to the DAQ, as seen in the data sheet. i've added the capacitor recommendation as in the data sheet, 0.1uF at the input and 0.0047uF at the output x pin. FYI, i'm using only the x-pin and left out y and z open circuit.

i'll update the schematic and amplitude of the signal once i've back in the lab

AWOL: Is gravity working correctly?

AWOL, you make me smile. ;)

Lefty

this is the schematic depiction (sorry if its a bit messy, drawn using simulink)

and the output signal

untitled.JPG

ADXL335.jpg

legolasx: this is the schematic depiction (sorry if its a bit messy, drawn using simulink)

and the output signal

Two flaws that I see. First you are powering a digital device via a voltage divider, that's not good. As a digital device doesn't have a constant current load, your voltage divider will be providing different Vcc voltage to the chip at different times. Change the 15k resistor to a 3.3 volt zener diode or power the chip with a 3.3vdc regulator chip instead of that two resistor voltage divider.

Second, you don't show a ground connection from the chip to a arduino ground pin, that is a requirement. If you don't have such a connection then your waveform is from just 60hz noise being read by a floating analog input pin.

Lefty

Re: the output signal: 20...what?

Is gravity working correctly?

AWOL, you make me smile.

Lefty

How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth? Sherlock Holmes

@ AWOL, its voltage versus time.

@ Lefty, if i drive the adxl335 with a direct 5v DC from the source will it be better, considering the accelerometer can accomodate the voltage. for the arduino ground, i'm not using any board from arduino, but the ground pin is the 0V-5 as in the schematic.

and the thing is, when i measure the voltage at the terminal board point, it gives out constant 1.65V, while reading in simulink it gives out the sinusoidal signal

@ AWOL, its voltage versus time

Megavolts per century?

Don't teachers tell you to label your axes these days?

(it may not seem like it at times, but I'm trying to help here - some cooperation and common sense would be useful)

The voltage divider should not be an issue.. It can provide 2ma at 3.3 V (more or less) The IC draws 250uA. A larger capacitor should help if there is an issue.

Not labeling the graph or providing the info is not helpful. I can assume it's volts and seconds... maybe not.

As I said previously there is always some noise on the output. Without knowing thee scales it's tough to comment.

Using a 40MHz scope... An MMA7361L sits at about 1.65 and has less than .1v noise -- looks somewhat sinusoidal.

An MMA7361L sits at about 1.65 and has less than .1v noise

Isn't that about 1/8th of a g? That's quite an error.

“quite an error” maybe/maybe not

I don’t know how much I trust that scope or the calibration signal.

I did say < 0.1V I really did not analyze it that much – some was HF ripple – some LF ripple.