Ensuring Consistent Sampling Interval When Reading ADXL345 Data via SPI at 400Hz

Hi,

Greetings for the day !

I am using ADXL345 with SPI communication in my project. I configured ADXL345 in

 -> Full resolution mode with +/- 16g,

-> data rate 400Hz,

-> SPI clock frequency as 327.6KHz

-> Operating in Bypass mode

-> Mapped Data Ready interrupt to INT1 and enable Data Ready interrupt.

I have to read 100 samples of X,Y,Z values from ADXL345 and i am reading the data only when data ready interrupt occurs. but the problem is there is no consistency in timing of data. (i.e., the time period varies from sample to sample).

You can find the screenshots of them below.

 
What should i do to get a consistent data from ADXL345 ?

Am i missing anything ?

Any help would be appreciated.

Thank you.

Rajkumar

Hi Raj,I've played with the ADXL345 a few times in the past. I just want to clarify, you want to read 100 samples at regular consistent intervals? Would it be easier to let your microcontroller handle the timing then instead as it has it's own clock and could give more accurately timed results? What microcontroller are you using?The accelerometers have quite a bit of functionality in them. There's a number of things that could be happening here. It could be an inactivity timeout, meaning additional time to wake up, although it should be relatively quick enough. You could try increase the inactivity time (TIME_INACT) to something larger than your sample time. What level have you set the activity threshold to (THRESH_ACT), maybe it's not triggering an event? Maybe try a lower value here?Another guess would be clearing the interrupt in a timely fashion by the microcontroller. Every time an interrrupt happens, it needs to be cleared by the microcontroller. Judging by your screenshots, the interrupt signal latches when an event happens and should only go low when you clear it. This should be the case, but it is something to check. Does the signal stay high, if you don't clear the interrupt?
-Mike

Hi Mike,Thank you for your response !I am using MK60FN1M0VLQ12 controller from NXP. I am reading 100 samples of (X,Y,Z)data only on getting the DATA READY interrupt. I am not enabling any other interrupts like TIME_INACT, THRESH_ACT any other. As of now, i am not getting DATA READY interrupt from ADXL with constant time period between consecutive occurrences. I am reading data only when DATA READY interrupt occurs and clearing the interrupt by reading X,Y,Z values from the ADXL registers.I observed that, the signal goes to High when Interrupt occurs and goes low when it is cleared. The signal will not become low until it is cleared.Can you tell me if I am missing anything to get DATA READY interrupt with consistent time period between consecutive occurrence of DATA READY interrupt.Any help would be appreciated.Thank you.-Rajkumar 

It is still not clear how you are initiating sampling of the accelerometer.The usual way to do it is to set up a  processor timer, and generate an interrupt from the timer at the rate you require, 400Hz, as  Mike said. On the processor timer interrupt, you send out a CS to the accelerometer, and a start command for one reading. This ties the accelerometer to the processor clock, the accelerometer data ready will subsequently be issued, and provided it arrives and the data is fetched before the next CS/start command is issued, then it should be very consistent. Normally, on reading the interrupt input, the latch should be cleared automatically, but some processors may require an explicit reset.Is that what you are doing?
I assume that the scope was set to some form of peak mode detection, scopes can miss short pulses in a long sequence that may actually be there, which is why a logic analyser wins hands down for this sort of work.
Cheers,Richard

For that accelerometer to the processor can be done with dell printer error 016-302 helped me to get the solution.