DIO counter glitch

Topics about the Software of Revolution Pi
Post Reply
csmf
Posts: 1
Joined: 09 Sep 2020, 12:27
Answers: 0

DIO counter glitch

Post by csmf »

Dear All,

I'm working on a device to measure frequency and current from multiple intputs. For this purpose I'm using a rev pi core 3+, DIO and AIO with the same physical order.

My program is made in C/C++ which is sampling one of the DIO's counter and multiple AIO inputs. The sampling is done at a fixed frequency of 20Hz.

I'm observed that the read values from the process image, using functions from demo code, every ~8-24 hours contains wrong/glitchy value for one cycle. Not only the DIO's counter value becomes wrong for one cycle, but most of the case one of the analogue input as well.

During my 10 days long testing I was collecting log files and I've seen, 10 times the glitch is happening. In all the case the counter's value and most case atleast one current input contains "glitchy" value.
An example:
DIO's counter value:
cycle x: 98732742
cycle x+1: 1345257472
cycle x+2: 98732791
The gradient of the increase stay the same, only a glitchy value read from the process image.

For analoge inputs I've seen transition from 1 (no current applied to any input) to 20, in the same cycle when counter's glitch happend.

For the counter I've set the InputDebound enabled. I've checked the RS485Cnt and it is stayed at 0 during the 10 days.

Version numbers of the boards I've used:
RevPi Core V1.2
RevPi DIO V1.4
RevPi AIO V1.3

Did anyone observerd similar issue?

Regards,
Marton
User avatar
dirk
KUNBUS
Posts: 1926
Joined: 15 Dec 2016, 13:19
Answers: 4

Re: DIO counter glitch

Post by dirk »

Hi Marton, this seems like a known issue that we have fixed in the Git Hub repository: https://github.com/RevolutionPi/piContr ... 4ff7b5b964
I'll send you a PN with a pre-release Kernel package. I hope that this will fix your issue.
Post Reply