InputShaping Plugin shows only zeros with Toolboard 1lc

[mfs12]

@Heartleander81 can you share the acceleroemter csv file with the new config (without R12 parameter)?

[Heartleander81]

@mfs12 from your is plugin I doun't find the file. In the Folder Accelerometer is only the csv files from manuelly test.

[mfs12]

@heartleander81, please run the manual test again and share the file.

[Heartleander81]

@mfs12 20_2021-10-08_12.32.09.csv

[Heartleander81]

ACC from duet

IS Plugin

Sorry for the Bad Picture Quality

[mfs12]

@heartleander81, @dc42,

The plugin behaves here correctly as the file reports a sample rate of zero. This is why only zeros are displayed. I am not sure why RRF reports a sample rate of zero despite recording all those samples.

From https://forum.duet3d.com/assets/uploads/files/1633689320062-20_2021-10-08_12.32.09.csv:

998,-0.031,-0.125,1.039
999,-0.031,-0.125,1.039
Rate 0, overflows 0

[dc42]

@heartleander81 did you resolve this issue, or are you still getting only zeros along the bottom of the plot?

[Heartleander81]

@dc42 There are still zeros displayed. Unfortunately.

[mfs12]

Another user reported a similar issue.

https://forum.duet3d.com/topic/25745/input-shaping/

[Heartleander81]

@mfs12 I have See it, thanks.

[Heartleander81]

Good Morning.
I have now updated to 3.4 Beta 6 and Pre 7.
Unfortunately the problem that the frequencies are not displayed is still there.

[dc42]

@heartleander81 please can you attach the latest .csv file that was generated when you ran the IS plugin.

[mfs12]

@Heartleander81, i added now an option in the "record tab" to not delete the generated files by the IS plugin.

[Heartleander81]

is-393cb560-none-X-0.csv

[Heartleander81]

@mfs12 When I look at the IS file in the Duet accelerometer, it shows frequencies.

[mfs12]

for some reason your device report a rate of 0. which is strange and causing the problems.

[Heartleander81]

@mfs12 OK. What could be the trigger for this?
Is there a way to do tests to find out what might be the cause?

[dc42]

@heartleander81 I am still trying to work out how this could happen.

[dc42]

@Heartleander81 I've managed to reproduce this by disconnecting the interrupt signal from the accelerometer. So I think your board has a hardware fault. Which revision of tool board is it? If it is a rev. 1.1 board, does it have a diode soldered on the underside in one corner of the board?

[dc42]

@heartleander81 I've added a test of the interrupt line in the firmware:

• Please install the tool board firmware from https://www.dropbox.com/sh/wtgcvkuaq52x3aq/AAADVEzSwypcIZSwExX0M4RQa?dl=0 on your tool board. It is compatible with 3.4.0beta6 main board firmware.
• When that's done, send this command to attempt data collection: M956 S1000 A0
• If that command reports failure, run M122 B20 to get the tool board diagnostics.