Hi, I have a huge probleme. I bought some adxl345 3 weeks ago and i can't get the input shaping to work. I've found some tutorials on the internet but they are on much older versions of mainsail and klipper and I think that they works pretty differently compared to the most recents version. My printer works beautifully under klipper with the mainsail interface and the wiring of my adxls are good but I'm sure that the problem come from the software side and I haven't really founded some documentation on the version that I have (1.2.1) and I'm terribly lost in all of this. can someone help me by giving me the link to some documentation eor a step by step guide?
#input shaping problems
47 messages · Page 1 of 1 (latest)
what's wrong with the documentation on the klipper site?
nothing is wrong but it tells to install numpy and matplotlib for an example. but it is already installed and it just does a bunch of errors. these documents didn't work for me.
Dunno. This seems very clear on what to do, If you are getting errors you could post what you did and what the error was . https://www.klipper3d.org/Measuring_Resonances.html#measuring-the-resonances
@sick girder MainsailOS has numpy installed for you as it could take long time and weaker SBCs. The Klipper docs on the other hand a written for a base Linux system
So how about describe your Problem … And to be clear in 99% it is the wiring of the sensor as SPI is not that easy and especially cable length is a big factor here
since numpy is already installed I added the following lines to my printer config like said in the documentation:
[mcu rpi]
serial: /tmp/klipper_host_mcu
[adxl345]
cs_pin: rpi:None
[resonance_tester]
accel_chip: adxl345
probe_points:
100, 100, 20 # an example
but then it stars talking about some micro controllers wich a don't have in my setup
so I assumed that the following steps were not for a setup like mine (with only a pi3 and no microcontroler) so there i^m kinda lost. i've seen tutorials to configure it as a linux process and then flash the device but it didn't work so I don't really know what to do.
To be clear, you first install the software following this: https://www.klipper3d.org/Measuring_Resonances.html#configure-adxl345-with-rpi
Then you skip the other software installation parts that are clearly for other devices, so you will end up here: https://www.klipper3d.org/Measuring_Resonances.html#measuring-the-resonances
There you can continue.
If it's not for your setup, simply ignore it and continue inside the tutorial
There is a legend on the right side, that you can get faster to the point where you have to continue.
thanks to your help I made it a step further. before when I added the lines for the adxl in the printer config the firmware woould keep crashing but now it works but i still have to configure it with a .cfg file but I don't know what to put in it and where do i put the file
I forgot to say that the thing i have to configure is the "secondary mcu"
As we don't know where exactly you are stuck I can just link the secondary mcu section of klipper: https://www.klipper3d.org/RPi_microcontroller.html
If you need more specific help, say at what exact step you are right now in the tutorial
you still only have one cfg file for klipper, even though it is controlling the pi host as well as the printer mcu.
i'm stuck at "remaining configuration"
Skip all of that and go back to the adxl config
THANKS !!!!
so now that I've included the adxl config to the printer config the firmware keeps crashing. it says "unknown spi_bus 'spi3a' "
whait sorry I messed up. I have to check the wiring
That is an unusual spi bus. Can you post your printer.cfg?
I think it comes from a file that i've seen people include in their printercfg. Its called "adxl345" cfg i added it and in it there is a line that says "spi3a" i think taht this file is a problem but when i remove it if i send accelerometer_query there is an error.
i've checked the wirring and it is not the problem. All the labes are about 3 ohm so they're good
yea, that adxl345.cfg will not work for your setup. The printer.cfg on it's own looks reasonable, what is the error? And you enabled SPI on the system?
you might want to use this https://github.com/nagimov/adxl345spi to query the chip outside of klipper.
GitHub
ADXL345 three-axis accelerometer reader (SPI interface) - GitHub - nagimov/adxl345spi: ADXL345 three-axis accelerometer reader (SPI interface)
so now i've been able to do the resonance test on the y axis and it went fine but when i unpluged the y axis adxl and switched to the other one and i verifi the connection with accelerometer_query it tells an error
it might be a faulty chip beacause it still works when i plug it back into the other one
you have 2 different adxl's and you just swap the connected pins? could be a bad chip, or you could accidentally be grounding something the way you have it mounted.
i know what happened
on the x axis the way I mount it is a bit wierd and i screwed it a bit to tight and it bended it very very slightly and i might have cutt some copper track of. i didnt know this until i swapped them (mounted the y chip on the x chip) and i broke the other one
so now it works but i have to wait until i get some new chips delivered
So now i'm stuck again. I've checked the connections on the boards with a continuity tester but they both look fine.
So mybe the error is somewhere else
Do you have an idea of what could be the issue?
@sick girder that is a SPI connection, physical connection is only a small part here. It runs a protocol and is very sensitive to cable lenght and cable impidance. I normally use a cut ethernet cable to connect the ADXL and make it as short as posible but never longen than 1m
But it worked for the fist test perfectlly but both of the chips stopped working as soon as i mounted them on the x axis
The error can not be caused by the cable
mounted may be the issue here ... The most versions of PCB that I now have a PCB via very close to a screw hole. You need to use a plastic washer otherwise you scrach the coating and make a short that destroys the chip
Fun fact i actually did screw the board with some spacers on the x axis but not on the y and it was the x axis who caused issues