3.4 rc1 1XD Probe problem after update

dc42

@x_pedro_x @tom33 the significant difference between 3.4rc1 (also 3.4beta7) and earlier releases is that in 3.4rc1, when a motor controlled by a CAN-connected board is stopped because and endstop or Z probe is triggered, the slight overshoot that occurs is corrected after a short delay by reversing the motor movement and sending whatever steps are needed. If you are using closed-loop drivers, perhaps they don't handle this situation very well? If so then reducing the speed of the homing or probing move may help.

x_pedro_x

@dc42 Thanks for the quick reply, on my eagerness to post I forgot valuable information:

I'm using a Duet 3 6HC with an SBC
Open loop steppers

In reality it does look like what you explained every now and then the stepper would make a sound like I had mixed the coils (usually after the 2nd probe location, but also on the 3rd) .

dc42

@x_pedro_x please post your complete config.g file so that I can reproduce your setup.

x_pedro_x

@dc42 Hi,

You may find all the files here

Thanks for the help,
Pedro Oliveira

tom33

@dc42 I don't have the problem for the homing just for the probing. And Reduce the acceleration or speed doesn't work for me

dc42

@tom33 @x_pedro_x this has now been fixed. Please try the new expansion board firmware at https://www.dropbox.com/sh/2dt7sbqpx6l74np/AADn4-lpcil1iqnWKkiVri3Ia?dl=0.

x_pedro_x

@dc42 I will have to wait a bit, unfortunately i destroyed my sbc (forgot that the camera link is not hot-plugable) and am waiting for a new one
As soon as i receive it I'll try the FW.

Marcossf

@x_pedro_x @dc42 I think we have the same problem than you, as David said in our post entry. We will test it tomorrow to see if the 3.4RC1+2 solve the issue.

David, do we use all RC1+2 dropbox updated files, or only the one for the 1XD board?

Marcossf

@x_pedro_x Hi Pedro, we've tested with the David's 1LC 3.4rc1+2 files and it's working fine.

tom33

@dc42 I try this today !

x_pedro_x

@dc42 Hi, I tested your files and the homing could be done properly.
Just to confirm I reverted to the public rc1 and the problem happens again, using rc1+2 behaves as expected.
Thanks for the help.