Topics created by Garth_42

@Garth_42 when you try to calibrate and it gives the message that motion was less than expected, does the motor turn forwards and back slightly, and does the message give you an estimate of the CPR? If so, is it close to the actual CPR of the encoder?

@o_lampe thanks for the thoughts. Why would cancelling during a PID tune affect the database? Does it create a temp file?

Yes, the user would need to make sure that they are using the correct FW version with a config if they use an older one.

I think that the major use cases are to make sure that if you make too many changes, or aren't sure why you are getting a specific behavior from the config after a change, you have backups and can find one that is working, as well as backup the most recent version of the config such that if you loose that data from a bad SD card/etc. it isn't gone forever.

It makes sense to reduce the backup check frequency as if a user is testing changes, then it doesn't sync many copies of those files.

@chrishamm ahh... I see. This makes sense. I assumed the FTP server in standalone was extant in SBC mode, too, and now I see my error there. I tested this as well and it is working on my board.

Thanks for the info on best practices. I haven't seen that information on the wiki.

Best,
Garth

@hebigt unfortunately that level shifter is for logic signals only. I suggest you use two resistors, as @T3P3Tony suggested. For example, 2K4 between the sensor output and the Duet input, and 3K3 between the Duet input and ground. These values assume that the sensor is capable of sourcing at least 1mA of current.

@garth_42
Hi all, I posted justa bit too quickly. I figured out the syntax. For anyone in the future, this is the syntax I used:
M117 {"Z-Axis Position: " ^ move.axes[2].userPosition ^ " "}

bda88bb5-4feb-4265-8800-74aeb91a185f-image.png

@deckingman Thanks for the response. That is the case, and all the tests were performed at 0.5:0.5 index.

The only way I'm aware of suppressing the message is to disable the driver entirely. I can move this to the wishlist and see if DC42 has any other ideas.

@o_lampe Ah I see, that is quite clever. Thank you for the clarification.

@garth_42 you can test if this is an issue with the timing/formatting of your network request vs what is been reported by making a move and using echo command to report what the printer status is during the move. very simplistically:

sent: echo state.status received: idle

then

sent: G1 Z25 F2000 sent: echo state.status received: busy

this is just using the console in DWC

@Garth_42 Sorry, but I had the same problem, how did you replace the file?

@Garth_42 said in Extrusion Differences with Mixing Extruder:

For example, a single G1 E1500 command will extrude about half of the material compared to 30 G1 E50 commands.

Is this something you can model mathematically? That is, the actual material flow rate during a movement?

If so, you may solve the problem by preprocessing the slicer's output, transforming the extrusions and gantry movement to fit your model. And then disable the duet's acceleration and jerk restrictions such that it will execute exactly the program your post processor generated.

Just a thought.

@bearer thanks for the knowledgeable response and helpful link. Very good to be aware of going forward.

@dc42, that would be great but if it isn't quick don't worry about it. We have other workarounds to the issue and while what you proposed is the most elegant solution, it isn't necessary yet. In the future it may become more important though.
@chrishamm thanks for the suggestion, I will keep that in mind as we are currently using the DSF for some other functions.

@seraser glad you finally compiled it! Sorry I didn’t respond in time.

Thanks both for your suggestions! I am slowly making progress. That repo is actually what I am using to develop a solution and am getting close. I am a little confused as to some of the matrix operations but I should be able to replicate it nonetheless.