A big X carriage problem

Hello everybody.

I am new on the forum so I want to introduce myself ...
I am a computer expert with a passion for electronics and technology. I have been interested in 3d printing for about 2 years.

My first printer was a Tronxy x5sa that, for a good start, arrived with the motherboard line driver broken ...
So 2 minutes after I finished mounting the printer, I started a printer upgrade with no object yet printed !!!
I'm not going to dwell on the upgrade, because by now the Tronxy has finished its days and has a devastated plate (too much adhesion of the PLA +) ...
For the Tronxy I created a "box" with a rudimentary filtering system that you see below:

3D Printer Box

I then decided to try to build a CoreXY with these features: 400x400x500, Duet3d wi-fi, 800W heating plate, 24v power supply (2000W max), Nema 23 motors, double Z motor with ball screws, BondTech extruder and hotend Mosquito Magnum Slice Engeenering. The movements in X and Y are mounted on rail mgn12h. All the hardware part is isolated from the electronic and power part. Signals and power are transmitted via a sort of bus built with aeronautical connectors (high currents). Duet directly controls only the extruder motor (the only Nema17). The other motors are controlled by external drivers (DM542 if I remember correctly ...). The plate is controlled by a solid state relay and I inserted a temperature probe under the plate connected to a thermostatic relay which in turn triggers a bistable relay, cutting off the current in case of too high temperatures.
The only printed part of the printer is the X axis carriage. Let me explain ... I have no problem against printed parts !!! But before reaching this result I changed a lot of configurations, especially due to the weight of the plate which is not the classic "thin plate". I can't use other plastic parts because I had destroyed the Tronxy ...

But now... I have a BIG problem:

This would be a cube...

In X the carriage is completely inaccurate.

I have disassembled the belt 5 times: without belt the carriage glides very well, both in X and in Y.
I redesigned and modified the belt return systems at least 3 times, making sure there was no friction.
I reassembled the belts ... and the problems start: whatever the tension of the belt, with the gear free to move on the crankshaft (anchor screws removed), I can feel the carriage jerking along the X axis. Once all the print has been reassembled, it slides in positive X to each single layer.

Is there anyone that can help me ? Any suggestion ???

I attach other photos of this "Frankestein" ...

Frank1

Frank2

Frank3

Frank4

Frank5

Frank6

Frank7

Frank8

Frank9

Frank10

Frank11

PS: I have some doubts about Duet3d's expansion card control on external drivers ... Has anyone had any kind of similar problem with a comparable configuration?
Thanks in advance to all.
Hello everybody (an sorry for my English... It's not perfect...)

Shifting layers at 45/135 degree angle indicates that one of the motors is not keeping up with the input. Check cables and connectors between controller and drivers. Also, check jerk and acceleration settings- either or both can cause layer shifting if they are set too high. Is there a reason you're not using the drivers on the Duet board? You may need to tweak the timing parameters for the external drivers to get them to operate reliably. The timing parameters are in the M569 commands in the config.g file.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Hi The_Digital_dentist.

Thank you for your kindly answer.
You give me something to check. Unfortunately I can't check connectors and wires until Saturday...
The acceleration setting is not high... I had lower all setting values for test.
I use external driver for two reasons:
1) I want to be able to change motors (nema 23) with other kind of motors with highest current if I need (I use plug-in connectors for engines);
2) the installed motors run with a max current very near to the max current that you can use with Duet3d internal drivers and I don't want to stress the mainboard.
However I can use direct connection for test and try what happen using this way.

About M569 command I use these settings:

; External Breakout Drives Setting
M569 P5 R1 T2.5:2.5:5:5 S1 ; physical drive 5 goes forwards (X)
M569 P6 R1 T2.5:2.5:5:5 S1 ; physical drive 6 goes forwards (Y)
M569 P7 R1 T3:3:6:6 S1 ; physical drive 7 goes forwards (Z1)
M569 P9 R1 T3:3:6:6 S1 ; physical drive 8 goes forwards (Z2)

The right timing setting would be T1.5:1.5:2.5:2.5 for this kind of external drivers.
I set a more prudential values with more long signals. At this moment I don't search for speed...
I want only a well running printer.

If I can't solve... I'm thinking to redraw the X and Y carriage with ball screws.
The problems may be spaces and moved masses... It would be an hard job...

Thank you again for your help.
I will repotrs the results of my checking asap.

Have a nice day.

wec71

Hi The_digital_dentist.

I checked all wire with a multimeter and I saw the same resistent value (1,8 Ohm) on any single wire.
I tested the X movement setting M569 on highest value than what is written on external driver manual.
I had no time to print a sample object, but a multiple series of move on right and left seems to be correct.
Is it possible that the value written on the manual are too much near to the limits ?
As soon as possible I will test a printed object.
Thank you again.

wec71

The pulse waveform from the controller can be distorted by the time it reaches the driver, especially if the cables are long. Also, in my experience, you can't really trust specs on Chinese made electronics, especially if they don't tell you how the testing was done, so the drivers may not perform exactly as specified.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

How have you connected the external drivers to the Duet? Have you used the Expansion Breakout Board or other level shifters to boost the signals to 5V for the drivers?

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Hello guys.

For The_digital_dentist: the cable to the controller are 20 cm max. I think you are right with China quality...

For dc42: the external drivers are connected to the expansion breakout board. To avoid 3,3v to 5v logic problem, I connect all the negative signals to 5v from the expansion board breakout (as specify in the Duet manual).
Maybe that the connection to 5v are not a solution, but i can't check without an oscilloscope... I search for a shifter levels ic, but I see that the max frequency is 80kHz. The declared max frequency of the external drivers
is 200KHz. I must do some check to see how 80kHz can limit printer speed.However I don't like to introduce a limit in speed...

Thank you again for your help.

wec71

Quote
wec71
For dc42: the external drivers are connected to the expansion breakout board. To avoid 3,3v to 5v logic problem, I connect all the negative signals to 5v from the expansion board breakout (as specify in the Duet manual).

The wiki page specifies that when using typical external drivers that have optically-coupled 5V inputs, you should connect the positive driver inputs to +5V (which you can get from either of the two PWM connectors on the expansion breakout board), and the negative driver inputs to the STEP-, DIR- and EN- pins on the output connector of the breakout board.

For faster response, post at [forum.duet3d.com].

Edited 1 time(s). Last edit at 05/10/2021 12:22PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Yes DC42.
You are right. It's my mistake.
Sorry, I wrote what I remember, but the connection that I made are right.

wec71

Updated news...
No problem solved. I changed the signal lenght to 10us, but nothing changes.
Now I have found a level shifter that I can try...
But I'm not sure that this would be the problem...
My next test will be connects X and Y engines directly to Duet, without
external drivers.
However... I have just now placed an order for linear actuators to
transform the X and Y assex in a full ball screw driven system.
I had always want to try this solution...
Regardless that I will solve or not solve the problems in the present configuration.
Thank you.

wec71

You have at least one problem with geometry of your printer. Belts on X should be parallel to X axis in XY and XZ planes.
Your very first image shows problem on a left side.