X axis seems to drift

#1

I’m doing some test rectangles and the X-axis seems to drift in one direction. I’d estimate 1 mm every 10 cm. It’s consistent and repeatable, and can’t be accounted for by pen slippage. What’s the process for diagnosing this?

I can reproduce it by repeatedly overstriking squares – the result is thin horizontal lines and thick vertical lines as each pass is slightly offset. Doesn’t seem the size of the square matters. For example:

G1 X150 Y150 Z15
G1 X50 Y150 Z35
G1 X50 Y50 Z35
G1 X150 Y50 Z35
G1 X150 Y150 Z35

I’ve tried swapping stepper drivers, changing step sizes, driver current and even swapping stepper motors but still the same result. Retightening the set screws and timing belt also has no effect.

If I try a more complex drawing the carriage quickly diverges and runs into the X-axis limit switch.

FWIW I’m using the command-line tool from https://github.com/duelafn/makeblock to send commands since I’m on Linux and can’t get GRemote to work. Also had to apply a patch for the Uno board (https://github.com/Makeblock-official/XY-Plotter-2.0/pull/17). Not sure either of those should introduce a one-axis bias though.

Example with a U-shaped object:

#2

Hi Sehugg,

Do you use XY Plotter 2.0? And the main board of your XY is Orion board?
Could you upgrade firmware for the XY Plotter with mDraw software or Benbox and draw a rectangle. Please kindly send the draw result.

#3

I got it working with GCodeParser. There was a lot of old Reprap code that was causing problems with the stepper driver.

#4

If it’s any help to you, when I was trying to adapt GRemote and GCodeparser for what I wanted, I also had a constant shift along the axis driven by a single belt. When I changed to MDraw this problem vanished. I put it down to two reasons:

  1. The XYBot.ino file contains a simple speed ramp-up/ramp-down that means the carriage has probably come to a complete stop before reversing direction and being driven back in preparation for the next pass.

  2. The revised stepper settings on the motor board might have also changed the charatceristincs when the stepper is halted and then undergoes a reverse in rotation.

#5