Using RAMPS 1.4 with a 5 wire stepper

I have some 5 wire steppers I want to use with the RAMPS board. I know that the two common drivers for this are the A4988 and DRV8825. However, these two are NOT compatible with a 5 wire unipolar stepper. Are there any alternative drivers I can use with the RAMPS 1.4 shield to make this work?

I don't think so. The RAMPS board is designed so you can plug in the A4988 etc boards.

Can you convert your motor to bipolar use by splitting the 5th wire into two to make it a 6-wire servo with two completely independent coils. If you can then just ignore the two wires to the centres of the coils.

...R
Stepper Motor Basics

I do actually have some 6 wire steppers as well (which, although unipolar, are supported by the A4988). However, they are pancake motors that don't seem to have much torque to them. I will test their torque later today, but how much torque would actually be needed for a delta style printer?

electricviolin:
but how much torque would actually be needed for a delta style printer?

How long is a piece of string ?

There is some rough guidance for measuring torque in Stepper Motor Basics

...R

Its about 1 m total (half up and half down to make a loop)

I an writing in this topic, even tho it has been dormant because i have a similar problem. We made a large (printing volume: 40cmx40cmx20cm) 3D printer from parts available in our firm and we atmega 2560 and RUMPS 1.4 for running it. It's a classical configuration so Z axis is run by one A4988 driver in full div32 mode running two nema 17 motors with lead screws. This part is running fine.

X and Y axes are run trough three separate drivers in half step mode (it's this or full step) and three unipolar (5 wire) motors (two on Y and one on X) getting pulses from RUMPS. they operate trough belt drive.

Up until now we only had one " successful " print, and that was a 1cm x 1cm x 1cm cube that was just barely off, but as soon as we try printing anything larger X and Y axes start running slowly.

As it is now X and Y axes have about 167 steps per mm while Z has 300, and we reduced the speed to ridiculously low but it's not helping. I can send motor/driver details tomorrow but the first question is could this setup even work?

If Yes, how, what are we missing, do i need to try and reprogram motor drivers (they have JTAG) or should we just give up, buy and run nema 17 and A4988 here as well?

I thank anyone for any support on this.

Memphis_Raines:
X and Y axes are run trough three separate drivers in half step mode (it's this or full step) and three unipolar (5 wire) motors (two on Y and one on X) getting pulses from RUMPS. they operate trough belt drive.

Why?

You say the Z axis is working fine so why not use the same technology for the other axes?

And how can you run two axes through 3 drivers?

Have you checked that each axis works separately with the required precision?

What 3D printer software are you using?

It seems to me you have built a complex system that will require someone local with good debugging skills to figure out. Is is a mechanical problem? Is it an electrical problem? Is it a programming problem?

...R

Thank You for the quick reply.

5 wire motors and drivers were already bought for another project when we started developing the printer. and three drivers are one for each motor, but the two running Y axes are run from the same pin of the RUMPS.

It is running Marlin - reperier - slic3r

Mechanically it behaves fine when You are doing setting up. After zeroing You give it a set distance and it moves wery precisely. But when You print, after a few layers it slowly starts drifting one way miwing each next layer a bit to one side. Oower wise we tesed out power supplies and they wewe fine.

As for the "why don't we just buy nema 17 motors" well, it requires teling my boss that I failed with this and that I can not make it work.

So are there some EEPROM options or something in the code itself that we need to chhange to adapt to unipolar motors?

Memphis_Raines:
Mechanically it behaves fine when You are doing setting up. After zeroing You give it a set distance and it moves wery precisely. But when You print, after a few layers it slowly starts drifting one way miwing each next layer a bit to one side.

Two things spring to mind …

You have not told us exactly what unipolar motors you are using (please post a link to the datasheet) and I wonder if they have a gearbox which probably would have considerable backlash.

Maybe you need to calibrate the axes to get the ideal number of steps per millimetre. You should be able to test that by printing a square an then measuring it accurately. Make sure to mark the sides before you remove it from the print bed so you know which is which.

…R

So, this is the basic data on the motors and drivers:




rest of the files


So, I have checked for precision and backlash, and a 10 X 10 X 10 mm cube that we successfully printed ended up being 10,04 X 10,04 X 10,03 mm. But when we tried printing a 20 X 20 X 20 mm cube problem in the picture below keeps happening.


Please display your image(s) in your post so we can see it(them) without downloading it(them). See this Simple Image Guide

...R

Thank You for the advice, I've edited the posts

The images are hard to read. Can you just post a link to the website where they come from?

Rather than printing a cube can you try just printing an empty square (walls only) that is just 1 mm high. Then the print head movements should be very simple so that you can see if it is seriously failing to move the appropriate distance.

It will also be a lot faster and use less plastic.

If you try that and it does not work then tell us if there are any strange noises or other symptoms that might help with diagnosis.

Have you made any changes to the standard configuration of Marlin?

This is a very wild guess ... you may get the sort of behaviour you are seeing if you were trying to put a very large number of steps (i.e. more than 32767) into an int variable. But I really can't see how that might happen if you are using off-the-shelf software.

I think if this was my machine I would write some very simple code using the AccelStepper library to make one axis at a time move back and forth by a specified amount without any extrusion.

...R

Here, I have added part of the PDF that are about these motors and drivers

SANMOTION - 5-phase stepping systems ABV.pdf (383 KB)

I did not realize it is a 5 phase motor. I don't know anything about them. The coil wiring seems to be very different from the bipolar motors for which the A4988 is intended.

If the proper stepper driver for those motors takes step and direction signals then you could probably connect it in place of the A4988 with some external wiring.

...R