DIY Inkjet Servo Macro Rail

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

nucleobyte
Posts: 51
Joined: Sun Aug 22, 2010 8:00 pm
Location: Virginia, USA

DIY Inkjet Servo Macro Rail

Post by nucleobyte »

I've been experimenting with a homemade stacking rail using parts from a discarded InkJet printer. These cheap printers use servo motors with encoders rather than stepping motors, so I wasn't sure what kind of precision could be achieved. At any rate, I got it working, and the results were acceptable, so I though some folks might be interested. I have not yet implemented any kind of PID control, which should be possible with the encoder feedback and a better motor controller. Here is a link to a (slightly shaky phone camera) video of the device in operation: https://www.youtube.com/watch?v=wFLTmwMCWn4

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Very nice!

The last couple of seconds show an interesting graph, but I can't figure out what step size it corresponds to.

What is the minimum step size that you consider acceptably repeatable?

--Rik

Saul
Posts: 1781
Joined: Mon Jan 31, 2011 11:59 am
Location: Naperville, IL USA
Contact:

Post by Saul »

Great & effective design ! I like these couplers from the tubing - in most cases they work better than machined from aluminum
Saul
μ-stuff

nucleobyte
Posts: 51
Joined: Sun Aug 22, 2010 8:00 pm
Location: Virginia, USA

Post by nucleobyte »

Rik,

Here is the graph:

Image

This was from a test run - I wanted to establish a baseline to compare with possible improvements. The run was 1000 sequential steps resulting from activating the motor for 7 ms with a 12 V motor supply. This resulted in movement with each step measured by the number of encoder counts.

To calculate the step size, one rotation on a 16 tpi thread would be 1.5875 mm. The encoder is counting 4800 counts per revolution. Estimating 140 counts from the graph, 140/4800 * 1.5875 = 46.3 um per step (high of about 170 count/56 um and low of about 100 count/33 um.)

The question then becomes, how precise (or consistent) is enough? You could probably answer that better than me! I'm not sure at what point variation in the step size causes problems. Turning down the motor-on duration shortens the step but increases the variation. Inverting the logic, I can tell the motor to move and stop after, for example, one encoder count -- but this results in 10 counts or more due to inertia. Motor voltage has a large effect on starting torque and effects this as well.

I guessing, but with a better motor controller and possibly a PID system, it might be possible to get about 20 um consistently enough for stacking, I think that would be good enough for about 5X magnification.

- Greg

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Greg,

Nice work!! Cool idea using an ink jet printer carriage.

Since you have the drive screw coupled with a plastic tube coupler, why not use a 200 or 400 step stepper motor instead of the DC motor and encoder? Steppers have a pretty good reputation working as the drive force for focus rails and the error isn't accumulative. I believe typical stepper motors (NEMA 17) have less than 10% step errors and the real error source ends up being the backlash from the thread/nut interface, and the screw thread error which I would assume is very small and repeatable. There are all sorts of programmable controllers for steppers available for DIU efforts too.

Best,

Mike

nucleobyte
Posts: 51
Joined: Sun Aug 22, 2010 8:00 pm
Location: Virginia, USA

Post by nucleobyte »

Thanks Mike,

I should have mentioned in the OP, one of my goals for this project was to see how much of the inkjet printer I could reuse. I was also simply curious as to how accurate it could be. Clearly, the printer makers in this case hp, get very high precision -- this drive train feeds the paper line by line for image priniting without banding. The printer contains a custom motor driver chip which I could not find documentation for, probably secret. They may also may be using the sine-wave encoder fully, I am converting it to square for much simpler handling.

The inkjet carriage I thought was a cool idea, but is not really a success, it is stable only in the hoizontal position. In the printer it is intended only for a very light load and weakly attached to the 8mm rod through a plastic bushing. I plan to switch to 8mm brass bushings on 2 rods. These bushing are available cheaply through the usual online markets.

- Greg

Post Reply Previous topicNext topic