Hello.
I am currently using the delta stage in a setup which involves scanning the sample. However due to the very small step sizes, the 28BYJ-48 motors struggle and some images have lateral movement due to backlash etc from the 3D printed gears. To address this issue I plan to implement metal gears. However it is also an option to increase the gear ratio which would in turn make the small step sizes more manageable for the motors. Apart from adapting the OpenSCAD stage and modifying the sofware stack to adjust the movement of the motors, are there any other issues that may arise that I will have to address? Or if there are any alternative suggestions I’d appreciate to hear. Thanks
You can set the backlash compensation in the Openflexure software (v2.11), which should help. However I don’t exactly know how effective that is on the Delta Stage. On the cartesian stage of the OpenFlexure Microscope only one motor moves for each axis, so backlash is simple. On the Delta stage all motors move for each axis of motion. What aspect of the Delta stage is necessary for your application?
Metal gears are not likely to help you much unfortunately. There is some backlash in the 2:1 plastic gears, but there is also a lot of backlash in the internal 64:1 gearing inside the stepoer motors.
On the Microscope, the printed gears have been changed a bit to fit more tightly on the motor and on the 25mm hex screw in the actuators. Using those latest gears from the latest Microscope release (beta5) might help you a little.
When you say increasing the gearing, do you mean going from 2:1 to 1:1 (fewer steps per stage mm) or to 3:1 (more steps per stage mm)? 1:1 and 0.8:1 gear sets are available from the Microscope instructions under Customisations. I have not tried whether either of those fit on the Delta stage, I think making the gear on the motor larger will make it hit the stage legs. A ratio of 3:1 will not be possible, the small gear on the motor would be too small to fit the locking screws, and the large gear on the actuator would be too big to fit between the motor mounting screws.
In response to the first question - the delta stage is neccessary due to its fixed optics module, as we are splitting part of the light out and directing it to a secondary camera.
We would be looking to increase the steps per stage mm to make the very small stage translation steps we are using more manageable for the stepper motors – this would involve modifying the stage which would be challenging. I’ll have a look at the backlash compensation on the software and try the updated gears.
Thanks for the detailed response.
1 Like
The alternative gear ratios adjust further in the opposite direction from what you want. I don’t think that it will be possible to get more steps per micron with the type of gear that we are using. There is a mimimum size/number of teeth that makes a sensible motor gear. There is also a maximum size of large gear that will fit between the motor lugs.
Your suggestion of metal gears might help, if you can get a gear that fits onto the motor shaft that has a much smaller diameter than the current small_gear.stl. Then you find a matching gear with a diameter similar to the current large_gear.stl and the gearing ratio will be bigger. You will have two remaining issues: you need to attach that large metal gear to the actuator bolt, and the sum of the diameters of the two gears will be less than the OpenFlexure design, so the motor mounting lugs will need to move.
The OpenFlexure Block stage will also move a sample in xyz. It has a smaller lever ratio and so smaller step size. The motion is characterised in the paoer about that stage. You will see the minimum incremental motion, which may suit your application.
With some research I decided to abandon the metal gears as I struggled to find 2 that were compatible - with suitable number of teeth and gear ratio, bore diameters that are required, pitch diameters that would work in the setup and matching modules. However, I have managed to increase to 30/12 for a 2.5 gear ratio which doesn’t require modifying the stage. This setup is working smoothly so far although I am yet to test it’s effect on vertical drift due to backlash.
I don’t think the block stage would be applicable for my setup.
Thank you for the assistance.
1 Like