clutch with friction disks to decouple stepper motor

iconoclastica · Post by **iconoclastica** » Sun Mar 24, 2019 3:02 pm

In an earlier discussion with Adalbert and Elf system the topic of decoupling the stepper motor mechanically briefly came up. If the motor turns easily when switched off completely there's no problem of course. But not all motors turn that lightly that I would want to have them connected to the fine focus mechanism. When finally my motors arrived it even appeared that they wouldn't turn at all hand-driven, for they are geared and require a lot of force to turn them against the gear (possibly exerting too much power on those fine mechansms),

So I have been thinking to design a clutch, so that the fine focus would turn freely when decoupled from the motor. At first it threatened to become rather complex 3-D printing, but gradually the ideas simplified into something very easy to make. In fact, while the 3-d printer and the laser-cutter are temporarily down for maintenalce I made it in a two hours with simple tools and scrap parts.

The hardest part is interfacing with the fine focus knob without making irreversable changes. Best would be to remove the knob and connect to its shaft, but I couldn't get it off.
My fine focus knob (Nikon Optiphot) is tapering, but it has a cylindrical base which allows for a collar. The collar has got to fit well. Mine needed to have the bore diameter 30mm, and by coincidence (if you may call it that with someone who rarely throws anything away

) I still had a failed part of an earlier condenser adapter lying that fitted perfectly. I only had to add a screwhole to fix it tightly. It is the red ring in the schematic drawing below.

The next part is a spacer to advance beyond the tapering part of the knob. For this I used the wheel that once pulled the paper through a NEC P6 matrix printer, taken apart long ago. The cap of a shampoo bottle might be just as fine. I drilled a hole in the centre and passed through it a bolt that is the core of all parts following. I glued the bolt with epoxy raisin to the (green) knob. The purple ring at its base is acrylic (plexiglass) to provide better holdfast to the glue between the collar and the spacer-knob.

Next two more acrylic disks (black) with adhesive neoprene (blue) on their facing surfaces. Sandpaper will work too. The outer disk is glued to the axis - an aluminium tube - that also carries the pulley. A soft spring (orange) pushes the disks apart unless the hexbolt that finishes the device is turned in (2-3 turns).

This is how it looks in reality (motor not in place yet):

Wim

Chris S. · Post by **Chris S.** » Sun Mar 24, 2019 10:49 pm

Wim,

Thanks for posting your clutch. If my experience is any guide (and you alluded to this), your first concepts were likely more complicated, and it took quite a bit of thought to simplify to the workable, elegant clutch you've shown here.

Nice addition to the art!

--Chris S.

Pau · Post by **Pau** » Mon Mar 25, 2019 1:57 am

Wim, this is a very nice project!

When finally my motors arrived it even appeared that they wouldn't turn at all hand-driven, for they are geared and require a lot of force to turn them against the gear (possibly exerting too much power on those fine mechansms)

Do you really need a geared motor to move a microscope fine focus knob?

Lacking your engineering skills, when I faced to a similar (not identical) problem I designed a very simple magnetic coupling and I can't be more satisfied:
http://www.photomacrography.net/forum/v ... 122#232122

iconoclastica · Post by **iconoclastica** » Mon Mar 25, 2019 5:59 am

Pau,

I am finding new applications for magnets every day, but this one I hadnot thought of yet. I certainly would have experimented with them if I had. With a sheet of rubber in between they should offer more than enough grip indeed. However, wouldn't the decoupling require too much force, or force perpendicular to the axis? I am thinking about electro-magnets now. Then the software could adhere and release to the axis...

re geared stepper motors: I am not sure. The flattest DoF I am working with is 0.19µm whereas one full turn of the knob does 100µm. So even a 400-steps per turn motor would be less than minimal. The 1:50 gears on the other hand perhaps is the motor I could find rather than the one most suitable. If I am interpreting its documention correctly, it will do 2400 steps per turn or 20 steps per micrometer.
There is, of course, some play in the gears. But with the Stackshot I have grown used to work unidirectional and overshooting the starting point to do so. I expect no issue there, but I am not ready yet to test it.
Now I have the driver board, I think I could have opted for microstepping. In December, when I started planning this project, I still felt too unsure about that. But it won't be a radical operation to change to microstepping and another motor in a later stage.

Saul · Post by **Saul** » Mon Mar 25, 2019 7:55 am

Used different stuff for this purpose - starting with flexible tubing, cap type connector, aluminum coupler, magnets ... Most recent ( on the BHMJ) - is with velcro - saved me space and it is more forgettable/fool-proof in the case of misalignment and easy removable (if needed).

iconoclastica · Post by **iconoclastica** » Mon Mar 25, 2019 8:47 am

Saul wrote:Used different stuff for this purpose - starting with flexible tubing, cap type connector, aluminum coupler, magnets ... Most recent ( on the BHMJ) - is with velcro - saved me space and it is more forgettable/fool-proof in the case of misalignment and easy removable (if needed).

How do you retract the motor when you want to go back to manual? From the photos it seems fixed into its position?

Saul · Post by **Saul** » Mon Mar 25, 2019 9:09 am

iconoclastica wrote:
Saul wrote:Used different stuff for this purpose - starting with flexible tubing, cap type connector, aluminum coupler, magnets ... Most recent ( on the BHMJ) - is with velcro - saved me space and it is more forgettable/fool-proof in the case of misalignment and easy removable (if needed).
How do you retract the motor when you want to go back to manual? From the photos it seems fixed into its position?

Go to manual - to adjust the focus ? Or something different ?
For focusing no need to retract the stepper - I'm using coarse only, it works. If you want to use fine knob, when stepper is on - just put 4 pole switch between the cable and stepper in order to disconnect coils.

mawyatt · Post by **mawyatt** » Mon Mar 25, 2019 9:44 am

Saul wrote:
iconoclastica wrote:
Saul wrote:Used different stuff for this purpose - starting with flexible tubing, cap type connector, aluminum coupler, magnets ... Most recent ( on the BHMJ) - is with velcro - saved me space and it is more forgettable/fool-proof in the case of misalignment and easy removable (if needed).
How do you retract the motor when you want to go back to manual? From the photos it seems fixed into its position?
Go to manual - to adjust the focus ? Or something different ?
For focusing no need to retract the stepper - I'm using coarse only, it works. If you want to use fine knob, when stepper is on - just put 4 pole switch between the cable and stepper in order to disconnect coils.

Please use caution when disconnecting a stepper when it's active (current thru coils). This can damage or destroy the controller because of the inductive "kick" caused by the disconnect.

I think most controllers have the ability to "turn off" the coil current, or if you can, remove the motor power before disconnecting. Either way this is the safe way to disconnect the motor.

Best,

Pau · Post by **Pau** » Mon Mar 25, 2019 9:51 am

iconoclastica wrote:..With a sheet of rubber in between they should offer more than enough grip indeed. However, wouldn't the decoupling require too much force, or force perpendicular to the axis? I am thinking about electro-magnets now. Then the software could adhere and release to the axis..

Initially I used a thin sheet of silicone or rubber between the magnets but now I find them unnecessary.
If the microscope focus is smooth like in my cases the magnets have enough lateral torque to drive them without issues and at the same time they can easily slip if the focus reaches its limit, if you want to manually focus while the motor is under current and to dismount them laterally sliding them. All this is very convenient in my case because I can use the same setup with four different microscope blocks and only one motor.

iconoclastica wrote:re geared stepper motors: I am not sure. The flattest DoF I am working with is 0.19µm whereas one full turn of the knob does 100µm. So even a 400-steps per turn motor would be less than minimal. The 1:50 gears on the other hand perhaps is the motor I could find rather than the one most suitable. If I am interpreting its documention correctly, it will do 2400 steps per turn or 20 steps per micrometer.
There is, of course, some play in the gears. But with the Stackshot I have grown used to work unidirectional and overshooting the starting point to do so. I expect no issue there, but I am not ready yet to test it.
Now I have the driver board, I think I could have opted for microstepping. In December, when I started planning this project, I still felt too unsure about that. But it won't be a radical operation to change to microstepping and another motor in a later stage.

The microscope focus block is actually an excellent gear box.
With microstepping I have more than enough precision to work with NA 1.4 objectives with a 100micron/turn block. The not-too-good Wemacro system can do 0.05micron slices (around 0.008micron if the controller allowed for single microsteps) with that microscope.

Saul · Post by **Saul** » Mon Mar 25, 2019 9:55 am

mawyatt wrote: Please use caution when disconnecting a stepper when it's active (current thru coils). This can damage or destroy the controller because of the inductive "kick" caused by the disconnect.

I think most controllers have the ability to "turn off" the coil current, or if you can, remove the motor power before disconnecting. Either way this is the safe way to disconnect the motor.

Thanks, Mike, fully agree, why I'm using coarse focus (never had a situation when fine was needed during the stacking). To set the start point (Wemacro) enough coarse focus.

remove the motor power before disconnecting

- in the Wemacro case power must be turned off, I'm losing time waiting for the controller to connect back through the bluetooth. Unless I misunderstood something ...

mawyatt · Post by **mawyatt** » Mon Mar 25, 2019 10:46 am

Saul wrote:
mawyatt wrote: Please use caution when disconnecting a stepper when it's active (current thru coils). This can damage or destroy the controller because of the inductive "kick" caused by the disconnect.

I think most controllers have the ability to "turn off" the coil current, or if you can, remove the motor power before disconnecting. Either way this is the safe way to disconnect the motor.
Thanks, Mike, fully agree, why I'm using coarse focus (never had a situation when fine was needed during the stacking). To set the start point (Wemacro) enough coarse focus.

remove the motor power before disconnecting
- in the Wemacro case power must be turned off, I'm losing time waiting for the controller to connect back through the bluetooth. Unless I misunderstood something ...

The Wemacro is a really nice controller but doesn't have a simple means to disconnect the motor power as you mention. If you are comfortable with modifying electronics then an option would be to to disconnect the motor power going to the driver package inside the case on the PCB with a switch. Be sure to do this so that capacitance is available (I would recommend 100uF) on the driver side of the switch so when the switch is disconnected the motor current has some capacitance to flow thru back to ground, this prevents a large voltage spike caused by the motor coil inductive current. The driver package pin is easily accessible, so shouldn't be too hard to do. Of course this will void the Wemacro controller warranty!!

Best,

Saul · Post by **Saul** » Mon Mar 25, 2019 11:15 am

mawyatt wrote:
Saul wrote:
mawyatt wrote: Please use caution when disconnecting a stepper when it's active (current thru coils). This can damage or destroy the controller because of the inductive "kick" caused by the disconnect.

I think most controllers have the ability to "turn off" the coil current, or if you can, remove the motor power before disconnecting. Either way this is the safe way to disconnect the motor.
Thanks, Mike, fully agree, why I'm using coarse focus (never had a situation when fine was needed during the stacking). To set the start point (Wemacro) enough coarse focus.

remove the motor power before disconnecting
- in the Wemacro case power must be turned off, I'm losing time waiting for the controller to connect back through the bluetooth. Unless I misunderstood something ...
The Wemacro is a really nice controller but doesn't have a simple means to disconnect the motor power as you mention. If you are comfortable with modifying electronics then an option would be to to disconnect the motor power going to the driver package inside the case on the PCB with a switch. Be sure to do this so that capacitance is available (I would recommend 100uF) on the driver side of the switch so when the switch is disconnected the motor current has some capacitance to flow thru back to ground, this prevents a large voltage spike caused by the motor coil inductive current. The driver package pin is easily accessible, so shouldn't be too hard to do. Of course this will void the Wemacro controller warranty!!

Best,

Thanks Mike, good information. I'm not going to change anything, I'm pretty happy with existing setup

iconoclastica · Post by **iconoclastica** » Mon Mar 25, 2019 12:38 pm

Initially I used a thin sheet of silicone or rubber between the magnets but now I find them unnecessary.

A quick test here showed that my magnets can't do it and slip. I suppose it all depends on the microscope and the size and thickness of the magnets. Rings may do better than disks.

at the same time they can easily slip if the focus reaches its limit,

That is a clever built in safety. It should not be necessary though (in my workflow, that is) for I always will scroll through the stack manually before engaging the motor.

I also tried a number of stepper motors to turn by hand. None of them powered on, but all of them I found inconvenient to have as extra friction on the FF knob. They don't turn smoothly but have a kind of gritty feeling to them. They all were also rather old motors. Are modern steppers different?

Perhaps I should stress that I use the microscope more often for observation than for photography. When observing, I am using the FF continously to get a sense of the 3rd dimension. It must move smoothly. I think some of us are using it predominantly for making photos and then it makes more sense to have the motor connected semi-permanently. So I want to be able to disengage the motor in one quick move.

Pau · Post by **Pau** » Tue Mar 26, 2019 5:55 am

iconoclastica wrote: A quick test here showed that my magnets can't do it and slip. I suppose it all depends on the microscope and the size and thickness of the magnets.

Sure, I bought several kind of magnets to find the right combo...they are inexpensive

Rings may do better than disks.

I've find not much difference. Rings allow to put a plastic centering axis, not needed in many cases as magnets of the same diameter are self-centering

It should not be necessary though (in my workflow, that is) for I always will scroll through the stack manually before engaging the motor.

I also tried a number of stepper motors to turn by hand. None of them powered on, but all of them I found inconvenient to have as extra friction on the FF knob. They don't turn smoothly but have a kind of gritty feeling to them.

I've found a sweet spot: I can use manually the fine focus drive while the motor is engaged: If it is on power the magnets slip and if it is off power the motor turns, the movement is less smooth but still convenient for precision manual focus

Perhaps I should stress that I use the microscope more often for observation than for photography. When observing, I am using the FF continously to get a sense of the 3rd dimension. It must move smoothly...
...So I want to be able to disengage the motor in one quick move.

When not doing photography I also disengage the motor. Engaging it again just takes few seconds, starting the electronics and entering the parameters in software takes much more time.