Fine focusing a microscope using StackShot

[rjlittlefield]

I was inspired by Rik's original post to have this clamp made for me by SRB-Griturn in the UK.

Excellent! I am curious about the price of custom machining these days. How much did this clamp cost you, and what was the procedure for ordering it?

--Rik

[Greenfields]

Its a simple and efficient process.

1 - email them a drawing with any instructions [e.g whether you want it anodized black]

2 - They email back with any questions or an order number, a price with shipping and an indication of the time it will take.

3 - Click on this page

http://srb-photographic.co.uk/specialis ... 1065-c.asp

4 - select the price they have quoted and put in the order number then pay.

5 - They make the item and send it to you.

The price for this item was £35.00

Henry

[ChrisR]

I think you'd want soft shoes on those screw-ends?

--

I must admit I wonder how important precision and rigidity actually ARE in this application. Rik's shoved-on tube would mould itself to the ridges on the fine focus knob, and it could be held with a sprung hose-clip.
A single small screw through the tubing, into a wooden dowel, would stop that slipping.
So what would be the result?
If the push-on were not tight, there could be some backlash at the end of the step as the torque relaxed. With ridges and a spring clip, I don't think it could keep slipping. I can't see a real problem.
With a single screw fixing into the dowel, there could be some eccentricity in the drive resulting in a per-revolution variation in the step size according to knob position. Well OK, so..?
I have trouble seeing how it could be anywhere near, say, 10%. Much below that, say 3%, and it's insignificant. Am I missing something?

[Greenfields]

Chris,

The screws are nylon. They don't need to grip tightly because the fine adjustment knob rotates with very little torque.

I am sure that there are lots of alternative ways of doing this.

I did try to copy Rik's idea but could not find a piece of tube to fit and don't have acccess to a lathe. I even thought of getting a small metalworking lathe just for the fun of it but sadly reason and my impatience won on that occasion: the cost of getting the clamp made for £35 made this the easiest option for me.

I asked for the screws to see if they worked. They do.

However, even without the screws the cylinder could be lined with tape for a push-fit onto the knob.

I wanted something which was durable and could be attached and removed quickly and reliably. The motor bracket is mounted onto an Arca tripod plate on an Arca clamp so everything can be cleared away or replaced quickly.

Henry

[ChrisR]

Thanks for the figure, Henry.
That's very good so my question is moot.

(Hmm, I wonder if they do flat adapter plates etc, I'll ask them.)

[rjlittlefield]

A few weeks ago, concern was expressed about lateral stress on the fine focus mechanism, caused by misalignment of the stepper motor and the focus knob.

I hope I clarified at that time that the two tapered cones collaborate with the vinyl tubing to make what is essentially a double-joint universal coupler that applies very little sideways pressure.

However, in related postings,

Greenfields showed us his bespoke adapter using a flexible link similar to http://www.ebay.com/itm/300752549179 (CNC Flexible Plum Coupling Shaft Coupler), and

dolmadis asked specifically about a flexible coupler like this:
http://www.ebay.co.uk/itm/5-6-6-35-8mm- ... 0631482123 .

At that time, I commented that

Sorry, I can't say. I have no hands-on experience with those couplers, and my imagination permits several very different results.

I have since rectified that lack of experience. I ordered samples of both those couplers. Here is what they look like:



The answer to dolmadis's question is that in my opinion these couplers are completely not suited to this application and would place the focus block at significant risk.

The problem is that these couplers are designed to transfer large amounts of torque while compensating for only small amounts of misalignment, both typical of what you might need in a commercial precision machine tool setup.

In contrast, the microscope focus application requires only tiny amounts of torque, while any DIY setup is liable to have significant amounts of misalignment. I have deliberately misaligned my setup here, so you'll get the idea:



I haven't figured out how to make a mechanical measurement, but based on their feel in my hands, I estimate the commercial CNC couplers as being over 100 times stiffer than the coupling with tapered shafts and vinyl tubing. No, that's not a typo and it's not an exaggeration -- over 100 times. These things are stiff. The helical coupler is relatively compliant for angular and longitudinal misalignment, but it's much stiffer for lateral alignment. This is because the web of the helix is over 3 times thicker laterally than it is longitudinally -- 2 mm pitch with about 1.5 mm thickness longitudinally, versus 5.3 mm wall thickness laterally.

I'm sure those commercial couplers are great for their intended purpose (CNC machining), but there's no way that I would rely on one of them to protect my focus block.

--Rik

[ChrisR]

I've always assumed that the motor in these applications would have to be on quite soft rubber mounts. ("Soft" == elastic with low modulus.)

Something like this would allow a lateral offset but demand axial movement, as the length of the coupler varies with the angle which misalignment could change with rotation.

My feeling is that a simple, radial peg-in-slot drive would be adequate. There's no real need to eliminate play/backlash as the rotation can be always in one direction. Misalignments would lead to eccentricity/ovality and ultimately in a variation in the focus step size, but if the variation is even as much as a few percent, I really don't think it matters much. The focus-step overlap isn't precisely calculated figure.

A toothed rubber belt drive is quite a nice way of accommodating misalignments, but it does present a consistent side-load on the focus knob.

[rjlittlefield]

My feeling is that a simple, radial peg-in-slot drive would be adequate.

I agree. Haven't tried it yet, but my imagination says this would work fine. Clamp the tubing to the focus knob, cut a slot in the tubing, mate that with a pin sticking out sideways from the motor shaft.

The only downside is that this would guarantee no slipping, which sounds like a good thing at first thought but might be annoying in practice because it prevents tweaking fine focus by turning the knob.

--Rik

[ChrisR]

Sprung tube clip ?
Or -
just turn the motor with the power off?

[Chris S.]

A toothed rubber belt drive is quite a nice way of accommodating misalignments, but it does present a consistent side-load on the focus knob.

I've used this approach (as shown here) quite happily for the past four years. The potential for side-load issues concerned me at first, but has turned out to be no problem in practice. Given this experience, I would submit that with a properly-adjusted timing-belt setup, there is little or no side load. The timing belt in my rig is not under tension; I set it just tight enough to engage the teeth on the timing pulleys, but not so tight as to displace these pulleys even slightly. My timing belt isn’t made of rubber, and isn’t stretchy. An important construction detail, I think, is to include an adjustment screw that lets you adjust the distance between the timing pulley on the motor and the timing pulley on the fine focus shaft. Then start with the belt loose, and increase the distance between the pulleys until the belt engages reliably, but doesn’t pull on the pulleys or the shafts to which they are connected. When this distance is established, you can use other screws to lock the assembly in place. While this may sound difficult at first, it's actually rather simple.

The only downside is that (a peg and slot drive) would guarantee no slipping, which sounds like a good thing at first thought but might be annoying in practice because it prevents tweaking fine focus by turning the knob.

My timing-belt drive also has this theoretical downside, but I find it not very burdensome in practice. When I first set up a shot, my StackShot controller is usually turned off, so the motor is also, automatically, powered off. In this scenario, turning the fine focus knob—thus freewheeling the powered-down motor, which has almost no resistance--works fine. When the Stackshot controller is powered up, I do avoid turning the fine focus knob—but find it very easy to tweak fine focus with the forward and backward buttons on the StackShot controller. If, on one's rig, this action isn’t smooth at first, one can adjust the StackShot controller’s "tramp" parameter (controls the motor’s ramp-up time from stopped to full speed). Adjusting the torque and speed of the motor (easy with a StackShot controller) may also help dial in this capability. On my rig, tweaking fine focus with the StackShot controller is no harder—and often easier—than doing so with the fine focus knob on the microscope block.

--Chris

[wpl]

I'm sure those commercial couplers are great for their intended purpose (CNC machining), but there's no way that I would rely on one of them to protect my focus block.

I am using one of those spiral-cut-ham couplers, as I call them. My microscope is a Leitz Ortholux (original version) and it appears very rugged. I never thought there was any reason to "protect" my focus block.

By the way, I enjoy seeing how other people are solving the coupling-to-the-focus-block problem.

[johan]

I'm going to come up against this challenge in the very near future, mating Stackshot and a newly-bought Labophot. Is the Rik method here still the currently recommended method?

Thanks

[ChrisR]

Johan I think your 'scope is a Labophot 2 -?- , which has slightly different focus, which may be pertinent to details.
(100µm per rev)

[johan]

yep, thanks, Labophot 2, 100um per revolution

[rjlittlefield]

I'm going to come up against this challenge in the very near future, mating Stackshot and a newly-bought Labophot. Is the Rik method here still the currently recommended method?

My current rig uses a focus block that was sawed from the arm of a microscope, with a stepper motor mount glued to the body of the block and timing belt drive to a gear that replaces one of the fine focus knobs. See details HERE.

However, if you don't want to damage or modify the microscope, then simply coupling to the regular fine focus knob is fine. This is an established method, used even by some commercial add-ons, for example HERE.

--Rik