Microscope headpiece dovetail dimensions

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

OK, here are my crosscheck results. There's good news and bad news.

The good news comes if I work backwards from Charlie's actual measurements of 39.57 mm diameter at 3.20 mm from the seating face. In that case I calculate that at 2.0 mm from the seating face, the diameter would be 39.57 - 2*(3.2-2.0)*cos(60 degrees) = 38.37 mm. That's comfortingly close to the values I calculated based on my tabs.

The bad news is that I cannot match Charlie's numbers for either diameter shown on the diagram. At the top, I calculate that the diameter should be 39.57 - 2*3.2*cos(60 degrees) = 36.37, not 35.88. At the bottom, I calculate 39.57 + 2*(6-3.2)*cos(60 degrees) = 42.37, not 42.8. Working backwards from Charlie's numbers, I get that the angle would be 54.78 degrees, not 60.

Can somebody else chime in here and resolve the discrepancy? Where have I messed up?

--Rik

Rylee Isitt
Posts: 476
Joined: Fri Apr 13, 2012 3:54 pm
Location: Canada
Contact:

Post by Rylee Isitt »

Ah! I didn't see your post!

I decided to measure (on my diagram) to see if my simulation agreed more or less..

Upper diameter of dovetail: 35.92mm
Lower diameter of dovetail (assuming that chamfer is removed and angle is continued): 42.78mm

Very, very close...

I think we're ready to draw up our tech diagram and decide on what our tolerances should be!
Charles Krebs wrote:This should be extremely close. I opened a brand new Olympus U-SPT tube and made some careful measurements. The outside diameter was exactly 42.0 mm (but as Rik mentioned the edge is slightly chamfered/rounded). The diagram, showing 42.8mm, indicates what the measurement would be if the edges were left "sharp" and the 60 degree cut was maintained sharply up to the seating face (not shown). The angle looks to be 60 degrees.

The depth of the dovetail was exactly 6.0 mm

(At a distince of 3.20mm in from the 35.88mm side, the diameter was 39.57mm... green line in diagram)

Hopefully all the old trigonometry neurons woke up and fired properly! :wink:

Image

Rylee Isitt
Posts: 476
Joined: Fri Apr 13, 2012 3:54 pm
Location: Canada
Contact:

Post by Rylee Isitt »

To recalc the angle based on Charles' measures, get the difference in width between top and bottom, but divide by 2 (you want to form an imaginary triangle on one side, not both).

( 42.8 - 35.88 ) / 2 = 3.46mm (the base of the triangle)

The height of the triangle is the height of the dovetail, so 6mm.

Tan T = ( 6/3.46 ) => T = tan-1( 6/3.46 ) = 60.03 degrees.

Seems to work for me!
rjlittlefield wrote:OK, here are my crosscheck results. There's good news and bad news.

The good news comes if I work backwards from Charlie's actual measurements of 39.57 mm diameter at 3.20 mm from the seating face. In that case I calculate that at 2.0 mm from the seating face, the diameter would be 39.57 - 2*(3.2-2.0)*cos(60 degrees) = 38.37 mm. That's comfortingly close to the values I calculated based on my tabs.

The bad news is that I cannot match Charlie's numbers for either diameter shown on the diagram. At the top, I calculate that the diameter should be 39.57 - 2*3.2*cos(60 degrees) = 36.37, not 35.88. At the bottom, I calculate 39.57 + 2*(6-3.2)*cos(60 degrees) = 42.37, not 42.8. Working backwards from Charlie's numbers, I get that the angle would be 54.78 degrees, not 60.

Can somebody else chime in here and resolve the discrepancy? Where have I messed up?

--Rik

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

rjlittlefield wrote:Can somebody else chime in here and resolve the discrepancy? Where have I messed up?
Never mind, it's my idiocy -- I was using the 6.0 as hypotenuse instead of one side of the triangle. It's the old cosine() versus secant() confusion.

Charlie's numbers are right. Time for bed...

--Rik

Rylee Isitt
Posts: 476
Joined: Fri Apr 13, 2012 3:54 pm
Location: Canada
Contact:

Post by Rylee Isitt »

I slept through most of trig in high school. I regret that now, beyond words, and every single time I have to do trig I end up relying on that "soh cah toa" mnemonic.

And don't ask me about trig identities. I don't know them.

So don't feel bad - you're probably a lot better at it than me ;)
rjlittlefield wrote:
rjlittlefield wrote:Can somebody else chime in here and resolve the discrepancy? Where have I messed up?
Never mind, it's my idiocy -- I was using the 6.0 as hypotenuse instead of one side of the triangle. It's the old cosine() versus secant() confusion.

Charlie's numbers are right. Time for bed...

--Rik

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Rylee Isitt wrote:And don't ask me about trig identities. I don't know them.
I'm willing to remember that cos^2+sin^2=1.

Everything else I look up.

God may know why we torture our math students with remembering double- and half-angle formulas. I certainly don't.

Mine was just a straightforward mistake from working too fast while I'm tired. Mistakes happen. That's why I cross-check everything. (Including terminology. Should have written cotangent, not secant. Definitely time for bed...)

--Rik

Rylee Isitt
Posts: 476
Joined: Fri Apr 13, 2012 3:54 pm
Location: Canada
Contact:

Post by Rylee Isitt »

rjlittlefield wrote:Definitely time for bed...
Oh, what a distant dream this seems. My bed is covered in microscope parts. It's going to take half an hour to put them back where they belong!

Charles Krebs
Posts: 5865
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

Here's how I figured it...

I measured the diameter across at 3.2 mm depth as 39.57 mm. I calculated d1 and subtracted twice that value from 39.57 mm to arrive at 35.88mm.

Knowing the base diameter (35.88mm), the angles, and the 6mm total thickness, the larger diameter works out to 42.8mm.
tan(30) = d2/6
d2= 3.464 mm
35.88 + (2)(3.464) = 42.80mm

Image

Charles Krebs
Posts: 5865
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

dang it Rik.... you owe me about an hour of ZZZZZZZ time working up that diagram and checking my numbers... :wink: :wink: :wink: :smt015 :smt015

Rylee Isitt
Posts: 476
Joined: Fri Apr 13, 2012 3:54 pm
Location: Canada
Contact:

Post by Rylee Isitt »

Okay, based on the measurements from Charles, I've put together a drawing. Feel free to suggest changes, but remember - the more fancy it gets, the more expensive it gets!

The measurements of the flange and external thread depth were obtained from another dovetail adapter I have. They are somewhat arbitrary and there may be better values to use here.

The diameter of the flange was determined based on the width of the BHMJ nosepiece.

Note that the tabs are recessed. This can be seen in my earlier "simulation" diagram. Our part would not, as drawn, sit against the upper surface of the tabs, but rather the upper surface of the nosepiece. I am not sure if this is how the actual BH heads fit or not. We might want to explore that a bit more.

Also, we should consider machining limitations. For example, I think we need to take a bit of material out where the dovetail meets the flange so that there isn't a sharp inside corner.

Image

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

What's driving the 35mm internal diameter?

--Rik

Rylee Isitt
Posts: 476
Joined: Fri Apr 13, 2012 3:54 pm
Location: Canada
Contact:

Post by Rylee Isitt »

I estimated a "wide enough" internal diameter by measuring the inside diameter of an existing dovetail to M42 adapter that I already have. It was 35mm. I also noted that the M42 to EF adapters I have use an inside diameter of 37mm, but that might be too wide.

We can go smaller but I don't think we want to go any wider. Just as long as we don't cause vignetting in normal usage, and also, the more of the light cone we intercept, the more possible problems for internal reflections.

On second thought, 0.88mm of connection between the dovetail and the rest of the part is pretty marginal, isn't it? Want to use a 32mm internal diameter to strengthen that connection?

Rylee Isitt
Posts: 476
Joined: Fri Apr 13, 2012 3:54 pm
Location: Canada
Contact:

Post by Rylee Isitt »

I reduced the inside diameter and also fixed a little problem with the drawing proportions in my previous version:

Image

Also, you can get the PSD here.

Rylee Isitt
Posts: 476
Joined: Fri Apr 13, 2012 3:54 pm
Location: Canada
Contact:

Post by Rylee Isitt »

More changes:

- Changed threading in drawing to match amplitude and pitch of actual M42 thread.
- Threading no longer goes down to the surface of the flange. A small reduction in diameter is provided to ensure that mating part binds with flange surface (thanks to ChrisR for this suggestion)

Image

You can get the PSD here.

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

This last one looks pretty good.

It's a nit at this point, but threads are still deeper than you have them sketched. The Unified Thread Standard specifies 60 degrees, so the depth of a sharp thread with 1mm pitch is about 0.866 mm. In the current diagram, take one of those off either side and 32.00 mm inner diameter turns into 33.7 outer diameter for the inner thread, which still leaves a wall thickness of 1.07 mm. That should be safe enough. The earlier design would have cut through completely.

The one concern I have at this point is to make sure that the fabricator understands what is important and what is not. As previously discussed, what matters is the diameter of the dovetail at a point the tabs will actually touch. Charlie measures that as 39.57 mm at a distance of 3.2 mm away from the seating face. An equivalent diameter at any other distance from say 2 to 4 mm would work equally well. But that spec of 35.88 mm at the base of what appears to be just a clearance slot might not say anything useful to the fabricator. 42.80 mm diameter at 6 mm distance is a good spec, but you probably don't want them leaving a sharp corner on the bottom of the dovetail for safety reasons.

So, before going any farther toward producing a finished spec, I would strongly recommend having a conversation with your fabricator about what kind of a spec they'd like to receive. If you send them exactly what you have now, I would hope that they would come back and ask about anything they don't understand. But it's also possible that they would just measure things like the clearance slot from the diagram and then agonize over making exactly what's shown instead of something simpler that would work just as well.

I once sent to an architect some drawings that I had intended to be just a computer-drawn sketch that should be tweaked to standard dimensions. The architect, however, decided that since the diagram looked so polished he should just take dimensions right off the file. Fortunately we discovered the disconnect before those custom windows got ordered -- every one in a different size... :roll:

--Rik

Post Reply Previous topicNext topic