Mantis Microscope Repair/Teardown

[nikonoid]

I recently purchased Mantis Macro microscope in "parts or repair" condition. This seems like a fairly recent model, similar to "Elite" but having a straight view path instead of bent.



It came in great physical shape, however, as promised, not fully functional. My unit produces a nice sharp image for left and right eye separately, however no matter how much I adjust interpupilary distance I still get "double" vision when looking with two eyes at the same time. When I rotate the pupil adjustment knob the left eye picture remains stationary while the right eye picture moves. I am not sure if that is a correct behavior. Maybe someone with working Mantis can check their unit to see if one picture or both of them are moving.

Information on Mantis construction and repair is fairly scarce and I was not able to find any teardown photos.

I will have to open my unit up for repair and plan to take good photos of the guts. I am looking for any sort of information or word of caution before I tear into it. Thanks.
 

[Samogon]

Would be interesting to see its guts.
And if it really worse to spend big buck for it in comparison lets say $500-700 Amscope or similar

[Zucca]

nikonoid is in my buddy list now. thanks for posting this, I'm looking forward your findings.

[nikonoid]

I started slow exploration of the Mantis.

The easiest entry is the door for desiccant bag. Beside the bag it exposes access to 5 set screws that probably are responsible for optical adjustments. These are accessible without disassembly of the head unit and therefore it is very likely that the microscope is assembled first and then optically aligned. This alignment is possibly requiring an optical bench or some sort of collimator.



Next I removed the glare protection hood and that exposed a large viewing lens. The lens was not fixed in place with anything but the hood. I hope it's precise position is not very important.



Under the lens I could see a number of mirrors.



So far I found a tiny spider and a bit of a web on inside. There is nothing that is obviously broken or out of place.

This is all for now. Next time I may try to open the bottom plate. It is fairly hard to photograph mirrors. I an considering rigging up a laser source trying to see and measure what the mirrors are doing.

[Samogon]

As i expected, nothing obvious. And it is scarry to displace linse or mirror and make things worse.
But dont let my words to discourage you

[nikonoid]

Today I had few minutes to diagnose the problem a little more. It looks like interpupilary distance adjustment is working ok after all.

It is the picture for each eye that is misaligned. When I look with a single eye I can see that left and right halves of my view are slightly misaligned with a small overlap in the middle!

I fashioned a crude optical bench to analyze it a bit more.

See two separate projections for each eye. Then each eye has two halves that are misaligned.

I think my next attempt will be to adjust these set screws:

Anyone can venture a guess what these do?

It is a strange L shaped configuration and I have seen in pictures before. It is possible that some are responsible for adjustment and some for fixation...

[tautech]

Does this help:
http://www.cmm.com.mx/pdf/vision%20engeneering/mantis/Mantis%20Operators%20Manual.pdf

[TiN]

Subbed, eventually I'll get hands on one of these. So far using usual rework binocular microscope to work on small stuff.
Couldn't imagine working on electronics without the one! Decent microscope is tool #2 after the good soldering iron. DMM is only tool number #3  .

[Zucca]

Those are wonderful machines worth every single penny, everytime I turn mine on there is a big smile in my face. I am so glad I got one, even if it's the old model, I am so happy with it.

nikonoid yeah it's a crude optical bench but rather clever. Nice work. Instead of a line I would put a sniper cross printed on a transparent paper at the bottom...



Surely you have already marked the initial screws position. Is it the Loctite Red THREADLOCKER what I see? I thik to get those screw moving you need then 260C of temperature...  pay attention not to met the plastic somewhere, use alu foil to protect the plastic parts.

[Samogon]

I see that Mantis Compact are sold on auction site under 2k new, are they suitable for electronics repair shop?
I see it has x8 magnification max. So does it worth or Elite is a winner here?
Mantis owners/users, can you advise?

Thank you

[mikeselectricstuff]

x4 is the best tradeoff between magnification and field of view for working under.
x6 can be useful, but I think x8 would be too much as the FOV would be too small

[Zucca]

Here my humble experience.

My x8 gets in focus at about 1cm distance from the object you want to observe:

1 - any soldering actions are difficoult since the lens is in the way, it is just like most pretty girls: "look at me but don't touch".
2 - a used x8 has probably scratches since it can be easely bumped on the target trying to get things in focus.

4x is the way to go 99%, 6x is my emergency call, 8x happens once a year.

Of course if you are working mainly on tiny mobile chips, you may want to consider a binocular scope.

[Armadillo]

That is one valuable piece of information there. Thanks.
Now the myth in my mind is debunked as far as answering whether it meets the needs in electronic repair.
I would prefer X90 but minimum X45 to be dealing with the ultra fine pitch [0.5mm and less] and high density chips in nowadays circuit and at minimum 150mm clearance space.

Wonder what this Mantis is actually use for, at such expensive price and whether it can mount a camera like a typical trinocular scope. ?

[Zucca]

I sold my 90x trinocular and kept my Mantis.

The image quality and the "3D feeling" is just another level, in Italy we use to say: "You can't compare chocolate with shit".
You have to try a Mantis, no forum post will help you to understand.

That said for fine pitch, IMHO, the Mantis shows its limits.

Same gold rule as alaways: different jobs, different tools.

[mikeselectricstuff]

Some figures for Mantis sompact :
x4 : width of field 30mm, distance 100mm
x6 : width 20mm, distance 70mm

the x6 lens is quite a bit longer, so the 70mm is from the 50mm lens diameter, not the much larger width of the mantis body, so still fine for working under

If I only had one, it would be the x4.

[mikeselectricstuff]

Another advantage of the Mantis is that as your eyes aren't right up against eyepieces, it's very comfortable to use for long periods. the 3D view is also really useful.
You can also nudge the position around with your forehead for hands-free adjustment if position and focus.

 

[Samogon]

Thank you guys!
So I probably keep my x45 for fine jobs, add mantis compact. Here is come bench space into consideration .

[oldway]

Perhaps it is interesting to remember that during a previous discussion, a staff member of Vision Engineering, manufacturer of Mantis Elite, participated in the discussion on this forum.

His name is gturemen and he may probably be able to give you the information you need to set and calibrate the device

https://www.eevblog.com/forum/testgear/frustrating-after-sale-vision-engineering-mantis-elite/.

[Armadillo]

Thank you guys!
So I probably keep my x45 for fine jobs, add mantis compact. Here is come bench space into consideration .

Try to probe the scalar chip confidently and you would change;
So I probably keep my x45 for fine jobs.   

[nikonoid]

nikonoid yeah it's a crude optical bench but rather clever. Nice work. Instead of a line I would put a sniper cross printed on a transparent paper at the bottom...

Surely you have already marked the initial screws position. Is it the Loctite Red THREADLOCKER what I see? I thik to get those screw moving you need then 260C of temperature...  pay attention not to met the plastic somewhere, use alu foil to protect the plastic parts.

zucca, thanks for a point about THREADLOCKER. Sorry for delayed response, I got very tied up with work.  I am hoping that this is not what they used, yet I will try the temperature if regular force is not enough to move screws.  By the way would you use hot air gun for THREADLOCKER or soldering iron to heat bolts?

As the line is concerned, I am not putting the line in that you see in the pictures. Without objective lens the mantis is focused at infinity, so the flash light is completely out of focus and represents a broad uniform light source. You would expect just a uniform white image from ocular projections. However it seems that the image of each eye is sliced in two and these two halves are misaligned. Then they intersect you can see a brighter section, when whey are apart a darker section between them. You can see that angle between the halves is not 0 and the dark line is wider on one end and narrower on another.

I am just hoping this misalignment can be resolved with the screw involved and is not an indication of something really broken on the inside.

So far I used make shift optics to do my testing. This week I should receive a proper mantis elite lens and test with that, before going deeper into the body.

[Zucca]

[By the way would you use hot air gun for THREADLOCKER or soldering iron to heat bolts?

Hot air and slow flow, with an iron it is difficoult to get an area (not a point) in temperature.

As the line is concerned...

You got me, I understood wrong that line business. After further thinking I got my mistake and crossed the sentence "Instead of a line". Thanks for your explanation.

Anyway it looks like those screws are tuned at the production facility and than sealed with red junk. If now the device is missaligned and the screws did not move, I suspect that mantis suffered a mechanical stress and the mirrors inside moved or jumped out from the correct position. Before playing with the screws I would look further inside to understand if the mirrors are well seated in position. I don't think it will be an easy job anyway....

What about to put a laser instead a torch light? You should then see in a good aligned Mantis two points, one at each ocular. I think in your case you will get 4 points, two each ocular.
By moving those screws you should then able to close the distance  beetween those points. Another risk could be that the image get rotated and it is not perfect aligned. Again by moving the laser you can see what the ocular are spitting out.

[nikonoid]

I finally had a bit more time to work on this.

I tried the adjustment screws. Despite appearance of threadlock they were not locked. Simple attempts to adjust them did not help at all.

I performed another experiment. Two post-it notes of different color were aligned and attached to each other. When looking at the composition edge on you should see one color with one eye and different color with the other. I used this to confirm that I indeed have a 3D image.

Than I decided to go for broke and pen the mantis. LED light are very easy to remove, just one screw and they pull out together with connector. Then lens carousel is another screw. Additional 4 screw separate the mechanism used switching lenses. It almost looks too complicated of a contraption for the simple function that it serves.


Viewing hood is easy to remove and that exposes a large lens that was held in place by that hood. It can now be removed.


Then two screw from the top and two from the bottom remove the bottom panel that is holding two mirrors at 90 degrees angle. I believe it is responsible for splitting image into left and right eye view. One mirror is fixed and the second is adjusted by 5 screws I shown before. They are the screws that are adjustable through desiccant port. In fact after looking at the insides on the microscope this is the only adjustable part.




With bottom removed you can see a number of flat mirrors, a large panel of flat glass that i think serves as semitransparent mirror. These mirrors are very hard to photograph. I may have to draw a diagram. So far I still do not understand the optical design.



The good news so far is that the microscope is fairly clean inside beside few strands of spiderweb. There is nothing obviously broken or out of place.

It is beautifully designed, appearing simple, elegant, yet mysterious.

I will keep exploring...




Sent from my iPhone using Tapatalk

[tautech]

These set screws: (not the knurled nuts)



What do they do ?

As the 'modules' are affixed to the ABS frame/cover and not to each other maybe a hot lab has allowed the ABS to move and distort the proper mirror and lens alignments.

IMO the knurled nuts hold each assembly firmly against the adjustment set screws BUT if the enclosure is distorted or the unit has been dropped an assembly or two might have moved from proper alignment.

Great WTF project. 

[nikonoid]

Basically three small set screws push the metal plate with a mirror away. Two large knob screws pull the metal plate in. The combination allows you to securely control two angles of the mirror.
The second perpendicular mirror is attached to movable mirror by something looking like a hot glue. It should provide some flexibility.
I will have to investigate this a bit more.


Sent from my iPhone using Tapatalk

[Zucca]

WOW

[nikonoid]

Success at last:

Last weekend I finally had some time to open the microscope a bit more. I removed the black bottom plate with glass aperture.



At this point I could see every bit of the optical system. This proved conclusively that nothing is broken or moved.

At this point it was clear the only part of the optical system that could be giving me trouble with double image in the middle is the right angle mirrors:



Mirrors looked to be approximately at the right angle. How can we make sure? Lasers of course!

If mirror are at the right angle they will return the beam that is parallel to the source. More over if you point laser exactly at the point where mirrors meet, they will return a single dot if they are aligned and dual dots if they are not. The distance between two dots is representative of the angle error. If the mirrors are perpendicular to each other and laser beam if perpendicular to line where they intersect, the mirrors will return the beam back to the source.

In a pinch I used good quality well aligned laser from a distance meter and several 1-2-3 blocks.




Simple experiment with a target 4 feet away from the mirror revealed about 1.5" error! Then I raised the contraption by 1" to measure different section of the mirror, the error was almost the same. Set screws were user to correct the error as much as I could. Once seemingly perfect, then I repeated the same experiment at 15 feet and adjusted again and well and I could. This time the error was about 1/8".

I was not sure what is good enough here. So I assembled the microscope back to check and ... IT WORKED!!! The double vision is gone. The picture is spectacular.

The design of the Mantis is ingenious. It is almost too simple - four plastic parts, bunch of mirror held by double tape and single large lens. Yet I still do not completely understand how and why it works. There is no separate image path for every eye. All the optics on inside is "wide enough" to cover both eyes.

Taking pictures of bunch of mirrors is almost pointless. It is like Alice in Wonderland: all you see is countless reflections and not the reality.
Here is a crude diagram of what I found inside, with some basic explanations:

[Samogon]

Congratulations!!!
And thank you for sharing this journey to unknown.

[Zucca]

So so cool... thank you so much. 

You should have done a Youtube video on this, i guess 100K views in less than a month.,,

[nikonoid]

Thank you guys, it was a cool project and I enjoyed working on it.

I am waiting for 4x lens and since lens protection caps for these cost quite a lot, I am making my own from 52mm clear UV filter.


I have never done videos before. Maybe one day.

[Koray]

Thanks for the light path diagram! Amazing!

K.

[Pinkus]

Cool repair documentation. Thank you for sharing it.
One hint:
As being said: the 8x lens usually does not provide enough space for soldering, and even with the 6x lens, this often is difficult. The 4x lens on the other hand often does not offer enough magnification.
However, the famous 6x SLWD ("super long working distance" I guess) offers you even more working space as the 4x lens. Thus I use it all the time. It is a massive chunk of glass and unfortunately pretty expensive ($1500 new, at least $500 used) but worth it.

Edit: here: https://www.thingiverse.com/thing:1330221 somebody is showing a DIY version of a SLWD lens. Before purchasing the genuine one, I tried this DIY lens myself, but the picture quality was very poor.

[nixxon]

Copy of PM:

In the original Manis, the concave (light focusing) mirror in the back of the unit could be tilted up/down to elevate/lower the right output images. The top concave mirror could be routinely adjusted with a knob in the horizontal direction, to move the left output image to match the distance between the user's eyes.

I actually disassembled a Compact as well to clean the optics. It was sold shortly after that. I don't think I ever adjusted the image up/down on the Compact, so I'm afraid I can't help you with the Compact. If the working principles are similar between the models, the flat mirror assembly in the middle of the head should be fixed and oriented in precise relation to each other (reflecting light 90 degrees every time). The concave mirrors seated in the outer shell should be adjustable, to adjust right/left output images either sideways or up/down.

[ruairi]

Fantastic thread, thank you.

I bought an original Mantis used a little while back but haven't had time to mount it on a bench and test it.  It came with a pair of x 10 lenses so I need to track down a x 4 or swap with someone. 

Cheers,
Ruairi

[toybuilder]

The "semitransparent mirror" is a beam splitter.  Half of the light passes through half of it reflects.  This creates two sets of beams -- the left eye beam is affected by the aim of left eye mirror, the right eye beam is affected by the right eye mirror.  Here's a simplified diagram (as I understand it):


The left and right eye focusing mirror (at least the concave mirrors on my original Mantis), as diagrammed, are aimed to overlap the exit beams -- but if the mirrors are tilted they will shift the beams paths creating the offsets needed for stereo viewing.