How to Diy objective parcentering.

[apochronaut]

The parcentering of a microscope's objectives in the nosepiece is one of the least noted quality parameters or specifications one is likely to encounter in a microscope's description. However, in older microscopes and lower cost new microscopes it easily is one of the most important overlooked performance characteristics, very close in fact to parfocality of the objectives as a prime function in establishing an instrument's ergonomic nature . Today, very seldom does a manufacturer or distributor even mention parcentering, while parfocality is a standard that is routinely referenced, if only to state the parfocal length, not if the objectives are actually tested to be parfocal.
With old school makers, where everything was produced in house and more recent OEM products manufactured in the home factory, a sense of standards and possibly even pride was/is applied to the precision of construction, and hand finishing of the nosepiece was/is a major q.c. concern.
With regards to parfocality glass has always been made from natural minerals in batches so each batch has some variability in refractive index, necessitating that an objective's internal tolerances be adjusted with very thin shims in order to meet spherical aberration specifications. The shimming has essentially the same effect as the adjustment of a correction collar on an objective so fitted, thus the point of focus and working distance of two otherwise identical objectives made from different glass batches can be different. Quality manufacturers have taken this into account when assembling a microscope for sale and select objectives with close parfocality. Very often serial numbers of the objectives would be very close because of the glass batching. While over time, this type of manufacturing variability has lessened due to more consistent batches and synthetic glass, nosepieces still have to be assembled with parfocality in mind.
Parcentering is equally a consideration too, though mediated by variability in machining tolerances not glass batch tolerances. Machining technology was heavily user controlled in the past and the results more variable than today, so great pains were taken to fit the objectives and adjust their centering. Many makers marked the nosepiece with an objective designation in the factory, guaranteeing alignment of focus and center to within their company's specifications. Objectives were matched in the nosepiece and were expected to stay there. By the 1970's , one mfg. guaranteed in brochures that their nosepieces were parcentered to 5 microns. Microscopes that have undergone repair, should have that alignment maintained. Moving objectives from one hole to another is a no no and when sn objective is replaced, it really needs to be parcentered but seldom is.
Today, machining is a little more controlled and potentially more uniform but instruments made for the entry level or general internet market can suffer from terrible parcentering due to sloppy q.c. at several junctures of manufacture.. Unfortunately and obviously, cheap internet stencil microscopes suffer from shortcuts in q.c. during manufacture, assembly and after. Used name brand microscopes can suffer from poor repair practices and component interchange. It is lucky if a microscope in either one of these categories arrives in a user's hands well parcentered but diy'ing a parcentering fix is not that difficult.

Most parcentering problems are caused by variability in objective machining ; a shoulder that is not exactly perpendicular to the optical axis. This can be exacerbated by a nosepiece having one or more ports that are not perfectly perpendicular to the radial axis and when both are under the optical tube, the optical axis is not centered to the optical tube.
Usually nosepieces are quite symmetrical or close enough but occasionally one pops up that has an obvious alignment problem at one port, usually as a result of a slightly skewed thread. Not a lot can be done about that, aside from replacing the nosepiece, which I have done. Without replacing it, two other options are available.
1) possibly one objective has a mis-alignment that is fortuitously the opposite of that occuring at the nosepiece port and the two will partially or mostly cancel each other out by placing them together If the two "splays" are not immediately subtractive, placing a very thin shim under the objective seat will cause it to tighten at a different position on the clock and adjusting the shim thickness can be used to get the objective's obverse position exact. This would occasion reestablishing a parfocal condition by shimming the other objectives but if your nosepiece is that far off, chances are parfocalization was in the cards anyway.
2) putting the objective least negatively affected by the splay in the offending port. Usually that would be the scanning objective. I suppose , for those of you with oliophobia, just put the oil immersion objective in there, since it is just an ornament anyway.

....but usually problems are caused by slight off camber alliances between objective and nosepiece, sometimes addfitive and sometimes subtractive. On a used or crappy microscope, if I can get 4 objectives within 20 microns of each other, I'm happy. I like to be able to change from one objective to any other and still have the target ohject reasonably well centered in the field. The biggest problem usually occurs with the highest magnification objective. When the field with the scanning objective is 5000 microns, and that of the 100X 200 microns it is easy for an object centered in the scanning objective to be outside the field of the 100X objective.

The first thing to do, is to test the microscope with each objective in each port. If you aren't picky whether the nosepiece rotates to the left or right , with a 4 place nosepiece you have 8 different combinations with the objectives in numerical sequence. if you are picky about direction , then you have 4 and 16 if you don't care about numerical sequence. Usually this solves parcentering problems and gets reasonable parcentering.

Sometimes not though, as in this recent case history. I acquired another interchangeable nosepiece for my AO series 4 and was able then to load it with a set of objectives dedicated to DF. A 20X .50 achromat, a 43X .85 L.W.D. dry achromat, a 50X .95 oil immersion achromat and a 97X 1.30 fluorite. These are all 1950's objectives but the set performs extremely well with an oil immersion DF condenser. The latter objective is fitted with a larger than factory funnel stop, allowing the objective to work at a peak N.A. and still achieve D.F., probably around 1.15 or even 1.20 N.A. The imaging is exceptional for such an old microscope with a 20mm f.o.v. and all the objectives original AO and completely parfocal. I really wanted this quite uncommon fluorite objective to work but when I put that objective in the nosepiece with the other objectives , it didn't matter which port I put it in, it still pointed over 100 microns off center. A 40 micron object centered through the 50X objective couldn't even be seen through the 97X. All the other objectives are within 10 microns of each other. It seems that either the shoulder or thread had been mis-machined in the factory.

There are two low tech. ways to fix this. One is more suited to infinity corrected microscopes since it involves using tapered shims , therefore lengthening the tube slightly and you have to procure the shims. Shims have reached the level of a rare and elite product these days. One retailer of them wants 10.00 each.
Because of the reversal of the image, it is sometimes difficult for some to envision which side on which to taper the shim. It is easy to think of it this way. You want to move the end of the objective in the direction that you want the image or an off center object to move. Shims are for relatively gross errors : more than 75 microns. For small errors I would use method 2. So, if using a shim, install all the objectives in the nosepiece and choose the best case scenario for their arrangement. I then get an idea of what thickness is required by loosening the offending objective on it's threads a little then tipping it slightly while looking through the eyepieces to get the desired alignment. After a couple of tries and perhaps the use of feeler gauges, you should get an idea of a minimum thickness of shim required. Choose one a little thicker, just in case. Then , after you have acquired the shim, it is just a matter of filing or dressing it on a sandpaper block or stone, lightly and incrementally to get a wedge form with the thin side installed on the side farthest from the object you are using as a target. So for instance, in my case history here, when an object was centered with the 50X, it was somewhere south of 7 :00 in the field, with the 97X. I will use the clock face for my reference points. If I were to have used a tapered shim to fix that, the thin side of the tapered shim would have been at 1:00. The downside of shimming is that you change the parfocality some and often you have to flat shim the other objectives as well in order to reestablish parfocality. Shimming changes the tube length a bit and might contribute to a bit of sa with certain finite objectives and wedged shims are tricky to shape and install. You have to compensate for clutching and slipping of the shim when tightening it each time you test your progress and at final installation. Always clean filings from the shim before draping it around the threads. I have used shimming with a couple of dreadful Chinese made planachros. One other downside is that once installed and aligned, you dread removing the objective because it takes a little time to realign it and always use the same port.
Method two is somewhat ruthless and some might be a bit squeamish about putting a file to their objectives but that was the insult I applied to the 97X fluorite. It is now about 20 microns west towards 9:00 of the 50X center. I might get it closer yet but it now pretty much works as intended. This process is a little better for finite objectives and a lot cheaper than buying parfocalizing shims at ridiculous prices. I use a quite fine very small good quality flat file. Mine was a Swiss made Nicholson tapered file but a small bastard file will do. You want fine, flat and sharp. First, plug the rear of the objective or cover it with an easily removable cover because it will need to be tested frequently. Basically the process is pretty much the same as that for the shim but you are dressing the objective mating surface : the area of the shoulder that snugs against the nosepiece. In this case though, there is no clutching and rotating of an intermediary piece but there is a constant additional number of degrees of rotaion each time you file a little and test the objective. This will cause the target object to move as much sideways as in the direction you intend, unless you compensate for the increased rotation by filing somewhat clockwise from the location directly across the clock from your target object. After about 10 or so filing sessions on the far shoulder in the roughly 2:00 to 3:00 location on the objective shoulder face, I was able to move my target object from outside the field at 7:00, to about 20 microns off center towards 9:00. Each time the file is used , the objective rotates a little more to it's stop and the target object moves, but not always in a straight line. You have to file a different location each time on the shoulder and at a slightly different location on the clock. For instance, now that I have gotten this far with centering the objective, Any further centering will require filing at approx. the 4:00 position. This isn't super precise like doing it on a lathe but the results can be excellent when done with patience and care. Always clean filings from the objective prior to testing.

[FredH]

apochronaut,

Thanks for this very informative post.

If I understand your last paragraph, you deduced from position of the image where the shoulder was contacting the nosepiece, and filed that point of the shoulder down about 0.0035 inches (enough to move the contact point about 45 degrees) to move the image close to where it would have been had the shoulder been perpendicular to the objective's optical axis.

My question is, knowing how far you had to move the image to achieve parcentricity, could you then predict how much material had to be removed from the shoulder to center the image? Otherwise, your method of filing and checking, filing and checking seems much better than the "super precise" method of using a lathe.

[Phill Brown]

Another great contribution thanks. Striving for perfection in production is the biggest waste of time and money resulting in bankruptcy or buy out.
There is an agreed standard, sometimes that is exceeded by chance.
Below that is usually left for the phase B tester to dispute and resolve.
The intended lifespan of commercial products seems to be an ever decreasing spiral into below average quality.

[apochronaut]

apochronaut,

Thanks for this very informative post.

If I understand your last paragraph, you deduced from position of the image where the shoulder was contacting the nosepiece, and filed that point of the shoulder down about 0.0035 inches (enough to move the contact point about 45 degrees) to move the image close to where it would have been had the shoulder been perpendicular to the objective's optical axis.

I hope this helps get a few stands more fun to use.

I may be reading this incorrectly but part one of your understanding is the opposite of what was intended and part two correct. Maybe I should go back and reread what I wrote.
The part of the shoulder to be filed is 180° opposed to the location of your target object ( not in the case of a stereo or some inspection mics.) and depending on how much you file, the object will move somewhere between directly towards the center and likely slightly clockwise from that. The next pass or few passes of the file will take place somewhat counterclockwise of the previous location due to a slight counterclockwise rotation of the objective as a result of material having been removed from the shoulder. It really is a trial and error method but with patience it works. It took me the better part of an hour to get the 97X fluorite adjusted and I am glad I did.
Theoretically a proper lathe job would be the best but unless the job is proper, a lot of damage could be done. Not everyone has a lathe or access to a good machinist and a lathe is not applicable to all cases of objective misalignment. This slow, incremental, empirical method can correct for any area of variance in the objective whereas lathing can only correct for burs or an off camber mating surface. Probably most factory machining errors are in the threading, so lathing the shoulder mating surface perpendicular to the barrel will accomplish nothing.

[apochronaut]

Another great contribution thanks. Striving for perfection in production is the biggest waste of time and money resulting in bankruptcy or buy out.
There is an agreed standard, sometimes that is exceeded by chance.
Below that is usually left for the phase B tester to dispute and resolve.
The intended lifespan of commercial products seems to be an ever decreasing spiral into below average quality.

Thanks, Phil.

[woyjwjl]

About a year ago, I sold all fixed nose wheels and switched to using two phase nose wheels.

I'm just pretending to understand this post when I say these things

[apochronaut]

In former times, there were little adjusting screws on each objective port on some stands. These were over time deleted in favour of more precision production methods. Metallurgicsl objectives have always or at least for 100+ years had a method of centering individually mounted objectives using either a thread or snap in saddle mount, more recently being replaced by nosepieces with individually centered ports.