Cheap LWD finite 20x objective for looking into 1mm of water

zzffnn · Post by **zzffnn** » Sat Sep 12, 2015 1:07 pm

No Mitutoyo or anything over $200 please, it has to be cheap.

I am looking for a long working distance version of a finite 20x objective. Highly prefer a decent NA of 0.4, common RMS mount and no cover slip type (upright use, 0.17 acceptable, but no 1-1.2 please). WD > 4 mm. Prefer 160 mm tube length, but I can work with any finite tube length since I can extend my tube length with PVC pipes if I have to. No infinity objective, please.

Application is bright field and oblique on upright compound microscope. I have seen some phase objectives, but not sure those would work well.

This Olympus M Plan seems to offer a WD of 4.6 mm at reasonable price, which is about right for my use (though longer WD is preferred):
http://m.ebay.com/itm/Olympus-Microscop ... 1441339362
I don't know what its tube length or cover lip thickness is though.

I know most members here like Nikon's M Plan 20x 0.4 in 210 mm or infinity versions, but those seem to go over $200 most of them times on eBay (too expensive for my casual use).

Please suggest more alternative's, if you know any. Thank you!

Charles Krebs · Post by **Charles Krebs** » Sat Sep 12, 2015 2:10 pm

This Olympus M Plan seems to offer a WD of 4.6 mm at reasonable price, which is about right for my use (though longer WD is preferred):
http://m.ebay.com/itm/Olympus-Microscop ... 1441339362
I don't know what its tube length or cover lip thickness is though.

These would have been be designed for no cover slip. The tube length is most likely 200mm.

ChrisR · Post by **ChrisR** » Sat Sep 12, 2015 2:47 pm

I have a Unitron 20x ELWD, which comes up fairly often, and they aren't expensive. If I remember correctly they're for a strange tube length. From memory there is some CA though. How much of a bother that is, after post processing, depends on the enlargement, your subject, your standards...!

It's possible that it should have a compensating eyepiece which the Olympus one you have found, also does.

I can't remember any other "cheap" 20x long WD objective, other than one which appeared a couple of years ago always with its outer shell missing (which didn't matter) "We" generally thought it was about a NA 0.35, and quite good. WD is sevreal mm, I forget exactly.

The only other option worth a mention IS infinite - the Mitutoyo clones, if you're lucky. There are several, which look very similar, though you'd still be doing well to find one for $200. The UMSCO is almost as good as a Mitutoyo, very much the same centrally at least.

zzffnn · Post by **zzffnn** » Sat Sep 12, 2015 3:23 pm

Thank you both very much.

Charles,
That makes sense as it is a metallurgical objective.

Chris,
I don't mind CA or SA, as long as it is not a huge amount of distortion. The primary purpose of this objective for me is sample isolation and preparation - when I use 20x to look through a petri dish with pond water, LWD allows more sample volume and space for my micro pipette to move in and out easily. I know an inverted scope is better for that application, but I don't have the space for such a big scope (for the cost of an inverted ~ $300, I would rather buy an ELWD 40x). I already have LWD objectives of up to 9x NA 0.2 WD 13.5 mm.

And of course, it would be icing on the cake, if the objectve's image quality is decent without cover slips. I can use it on a microscope for extreme (afocal) macro. Already have slim fiber optics with 5 mm output ends.

Unfortunately, my intended scope set-up cannot use infinity objectives.

Edit:
I did see this Unitron, but its NA is only 0.3. Do all the Unitron 20x LWD have this low NA?
http://m.ebay.com/itm/UNITRON-Microscop ... 1789371301

zzffnn · Post by **zzffnn** » Sat Sep 12, 2015 3:47 pm

So I searched "Unitron 20 microscope" at eBay and found many listings that read: M Plan 20 NA 0.4 (or 0.45) and tube length 170 mm. They don't say LWD though.

The following is a typical example:
http://m.ebay.com/itm/MICROSCOPE-PART-O ... 2016942802

Chris, does you LWD Unitron says LWD on its label? Does it have a catalog or serial number on it? Thank you.

I have found many different numbers on those Unitron 20x objectives, including:
92432
80411
12207
82734
80394

Such numbers may be unique to each objective (like a batch/product number), so it may not be useful?

ChrisR · Post by **ChrisR** » Sat Sep 12, 2015 5:19 pm

That one (281789371301) looks like the one which I have. I was going to suggest it was 240 TL, but I didn't remember anything else that long.

I made some reasonable direct-projection images with it with some PP. Then other optics came along.
I wonder if it's designed for a compensating eyepiece - perhaps other folk will know? I believe these are objectives which can come with many manufacturer names.

zzffnn · Post by **zzffnn** » Sat Sep 12, 2015 8:35 pm

Chris,

I actually saw that one before opening this thread and before its listing ended. But I did not click the "Buy It Now" button at eBay because it has a low NA of 0.3, which is not much better than that of a regular 10x?

I am guessing that 20x 0.3 would have a WD of over 10 mm, otherwise that low NA is hard to justify? Chris, do you remember what WD yours has?

It is a difficult decision, since that Olympus M Plan LWD 20x has NA of 0.4 / WD of 6.8mm and is listed at $128 shipped. The Unitron is not that cheap, considering the unknown WD and its low NA..........

Chris S. · Post by **Chris S.** » Sat Sep 12, 2015 9:46 pm

Fan,

For your particular use and budget, I wonder if it’s worth raising a question mark about one of your requirements:

zzffnn wrote:. . . 0.17 acceptable, but no 1-1.2 please . . . .

The question is as follows, and the best person I can think of to answer it is Rik: How well would an objective with an NA of .40, designed for use with a 1.2mm cover glass, be expected to work when used with no cover glass to find and manipulate subjects? As you’ve said, you won’t be photographing through it, or doing detailed analysis. And you’ve also said, some spherical aberration is acceptable.

I have a hunch that such an objective would work OK for your particular situation. If I recall correctly, spherical aberration is vastly more dependent on NA than on cover glass thickness. I could look up the formula in one of Rik’s posts on the subject, but would want him to give perspective on the results of the calculation anyway, so will raise the question and leave it at that.

As I’m sure you’ve seen, there is a 20x/0.40 ELWD 160/1.2 phase objective on eBay for $170 or best offer. I suspect a reasonable offer would stand a good chance of getting accepted. The phase part shouldn't be a problem for use as a finder objective. Item number is 171874234994. (I would normally not list the item number, but your need here is fairly unusual, and I don’t think there will be a run on this item.)

--Chris S.

Pau · Post by **Pau** » Sun Sep 13, 2015 12:14 am

If I recall correctly, spherical aberration is vastly more dependent on NA than on cover glass thickness. I could look up the formula in one of Rik’s posts on the subject, but would want him to give perspective on the results of the calculation anyway, so will raise the question and leave it at that.

As I’m sure you’ve seen, there is a 20x/0.40 ELWD 160/1.2 phase objective on eBay for $170 or best offer. I suspect a reasonable offer would stand a good chance of getting accepted. The phase part shouldn't be a problem for use as a finder objective. Item number is 171874234994. ..

This one is for inverted microscopes, ie to see through the main glass of the microscope slide. I've not the calculation but intutively I think that the spherical aberration will be notable (I have some experience with students putting the slide upside down and the image is clearly degraded with a 10/0.25 objective) You can test the idea: just put a prepared slide upside down and take a look with your 10X

rjlittlefield · Post by **rjlittlefield** » Sun Sep 13, 2015 1:02 am

Chris S. wrote:
zzffnn wrote:. . . 0.17 acceptable, but no 1-1.2 please . . . .
The question is as follows, and the best person I can think of to answer it is Rik: How well would an objective with an NA of .40, designed for use with a 1.2mm cover glass, be expected to work when used with no cover glass to find and manipulate subjects? As you’ve said, you won’t be photographing through it, or doing detailed analysis. And you’ve also said, some spherical aberration is acceptable.

I have a hunch that such an objective would work OK for your particular situation. If I recall correctly, spherical aberration is vastly more dependent on NA than on cover glass thickness. I could look up the formula in one of Rik’s posts on the subject, but would want him to give perspective on the results of the calculation anyway, so will raise the question and leave it at that.

You're probably recalling that wavefront error goes as NA^4 at fixed thickness, but only as thickness^1 at fixed NA.

I think the hard question here is just how much error are we talking about, and more importantly, what does that do to the image?

I don't have a calculator handy to address exactly the question that was asked, but I do have one that goes the other direction: adding some thickness of dense medium where none is expected. The answer in that case is that adding 1 mm of NA 1.5 glass to an NA 0.40 objective introduces about 0.45 lambda wavefront error (lambda = 546 nm).

0.45 lambda is not a catastrophic loss like 0.68 would be, but it definitely cuts the contrast and sharpness by quite a bit.

This is a fine-detail result that I calculated last year:

This is also a situation that can be investigated experimentally, so I hauled out an M Plan 20X NA 0.40 and visually looked at a high contrast target (copier toner dots on white paper), with and without an ordinary 1 mm slide dropped over the target.

In that test, I saw much less visible difference between the two cases than is suggested by the computed images shown above.

Using a simple afocal setup, I tried to capture a pair of images that could be displayed side by side to illustrate the difference. However, that attempt failed because I kept getting mismatched focus points. Shooting some stacks and picking individual frames would have solved that problem, but my stacking gear is set up doing other things.

In any case, I would characterize the degradation as "might be acceptable, might not, depending on tastes". The effect of introducing the extra glass turns a crisp lens into a not crisp one. How much that would affect the intended use, I can't say. It's an experiment that Fan might be able to run with objectives on hand, as a way of evaluating the idea.

But re-reading Fan's list of desired features, I notice "it would be icing on the cake, if the objectve's image quality is decent without cover slips. I can use it on a microscope for extreme (afocal) macro". I don't think NA 0.40 with a 1 mm cover glass error would be very good for that use.

--Rik

ChrisR · Post by **ChrisR** » Sun Sep 13, 2015 2:48 am

If you're going to "use 20x to look through a petri dish with pond water", I suspect the optical quality will be affected by the water quite a bit. More than the diffrence between NA 0.3 and 0.4.

At 10x on an Olympus, going from NA0.25 to NA 0.45 and using the correct eyepiece instead of the wrong one, I wouldn't say the visible difference exactly jumps at you, though it's a bit brighter!

The Unitron WD is "several" mm, not out there with the Mittys and SLWD Nikons.

zzffnn · Post by **zzffnn** » Sun Sep 13, 2015 5:52 am

Thank you very much, gents.

Your excellent comments made me rethink my requirement. I should probably drop that "icing on cake" consideration and buy, in the future, another LWD no cover 20x 0.4 just for extreme macro.

For my CUURENT primary purpose - looking/videotaping through 1mm of water held in petri dish on an upright scope (assuming subject protist is at the bottom bellow 1mm of water), I am guessing:

A) an inverted biological objective corrected for 1mm of glass
is probably a bit better than
B) an upright metallurgical objective corrected for no cover slips?

In A), the inverted biological objective expects 1mm of glass (RI=1.52), but sees 1 mm of water (RI=1.33).

In B), the upright metallurgical objective expects air (no cover, air RI=1.0), but sees 1 mm of water.

I don't know the mathematical formulation for theoretical computation. Rik, maybe you can help us out?

I am just guessing - if I have to bet now, I would guess A) produces less optical error and better image? Sorry, I do not really have the equipments to test, so mathematical calculation/prediction is good enough for me.

I could be totally wrong on this (please feel free to correct me):
Based on my very rough approximation of some formulations and calculations that I have seen, it seems to me that:

a NA 0.25 objective that expects one 0.17mm cover slip can tolerate about eight additional cover slips (over 1mm of glass);
a NA 0.3 objective that expects one 0.17mm cover slip can tolerate about three additional cover slips;
While a NA 0.4 objective that expects one cover slip can only tolerate one additional cover slip.

ChrisR · Post by **ChrisR** » Sun Sep 13, 2015 7:23 am

I'm not a pondwater guy, but for good macro photos at 20x, doesn't everyone use a coverslip? Maybe a well slide?
Gives you a more stationary subject, a flat surface, and less water.

zzffnn · Post by **zzffnn** » Sun Sep 13, 2015 7:41 am

ChrisR wrote:I'm not a pondwater guy, but for good macro photos at 20x, doesn't everyone use a coverslip? Maybe a well slide?
Gives you a more stationary subject, a flat surface, and less water.

You are most likely right, though I am not sure everyone does that. I imagine some thick insects may look good under lwd 20x and deep stacks? I did read Rik's excellent threads about shooting through water and am aware of the benefits.

rjlittlefield · Post by **rjlittlefield** » Sun Sep 13, 2015 9:05 pm

zzffnn wrote:I don't know the mathematical formulation for theoretical computation. Rik, maybe you can help us out?

Maybe. This has been an interesting problem. I've been fiddling with it for hours.

First, let me explain that I don't know any simple formulas for calculating spherical aberration with large apertures. My usual method for calculating from theory is to construct a spreadsheet that solves an optimization problem that is based on first principles of refraction. By "first principles", I mean nothing more than 1) the optical path length through dense medium is equal to the refractive index times the geometric path length, and 2) the actual ray path through multiple media from A to B is whatever minimizes the optical path length. Calculating it this way is computationally expensive, but it has the advantage that I can avoid possible misapplication of more complicated formulas, which I consider a huge risk.

What I've ended up doing today is to extend my spreadsheet so that it understands how to correct an objective for a slab of arbitrary depth and RI, then calculate for a slab of different depth and RI.

The numerical results for NA 0.4 are as follows:

In A), the inverted biological objective expects 1 mm of glass (RI=1.52), but sees 1 mm of water (RI=1.33).

I calculate wavefront error = 0.05 lambda (not significant, in my view).

In B), the upright metallurgical objective expects air (no cover, air RI=1.0), but sees 1 mm of water.

I calculate wavefront error = 0.41 lambda (definitely significant).

I don't know any intuitive explanation for why it is so much more painful to go up from RI=1.0 to RI=1.33, than it is to go down from RI=1.52 to RI=1.33.

My calculations agree with intuition that it is equally bad to be expecting a block that isn't there, and to confront the same block when it's not expected. For example, using the inverted 1 mm objective without cover is predicted to be exactly as bad as adding 1 mm cover to the metallurgical objective.

But if I think about having a 1mm thickness of RI=X, and continuously varying X from 1 (air) to 1.52 (glass), then the relationship between RI and wavefront error is notably nonlinear. The "equal pain point" (EPP) where I get half the wavefront error is down around RI = 1.12 . So it's much less traumatic to go from 1.52 down to 1.33 (about halfway to the EPP), than it is to go from 1.00 up to 1.33 (almost three times as far as the EPP).

I redid the calculation for Chris S.'s objective, corrected for 1.2 mm of glass instead of only 1 mm. The answer is not much different: looking through 1 mm of water then gives wavefront error = 0.061, still not important in my estimation.

I don't have a good feel for why the dependence on R.I. is so nonlinear. My natural skepticism leads me to be suspicious of some error in the calculation. But the approach I'm using doesn't lend itself to introducing nonlinearity as a bug, so I'm going to tag these numbers as definitely worth a physical test.

a NA 0.25 objective that expects one 0.17mm cover slip can tolerate about eight additional cover slips (over 1mm of glass);
a NA 0.3 objective that expects one 0.17mm cover slip can tolerate about three additional cover slips;
While a NA 0.4 objective that expects one cover slip can only tolerate one additional cover slip.

The criterion for "tolerate" must be pretty stringent in these rules.

I calculate that they correspond to wavefront errors of 0.11 lambda for the NA 0.25 case and about 0.08 lambda for the other two cases.

Experimentally, I can certainly detect SA when adding an extra cover slip to NA 0.40, but to my eye the image does not degrade obnoxiously until I add several of them. Even looking through a full 1mm slide (giving wavefront error = 0.38 in my calculations), I see a lot of haze but no loss of high contrast detail.

Have you run a test like this to see how your own visual experience matches up against the words in the literature?

--Rik