Original page created on 18/09/2022; updated on 11/11/2022.

New slide holder for 35 mm lens

As I said before, a short focal length lens allows using a much shorter slide holder. I have just purchased a 35 mm F2.8 macro lens. This lens is a bit old: in particular, it doesn’t have an internal focus drive. However, this is a very good lens optimized for macro and allowing the 1:1 ratio without accessories, as it should.

Another advantage over my good old 100 mm macro bellows is that it has an electric aperture drive, so it can be operated in automatic. The downside is that it has no aperture ring, hence cannot be used with extension rings.

Taking dimensions for the new slide holder

The dimensions of the new lens don’t allow any part of the previous device to be reused. To dimension the new one, I have to take note of certain values, such as the lens diameter, its distance from the subject according to the desired magnification, etc. I therefore put a ruler on the table and view it through the lens.

Display of the width on the rear screen of the camera:

Measuring the object-to-lens distance with a caliper:

I take a reading for a few different distances. The width W is the value visible on the rear screen; the distance D1 is the distance between the subject and the front end of the lens; the distance D2 is the distance between the subject and the focus ring. This is because the front end of the lens moves during focusing: the slide mount moves with it and must not be allowed to bump into the focus ring.

Diameter of the front part of the lens: 54.6 mm.

W (mm)	D1 (mm)	D2 (mm)
25	24	55
36.5	40	65
45.5	53.5	74

I deduce from these readings that, to allow some latitude in shooting, the holder body will need to be about 50 mm long.

As I plan to fit it directly on the lens, it will need an internal diameter close to 54.6, say 54.8 to 55 to have a minimum clearance.

New holder drawing

The new 3D-printed slide holder

The object arrived after a week. The first negative impression was that, although the whole thing showed the ridges I expected, the surface of the most important part, the plate that holds the slide or film, was a real ploughed field, the edges were in sharp relief, the rectangular opening was collapsed in the centre.

On reflection, this is not surprising, depending on the manufacturer’s choice of orientation of the object to be printed. I think they placed it with the square part down. Therefore, the plate was overhanging. I don’t know if any supports were provided, but the result is clearly bad. Perhaps I should have planned to print in two separate parts: the cylinder and the square, even if it meant gluing. It would have been necessary to use a centring system based on slots, which would have the same problem of overhang, but on a much smaller surface, so probably less annoying.

Machining the plate

Here is how the surface looks after milling it partially to a depth of 0.4 mm: you can still see several cracks! In fact, even after removing a tenth of a millimetre more, not all the defects will be removed.

Click to enlarge. The rough part is at the bottom of the overview, but at the top of the enlargement. You can clearly see the defect, not too annoying in fact, on the rectangular opening.

After milling, the surface still needs to be sanded, but it will not be in a really satisfactory state.

Mounting the tube on the lens

A more optimistic note: the dimensions are well respected, and the tube fits with friction (rather rough!) on the lens barrel. There is a slight clearance (less than 5/10 mm) of the slide in the holder. The same applies to the filmstrip.

Holding a slide in the holder

I used “hair clips” which are very popular with brass kit builders, and which can be found at low cost on the web. One of the branches is cut, the other is half cut and then bent. The fixing, through the existing hole, is made with M2 screws (I had planed holes in the plate, which I just had to tap, with caution given the reputation of fragility of PLA). The remaining branch is covered by a heat-shrink tube.

Intact clamp in the background.

Here is the positioning test of a slide after mounting the clamps.

I notice that the spring of these clamps is a bit too strong: inserting a slide is a bit hard. I also need to find a solution to keep these clamps open during the placement of a filmstrip.

Second attempt at 3D printing

To say the least, the above 3D-print is not really satisfactory. The probable reason for this was given above. I make a new drawing, but this time in two parts: the rectangular plate on the one hand, and the tube on the other.

When ordering from All3DP, I was pleasantly surprised to find a much cheaper offer than the first time (€11.63 instead of €20.68 all-inclusive), and moreover in France, with reduced shipping costs and delivery time. I have to say that I chose the standard finish, not “sanded”, because, considering the finish obtained the first time, I had a doubt…

Order received…

The first impression is rather good, in any case better than the previous time: the stripes are present, but finer, and the roughness is much less strong.

… but disappointing

Note: None of the dimensions respect the initial design. Of course, this is a matter of a few tenths of a millimetre, always less, but this causes several problems:

• The inner diameter of the tube, calculated to fit over the lens with enough clearance to interpose a velvet flock adhesive, is too small for this.
• The width of the film pass is too narrow, it does not fit.
• And above all, tongue and slot assembly is impossible, the slot being too narrow and the tongue too wide!

Here is a table comparing the theoretical dimensions and those of the object received.

Note: for the diameters, the dimension used is an average of three or four values taken.

Dimension	Drawing	Object	Comment
Slide path width	50.1	49.92	Just passes
Film path width	35.1	34.82	Don’t pass
Window width	37.0	36.6	Not a problem
Window height	25.0	24.5	Not a problem
Slot, ext. dia	58.8	58.3	Don’t fit
Tongue, ext. dia	58.7	58.6
Slot, int. dia	54.7	55.0
Tongue, int. dia	54.8	54.6
Slot, width	2.05	1.5
Tongue, width	1.95	2.1
Slot, depth	1.0	0.6
Tongue, depth	0.9	0.7

Maybe I’m naive to have chosen such small clearances…

Another defect: the marking groove on the top of the tube has caused a sort of raised bead inside, which hinders even more the fitting on the lens.

Illustrations of dimensional defects

Slide path, expected dimension 50.1.

Film path, expected dimension 35.1.

Slot outer diameter, expected dimension 58.8.

Outer diameter of tongue, expected dimension 58.7. This value is larger than the slot, so it is not possible to interlock.

Necessary machining

Let’s start with the easiest one: taking one or two tenths off the film strip guides. This can be done by light milling.

More difficult: the reworking of the slot. I will widen it with a ⌀ 2.5 milling bit. The difficulty is to make it go round a circle, without the centre of this circle being materialized. So, I have to make a centring piece, a 36.6 × 24.5 rectangular block, to the dimensions of the window. The screw screwed into the martyr of the cross table will act as a pivot.

Electrician’s tape was added to eliminate the residual clearance. This photo was taken after machining.

Machining the slot: simply rotate the workpiece slowly with the milling bit engaged. Although used at minimum speed, the bit jams, fortunately without damage, but removing the melted PLA coating around the bit is difficult! Many burrs remain, even after three passes with no change in depth. Long live PLA!

Compensating for the clearance of the tube on the lens

As already mentioned, the small diameter of the tube does not allow for lining the inside of the tube with a velvet flock adhesive. However, there is still a gap which I compensate with a 0.13 mm layer of Evergreen, fixed with thin double-sided adhesive.

It can be seen that the (transparent) Evergreen strip stops at the bead unduly created by the 3D printing. The resulting joint is clearance-free, but a little too hard.

New holding clamps

I don’t want to use the single hole in the clamps to fix them to the plate anymore because they can rotate and risk touching the inserted slide. I therefore drill two ⌀ 1.5 holes spaced 5 mm apart. They will be fixed with two 1.4 × 8 P-tite screws.

Assembling the two parts

After careful — and laborious — deburring, the gluing is done with a two-component epoxy glue.

Assembly of the clamps and testing

Click for a closer look at the clamp.

Testing the placement of a thick slide frame:

Holding the clamps open for placing a filmstrip

I finally came up with a solution that is a bit DIY, but easy to implement. These are 2.5 × 4 section wedges, made of recycled Plexiglas. They are attached to the plate by a thread (nylon or other) to avoid losing them.

Conclusion

Despite its attractive price (at least for PLA), FDM 3D printing (by depositing molten filament) does not seem to be suitable for functional parts requiring some adjustment. The experience was interesting, but I would not have spent more time making my slide holder from scratch.

First tests

Here is a photo dating from 1985, connoisseurs will guess where it was taken… The scan was made by pointing at a white screen lit by a 5000 K LED lamp. Aperture F/5.6, exposure time 1/25 s, ISO sensitivity 100. The colour temperature set on the camera is 10,000 K, as there is a strong blue cast in the photo, not totally eliminated here.

Click to enlarge. The film grain is clearly visible in the enlargement (without compression).

The magnification ratio is about 0.66. The slide is almost full frame (you can see the edge defects). Keep in mind that a 35 mm lens frames in APS-C like a 52 mm lens in “full frame”.

Here is the result after a few quick interventions on the image.