The simple new holder is made from a piece of scrap 6061 aluminum. The only slightly 'critical' dimensions are 1 through 5:
1. The large hole; it is 0.253" (6.48mm), for a slip-fit over the existing pivot bushing (which has an O.D. of 0.250").
2. The small hole; it is 1/8" (3.16mm), for a press-fit of the 1/8" stainless rod (5/8" long, super-glued and hammered lightly into the hole to protrude about 1/2" (12.7mm). This rod is where the cable wire loops over the holder to work the filter in and out of the light path.
3. The distance between the holes: center-to-center is 1.16" (29.5mm). This dimension was gotten from the original plastic holder.
4. The bar thickness; this has to be less than the length of the existing pivot bushing, so that when tightening the bushing screw, the bushing is secured, but the bar is still free to pivot.
5. The semi-circular slot (groove) for the filter glass is about 3/32" (2mm) deep, cut with a round saw or abrasive blade using a hand-held rotary tool (Dremel or equivalent, in the US), and having a width of a little more than the thickness of the filter.
The overall length of the bar is arbitrary, mine was about 1.8" (46mm). The bar width was about 0.485" (12.3mm).
I added the washer over the inner end of the bushing to keep the bar from wobbling as much.
The best way to find the position of where the groove needs to be cut is to mount the bar in place in the microscope, swing it down into position, and mark the bar with a dotted line (see image) where the OD of the field condenser diffuser glass is, behind the bar. This will give you the best place for the filter, whether 26mm or 32mm. Then remove the bar and cut your groove, glue in the filter, etc.
The groove for the filter is filled with Hi-temp RTV, the edge of the filter is cleaned and smeared with the same, and the filter is bedded into the groove, and the whole thing propped up to keep the filter planar with the bar, until the RTV hardens. Then an additional fillet of RTV is applied to the bar-filter joint for extra strength. One could even skip the making of the groove (the most problematic part of the fabrication) and glue the filter edge directly to the side of the bar, but I didn't try this.
To test the RTV, I ran the lamp for several hours with the scope baseplate screwed on, to see if the temperature would be a problem, or if the RTV would fail, etc. No problems; the RTV got harder, the temp of the bar got up to no more than 130F (oddly 300F on the diffuser glass). The RTV supposedly can take 700F. Tugging on the filter, the RTV seems strong enough, but I didn't pull it out of the groove, so am hoping for the best. In service it is under almost zero stress, so I think it should work OK over the long term.
It may seem that the bar blocks some of the light, since some of the filter is recessed into the groove, but the lamp is so close to the blue filter, that if one traces the light from the filament of the lamp to the edge of the diffuser, one can see that only the center 75% or so of the filter is used. Actually, the larger 32mm filters might be a bit better (if one can find a filter that is not too dark blue.)
To help the positioning of the filter when swiveled into place, some sort of 'stop' can help. The microscope has a pin for this purpose, underneath in the base, right under the blue-filter control knob; the 1" lever worked by the knob runs up against the pin, to control the in-positioning of the filter. I've seen some of these pins bent to adjust the stopping of the lever, but personally prefer leaving the pin straight, and instead sliding a rubber cylinder, or the like, over the pin. This way, the pin doesn't break off from being bent, and changing the position of the stop is just a matter of changing the rubber cylinder size (one could even just wrap tape around the pin until it works).
nicely made part , Bill. Looks permanent.