In the days when I used a Cokin P square filter system I tried a few polarising filters, and ended up with one made by Stealth Gear which gave me pretty decent results. Unfortunately, since moving up from 84mm (Cokin P) to 100mm to cope with some of my wider lenses I’ve been having to make do without a polariser. A lot of the 100mm versions seem to be tied to a particular filter holder and probably won’t work with my humble Sioti unit. They’re also very expensive …
One of the options I’d been thinking about for a while was a bit of recycling, in the form of pressing the Stealth Gear polariser back into service – it would need some sort of housing to hold it in a 100mm filter holder, but since it has a roughly 73mm diameter glass area it should work with quite a lot of my lenses, even on a full frame body.
I had thought about making something up with layers of plastic sheet (as used in the flare filter I made last year), but stalled at the idea of making a wheel that would engage with the ‘teeth’ on the outside of the Stealth Gear filter. That would need its own set of accurately hand cut teeth, and while I’m not too bad with a modelling knife I think making the wheel would be a serious (if not impossible) challenge.
However, fast fowarding to now, with a 3D printer desparately wanting to make itself useful it seems that this might be a good time to design something and see how (or if) it works.
As ever the parts were designed in Fusion 360: there are three of them – the body back and front plates and the control wheel. The drawing below gives a reasonable idea of how they look:
Note that the toothed filter is included in the drawing for clarity: it was used to design and position the control wheel – it is not intended to be printed.
The back and front body plates have matching recesses to fit the filter and the control wheel. I made the control wheel recess on the front a bit deeper than the filter recess so that the wheel can be a bit thicker than the filter – this makes it a lot more comfortable to turn.
While I could have designed the control wheel with a shaft either side to fit into holes in the front and back plates, I opted to use a countersunk bolt and captive nut to hold the wheel in place, as I think these will be stronger. This might be unnecessary, but seemed like a good idea at the time. The stl files for the three parts can be found here.
The front and back plates of the filter body were printed using Matte Forge matte black PLA: print settings were head temperature 210° and bed temperature 50° (52° for the first layer).
To make it contrast with the body of the holder, the control wheel was printed using Technology Outlet blue PLA: this was printed with a head temperature of 205° and bed temperature of 48°.
This has to be one of the least challenging assembly jobs ever, as including the filter there are only four parts.
- Filter was placed in the recess in the back plate,
- Control wheel was placed in its recess in the back plate,
- Front plate was attached using four 6mm M2 countersunk screws,
- Control wheel was secured using a 6mm M2.5 countersunk bolt, with a captive nut held in the recess on the front plate. I took care not to overtighten the bolt, as there’s not a lot of thickness to the plastic.
That’s about it! The completed filter looks like this:
One minor flaw with the design is that as you rotate the control wheel finger pressure pushes the filter down in the slot in the 100mm filter holder. To prevent this I’ve added a small stop to the back plate – this rests on top of the guide on the filter holder to keep it in place, and also ensures that the filter is centred when it is slid into the holder. The stop isn’t included in the stl file, as it is just a small screw fixed into a hole drilled in the side of the back plate: its position is dependent on the length of the guide slots on the filter holder.
As far as performance is concerned, the adapter was tesetd on a full frame camera using my Canon EF 20-35mm f3.5-4.5 lens My first tests, using the back slot of the filter holder, show a bit of clipping in the corners of the frame at 20mm but it appears to be all good for 24mm and beyond. I’ll be doing some more testing to confirm this, but I’m pretty happy if it’s usable down to 24mm, as I’m not likely to want to use a polariser with a wider focal length.