Preserving Autofocus on Modded DSLR's

[beeboman82]

Hi all,

I know this is an older thread, but I'm hoping it is still useful for me to add another vote that an extra half turn (6 "points") clockwise/inward appeared just right to regain autofocus.  This was on a T4i/650D testing with the EF 50 mm at f/1.8, which I think should be a pretty good test.  My LPF-2 was also 0.6 mm thick.  Initially I only returned the screws to their original positions for telescope use, but got the itch to use lenses and needed to get infinity focus back (haven't tested yet, but I'm assuming that if I got autofocus right then infinity focus should be possible).  Wish I hadn't followed Gary Honis' suggestion to super glue the screw positions at that point...

Many thanks to Kevin and Michael for doing the legwork!

Now I need to find out the adjustment for my Nikon D5300.  I removed the single filter that one has (full spectrum, no replacement glass) and measured this filter at 0.75 mm thickness.  IF, it has the same 0.5 mm thread pitch as the Canons, then I guess that would be 7.5 "points" clockwise.  Unfortunately, it is harder to find parts information for Nikons than for Canons.  I have the part number for the D5300 sensor screw as 10C8J, but just can't find any details.

Brian

[beeboman82]

Just worked on my Nikon D5300.  I confirmed the 0.5 mm pitch by visual inspection (2 threads per mm).  For the 0.75 mm thick filter, I first tried about 7.5 points clockwise.  As others have noted, consider my actual adjustments as only approximate.  Autofocus was better, but definitely still too close.  Since the empirical adjustment for a 0.6 mm thick Canon filter wound up about 6/5 = 1.2x larger than calculated, I considered trying a 1.2x factor on the 7.5 points up to 9 points.  This is now doubling up on 1.2x factors, so not mathematically satisfying.  I don't have a very fast auto Nikon lens, but best I could tell, 9 points (3/4 of a turn clockwise) got pretty close, but was maybe focusing just a tad too close still (probably good enough for my current lenses).  Unfortunately, I couldn't physically turn 2 of the 3 sensor screws past 9 points anyway, so I think I will live with 9 points on each.  I'm hoping I at least got enough for infinity focus, as that was my true goal with autofocus as nice to have.

This has me thinking, if one considers autofocus as a must in a modification, that might be the best argument I can think of for installing some sort of replacement filter in front of the sensor.  Otherwise, the price of these filters still seems steep if one is working with the budget of self-modifying an older/used camera.

I can share some focus pictures if there is interest.

I also may eventually try LPF-2 removal on my Canon 6D.  Will post if I get to it...hoping it is the 0.6 mm = 6 points of the T4i/650D.

Brian

[SpaceTinkerer]

This is an old thread but I believe I can contribute.  I have converted over 200 Canon and Nikon cameras for astrophotography and IR.

The greatest danger is sensor tilt. In cameras with torx screws and springs, if one screw is misaligned you get sensor tilt, meaning stars will not focus correctly from one side.

A safe approach for astro or full spectrum clear glass conversions: Mark the exact screw positions with a thin marker. Count screw turns so you can return at the exact same position. Measure internal filter thickness and replace them with an equally thick filter or clear glass.

The safest approach: Contact me and I can do it for you. 🙂

spacetinkerer.com/shop/

Cheers,

Chris

[michael8554]

I've converted many Canon DSLRs, and used to measure the sensor-to-mounting plate distance at the three Torx screws with a digital micrometer rig.

Now I mark the screws and mounting plate, and record how many Torx "points" it takes to fully tighten each one.

On reassembly, fully tighten, then back off the required amount.

Michael

Edited March 19 by michael8554