Did a few more experiments in the last few days with PR48, my net findings were:
- Reducing Z layer height doesn't reduce exposure requirements much (or perhaps at all) - 15s @ 100XY/50Z results in similar under-formed features as 18s @ 100XYZ
- A good quality multi-coated but non-UV coated lens protecting filter (http://www.amazon.com/gp/product/B0009K9Q1A) can protect the projector lens opening w/o impacting exposure times or creating optical artifacts
- Fine features (e.g. 1mm square pillars, threads) need at least 20s per layer exposure at 100XY/50Z
Here’s the baseline 18s @ 100XYZ print from last time:
Note that the fine features failed to fully form at this resolution, but the support pillars are all distinct (you can see the grid of dots through the part on the top).
Here’s 15s @ 100XY/50Z:
With a 16% reduction in exposure and half the layer height, the pillars and body form, but the fine top features form even more poorly.
Here’s 15s @ 100XY/50Z with the Hoya 58mm Digital MC filter:
In this case, the fine features seem to actually form slightly better (especially the left half-torus); this is probably a random fluke but demonstrates that the filter doesn’t seem to increase exposure time (e.g. minimal light blocked).
Here’s 20s @ 100XY/50Z (also with the filter, I’m standardizing at leaving it on now):
The small features all form now, and the threads look much better and fully thread-like. 1 of the 4 threads is now large enough that it engages an M3 nut, although the other 3 are undersized and the nut simply falls through. The pillars still form cleanly without being absorbed into a body enlarged by light bleed; you can see the individual grid of attachment points through the top of the part.