1. The previous "hiccup" (using an acetone-containing de-greaser spray accidentally) is sorted now. Had to replace the display on the affected modules, which was both costly and a PITA (and had to wait for the replacement modules to arrive). The new ones are on, and all is well (well, almost - please keep reading...)
2. Decided to sacrifice module S/N 1 , and improvised an accelerated aging test. Put it in for about 5 days under UV light (in my PCB etching lab) and cut it open to see the inside. Epoxy resin showed ugly yellowing but no other bad signs (cracking, shrinking etc). However, found that the resin did not reach in between the backlight and electronics modules - probably because of it's viscosity is too high. Yet, it managed to partially block the 6-pin socket (functional, but has an unpleasantly tight feeling when inserting the connector). I simply forgot to take into account that these connectors are open from the rear:
3. To solve these problems made two changes: (a) sealed the backside of the pin connector using silicone glue before casting,
and (b) ordered special UV-resistant ultra-low-viscosity epoxy resin. This one takes 2 days to cure so experimenting slowed down a bit...
4. The biggest bummer was such a stupid mistake that I could not believe myself: normally I use wooden clothespins to temporarily hold whatever needs holding, but for some reason reached for plastic/rubber padded pins this time. Not sure what chemical reaction did this, but the padding lifted the topmost layer of the paint on finished housings. Will try to salvage them by masking the display and spray painting them over again, but it will be a costly bummer (again...) Stupid me
5. But I keep on crawling ahead. Finished the wiring harnesses now, with labels added. Found that 7ft (2.1m) might give enough leeway for bigger bikes, so made them that long (instead of the earlier 5ft).
- continued -
6. I was never 100% happy with the surface evenness of the housings. No matter how long I hand-sanded them, there were always some minor imperfections remaining. Also, the epoxy mid-layer (used to cover 3D printing striations) has an ugly tendency to form shallow humps and bumps when curing, depending on the underlying surface's evenness. Decided to bite the bullet, ditch the epoxy mid-layer and invest in a precision belt sander. It is a small unit, with the belt running on the edge of it (so that it can reach into the backside of the angled mounts). The display housing is for scale:
Using 240 grit belt and some patience, I was able to reach even better surface quality. (I know, I'm a perfectionist...) :blush: The angled mount bottom side on the left is made with this new method, the one to the right is the "old" one. Note that this picture was taken in grazing light - all surface imperfections are vastly exaggerated:
While already putting parts together for mailing, started to look for packaging material. Hope to get some decent cardboard boxes - could print a nice logo on them
2. Decided to sacrifice module S/N 1 , and improvised an accelerated aging test. Put it in for about 5 days under UV light (in my PCB etching lab) and cut it open to see the inside. Epoxy resin showed ugly yellowing but no other bad signs (cracking, shrinking etc). However, found that the resin did not reach in between the backlight and electronics modules - probably because of it's viscosity is too high. Yet, it managed to partially block the 6-pin socket (functional, but has an unpleasantly tight feeling when inserting the connector). I simply forgot to take into account that these connectors are open from the rear:
3. To solve these problems made two changes: (a) sealed the backside of the pin connector using silicone glue before casting,
and (b) ordered special UV-resistant ultra-low-viscosity epoxy resin. This one takes 2 days to cure so experimenting slowed down a bit...
4. The biggest bummer was such a stupid mistake that I could not believe myself: normally I use wooden clothespins to temporarily hold whatever needs holding, but for some reason reached for plastic/rubber padded pins this time. Not sure what chemical reaction did this, but the padding lifted the topmost layer of the paint on finished housings. Will try to salvage them by masking the display and spray painting them over again, but it will be a costly bummer (again...) Stupid me
5. But I keep on crawling ahead. Finished the wiring harnesses now, with labels added. Found that 7ft (2.1m) might give enough leeway for bigger bikes, so made them that long (instead of the earlier 5ft).
- continued -
6. I was never 100% happy with the surface evenness of the housings. No matter how long I hand-sanded them, there were always some minor imperfections remaining. Also, the epoxy mid-layer (used to cover 3D printing striations) has an ugly tendency to form shallow humps and bumps when curing, depending on the underlying surface's evenness. Decided to bite the bullet, ditch the epoxy mid-layer and invest in a precision belt sander. It is a small unit, with the belt running on the edge of it (so that it can reach into the backside of the angled mounts). The display housing is for scale:
Using 240 grit belt and some patience, I was able to reach even better surface quality. (I know, I'm a perfectionist...) :blush: The angled mount bottom side on the left is made with this new method, the one to the right is the "old" one. Note that this picture was taken in grazing light - all surface imperfections are vastly exaggerated:
While already putting parts together for mailing, started to look for packaging material. Hope to get some decent cardboard boxes - could print a nice logo on them
