I am studying watchmaking via the British Horological Institute's Distance Learning Program. The practical lessons of the first year involve a lot of filing and precise measurement. I've made some hand levers, a small try square from gauge plate, and a hardened scraper. I scored well on those assignments. But now I'm working on the next assignment and I want to squeeze out some improvement on my accuracy when filing squares.
The assignment requires filing a piece of 1.5mm brass into a piece 1.2mm thick, 25mm wide, 30mm long.
I dealt with the thickness by filing it on a cork pad -- this kept it level with the file, and then I finished it page of sandpaper. I measured the accuracy using my micrometer. It was averaged to 1.20mm, with only 0.02 deviations, so I was happy enough.
The edges, 25x30mm, I feel like there is another tier of accuracy I could get to, but my current techniques don't achieve it.
Here's a range of the tools I employ:
My approximate technique is:
1. Decide on master edge, file it roughly flat.
2. Hold it against my try square edge to see how flat it is.
3. Try to eliminate the high spots using either (a) needle file (b) my stainless stick with some sandpaper glued to it (c) or drag the entire face across a sheet of sandpaper. The needle file seems to round the edge, as the file flexes a bit. If I drag (c) the piece, I have to do it unidirectionally (pull it perpindicular towards me), otherwise when i push it away from my i build up a slightly different angle). The "sandpaper stick" does ok for small patches, but it's tedious to reapply fresh sandpaper after it degrades.
4. Recheck the edge flatness with the try square, repeating step 3-4-3-4 until it seems "straight".
5. Then I start doing the same process with the adjacent ends, using the try square taking data from the "master edge".
6. Finally wrap up the opposite side of the master edge using one of the other sides as the datum for the try square. [this seems most dodgy -- accumulated square errors as i work from master to side to the oppositeMaster]
The results? It's pretty darn square. It's just that the try squares never are light tight and the light crack results are very hard to repeat. I even think that micro grit messes up the reading. The try square has 1mm thick blades, and the base is a heavy ground block. I bought a preisser bevel square with narrow, unnicked edges and i have trouble getting consistent light-tightness measurements using that too though.
I feel like the last few percent perfection on this is limited by me: 1) not using the try square or bevel square well enough (are their tricks tips to using them?) 2) my last bit of micro-finishing work not quite fine enough. (but I think any micro-finishing effort is doomed if i am not exactly where the high spot is, so perhaps issue #1 is most important to deal with first?
Any advice on how to get closer-to-light-tight accuracy and squareness on tasks like this?
Thanks
The assignment requires filing a piece of 1.5mm brass into a piece 1.2mm thick, 25mm wide, 30mm long.
I dealt with the thickness by filing it on a cork pad -- this kept it level with the file, and then I finished it page of sandpaper. I measured the accuracy using my micrometer. It was averaged to 1.20mm, with only 0.02 deviations, so I was happy enough.
The edges, 25x30mm, I feel like there is another tier of accuracy I could get to, but my current techniques don't achieve it.
Here's a range of the tools I employ:
My approximate technique is:
1. Decide on master edge, file it roughly flat.
2. Hold it against my try square edge to see how flat it is.
3. Try to eliminate the high spots using either (a) needle file (b) my stainless stick with some sandpaper glued to it (c) or drag the entire face across a sheet of sandpaper. The needle file seems to round the edge, as the file flexes a bit. If I drag (c) the piece, I have to do it unidirectionally (pull it perpindicular towards me), otherwise when i push it away from my i build up a slightly different angle). The "sandpaper stick" does ok for small patches, but it's tedious to reapply fresh sandpaper after it degrades.
4. Recheck the edge flatness with the try square, repeating step 3-4-3-4 until it seems "straight".
5. Then I start doing the same process with the adjacent ends, using the try square taking data from the "master edge".
6. Finally wrap up the opposite side of the master edge using one of the other sides as the datum for the try square. [this seems most dodgy -- accumulated square errors as i work from master to side to the oppositeMaster]
The results? It's pretty darn square. It's just that the try squares never are light tight and the light crack results are very hard to repeat. I even think that micro grit messes up the reading. The try square has 1mm thick blades, and the base is a heavy ground block. I bought a preisser bevel square with narrow, unnicked edges and i have trouble getting consistent light-tightness measurements using that too though.
I feel like the last few percent perfection on this is limited by me: 1) not using the try square or bevel square well enough (are their tricks tips to using them?) 2) my last bit of micro-finishing work not quite fine enough. (but I think any micro-finishing effort is doomed if i am not exactly where the high spot is, so perhaps issue #1 is most important to deal with first?
Any advice on how to get closer-to-light-tight accuracy and squareness on tasks like this?
Thanks