| Gallery, Projects and General > How do I?? |
| Accurate Drilling - centering microscope, USB cam, optical punch? |
| << < (6/8) > >> |
| Lew_Merrick_PE:
--- Quote from: Jasonb on June 26, 2011, 12:30:25 PM ---Lew surely the whole problem that Chris has using a depthing gauge is that he does not know how far to wind in the dial or read off the DRO so there is little point in establishing a zero. At best the hole will be located a set distance from one edge and at its intersection with an arc, at worse it will be at the intersection of two arcs so you can't use co-ordinates to position the hole. As for round holes when clock making isn't that what they make broaches for, bit like a reamer but produces a round hole in sheet material. --- End quote --- Jason, The positions you cite (distance from one edge & intersection with an arc or an intersection of two arcs) is geometry that can be calculated in terms of X and Y coordinate values. Yes, there are many times when transferring geometry is simpler, but it is rarely more accurate -- and accuracy is stated as a concern. I admit that I have never made a clock. Travel and test dial indicators, yes. Clocks, no. Finding the coordinates of an intersection between an arc and either another arc or a line is the base application for the solution to a Quadratic equation that we all had to memorize at some point in school. We do it quickly and painlessly today using CAD software. ??? [/quote] Edit, is this the type of drill bit? [/quote] Those are similar, but they are ground for drilling (burning might be a better descriptions) holes in hardened steel. If you found one in the proper size, you could regrind the point. The sheetmetal drill bits of my youth were fairly common -- and I spent more time than rational trying to find a source for them a few years back. I ended up making a small set of them for a client and they were thrilled. . Lew |
| kwackers:
I can't believe people are seriously considering replacing a perfectly simple tool that can obtain the ideal mesh between two gears and transfer those points to a plate with as close to perfect accuracy as it's possible to get with a system that still requires the tool, suitably modified to enable the distance between the two points to be accurately determined and then the numbers crunched in a CAD package before finally being entered into the DRO. Why make things so complex? |
| Miner:
I've certainly learned far more about just how clockmakers lay out the holes than I knew before. Since I obviously still know very little about this my comments could be a total waste of space on this thread. It seems to me that the inaccuracys are mostly due to the prick or center punching as your depending on feel and normal eyesight to get those punch marks aligned at the true center of your layout lines. If the punch marks are even slightly out then the drill will drop into that punch mark and you then have a misaligned hole. I'd delete the punch marking completely since there's no real way to get consistant and accurate results. As people have mentioned, Those spindle mounted web cams would be your best bet IF you can get a consistant alignment with the cameras cross hairs and the spindle C/L. IMO a dedicated web cam mount like a end mill holder that uses set screws to retain the end mill and fits the mills spindle would be a priority. If your spindle allows rotation of that web cam/end mill holder like a mill with a M/T internal taper would, Then marking both the spindle o.d. and the holder so they always go back into the same position would be a nessisity. Those high end German made 3-D Tasters use the same idea to get accurate and consistant results. The old school clock makers didn't use anything like a web cam and everything was done by hand and eye. But there's no doubt if they had what we do today they would have dumped the hand methods in a second. Other than your absolute need and the requirements to use a depthing tool, For myself I've given up on center punching work that requires accuracy and use the coordinate system. You then start from a referance point. The old school (non cnc) super accurate Moore jig boring and grinding machines used that system, So if it was good enough for them............. Pete |
| John Hill:
--- Quote from: kwackers on June 27, 2011, 06:50:22 AM --- I don't follow, where did the paper drawing come from? Did you draw it yourself? If so how's that easier than simply marking out using a depthing gauge? I think people are getting carried away here, marking out clock wheels using a depthing gauge and then drilling the points isn't a big issue. It doesn't need to be super accurate on a per hole basis but it does need to allow for accumulated error. The system of using a depthing gauge and drawing arcs fits those parameters perfectly. As soon as you start measuring and calculating you accumulate errors and risk the whole thing going pear shaped. --- End quote --- You need some idea of the layout of the gear train unless you have designed the clock entirely in your head. Here is my suggested process in some detail. 1. sketch layout of gear train. 2. use depthing gauge to determine distances between drill points and mark these on your sketch. 3. drill first hole. 4. refer to sketch for distance and direction to next hole. 5. use DRO to position for next hole. 6. drill 7. rinse, lather and repeat. Note that each position is determined relative to the one before therefore errors do not accumulate. |
| PekkaNF:
Really great thread. Like this formum. People from different backgrounds and disiplines tell how AND why they use the method and how it works from them. I never build a clock and have found it hard to understand WHY they do gear train the way they do, but I think it's dawning to me now. In principle, it's not that fundamentaly different from how gear train is build on banjo before thread cutting... IMHO if centre punching is used, it needs practice. Least for me: I need small and sharp 60 decree punch to find the centre of two scribed scores, held perfectly straight up, then gentle tap and checked with magnifier. Haven't noticed difference between 60/90 decree punch marks when I have used NC-spotting drills, twis drills seems to need 90 decree punch, although I have found out that too big "crater" for the drill size is not a way to go. Does anybody use a conical wiggler or sticky needle to center spindle to centre punch? How small holes we are talkking on this thread? I have't found smaller than 3 mm spot drills, but years ago we used PC-board drills to make 0.5 - 1.2 mm holes. They were britle and run very hight rpm, but worked then fine on good drill press. Pekka |
| Navigation |
| Message Index |
| Next page |
| Previous page |