Precision hole drilling

This is where autocad is a great tool. Draw it and print full scale. check paper against part.

From there, you have all your holes centers and diameters. use a center drill or spot drill (my favorite) then drill each point. A CNC machine is best , but a DRO works too. Do your milling at the same time.

Karl

P.S. Find me a deal on a Leblond servo shift and I'll make the parts for you.

For all of the holes, you should be spotting them with the short flex-free center drill to get started on center without wander. You also need to plunge the center drill down solidly to start which also helps ensure it doesn't wander. After center drilling switch to regular drill bits, but work your way up to the final hole size in a few steps if it's a larger hole. If you really want a good hole, you drill up to a size a few thou short of the final size and then ream to final size.

DRO ??

Bob Swinney

The material is brass (it was a great idea to use brass).

I calulated all hole positions using a spreadsheet. See "Encoder Mounts" sheet here:

To fit it over a 17mm motor shaft, I had to drill a 43/64 hole. (17.065mm).

I do not know of any way how I could drill that hole precisely, with a huge (comparatively) MT3 drill bit, as any drill bit would wander away at least somewhat, I think.

So what I did was, I drilled the 17mm hole aproximately where it should be. Then, with DRO, I located the center of the hole and then moved to what should be a point (0,0), based on the calculated position of the center of the hole. (kind of a backwards thinking process).

For smaller holes, all I did was start them with a center drill and then drill with a drill bit.

It actually seems to have worked, as the encoder works just fine.

When making that mounting plate, I realized that there is a lot to precision drilling, and want to ask now if anyone knows tricks for drilling precise holes. I have two more plates to make.

The required precision for locating encoder base is 0.01". (which is not that bad).

On a related note: to bolt the base to the motor, I had to drill four holes for the mounting bolts. How can I precisely measure the distance between holes. I tried using a caliper and it worked, obviously, but I can not be totally sure how accurate I was. I would measure the distance between two points in holes closest to one another, then farthest, and average the two. But it felt that there was a lot of wiggle room in those measurements. SEM has a manual for the motors in question and it specifies the distance, but based on what I drilled, the distance is slightly wrong. (the motors were made 20 years ago)., I was lucky that I drilled the holes slightly oversize. For the next pair of plates, I would really like to drill to-size holes in the right place.

i i

"Robert Swinney" fired this volley in news:FuydnWl8U6257qHRnZ2dnUVZ snipped-for-privacy@giganews.com:

Drill relocation option.

Special tool on his mill.

(Digital Read-Out)

LLoyd

I have a DRO. The issue is, if I move to point (1.2345, 6.5432), and plung the drill down, will the hole really be in thast point or somewhere close due to drill wandering.

i

The hole will be at that point it you use a suitable center or spotting drill (short and stiff), and plunge fast enough that it bites properly before having and opportunity to wander. You should also be using a fairly high spindle speed for the small drills.

Please just drill them close to size and then bore them. You must have a boring head.

JC

Ignoramus3537 fired this volley in news:vcudncRmQPR66qHRnZ2dnUVZ snipped-for-privacy@giganews.com:

Iggy, you understand he doesn't mean plunge the 43/64" drill. He was talking about a short, rigid spotting drill.

Unless you've got lost motion in your system, it should be close enough; WAY tighter than your spec on those mounts.

It cannot hurt, after dialing to your location, to lock both ways and just barely gently snug up the quill lock. Re-check that the way locks didn't move the table, and that the quill lock didn't push the quill out of line. Unless your machine is quite worn, it shouldn't.

I personally wouldn't drill such a hole in that thin stock. I'd drill/ream or bore it to finished size. A trick I'm partial to is to mount a boring head in the tailstock, and use the head's vernier to adjust the hole size. There's no difference in the finished job between doing that or just using a boring tool on the post, but it is faster and more convenient for me to do it that way if I don't already have a boring tool holder in the toolpost. (I have a homebrew indexable post that is anything BUT "quick-change").

LLoyd

"Lloyd E. Sponenburgh" fired this volley in news:Xns9DB46E36368AFlloydspmindspringcom@216.168.3.70:

It might have been appropriate to say, "when I'm drilling in the lathe..."

Duh!

LLoyd

Karl,

I disagree. If he has the money for the CNC, then he should buy a multi-spindle drilling head from those AutoDrill guys and do it in one stroke. That way, the head manufacturer is 100% liable if the pattern comes out wrong. A drill bushing plate could help with the oversized issue too... But to create that plate is more work than the single part I presume he needed.

As John said, the way that accurately-located and -sized holes are made is to drill, then bore. You won't get accurate location or size with a drill bit -- depending on what you mean by "accurate."

You can drill and then ream for good size and roundness, but that won't give you location. Or drill close to location and then bore for size, roundness,

*and* location.

This is what toolmaker's buttons were all about, and the basic idea has been the way to produce accurate holes for over 100 years.

I lock the table and drill mostly to size with cheap or resharpened drill bits, which are more fast than accurate, then swap to the boring head and take a light cut for concentricity. Finally, since my mill doesn't have much headroom, I drill to size with a good S&D bit in a collet. So far hole size and location has been good to 0.002" or better.

These are better than I expected:

They are only 3" long beyond the collet so you can mill and drill at the same head or table height setting. I bought them locally, that site is the first one that Google returned.

I once worked for a very nice, but very large, guy named Vito who inspired a lot of similar comments.

jsw

Well, you need a CNC machine to do that! The problem is that without accurate measuring of table travel, it is hard to do. You can use a boring head to bore the central hole, and any centering boss or recess at your (0,0) coordinate. but, then you need to spot and drill 4 holes at the corners. A machine with worn Acme screws and no DRO will give you fits as there is differential wear on the screws, so even relative movements are wrong. (My old manual Bridgeport had this disease, nothing would ever fit no matter how careful I was.)

Yoiu can try scribing marks with calipers, or even surface plates and height gauges, but it all comes down to how accurate you can dial in the drill bit over the mark. I was never very good at this.

Jon

laser printers have gotten a LOT better since the early ones, but they are still very inaccurate. If you can get .050" accuracy over the whole sheet of paper, I'd be amazed. I'd NEVER try to do precision work this way. print a document twice, once with mirror printing on, and hold the two up to a light to see how well they align.

Jon

OK. With the DRO, it is doable to get well-centered holes. For the central hole and any locating bosses, use a boring head and boring tools. You can set the exact diameter needed, the hole will be aligned very accurately. For the bolt holes that will likely have a little slop anyway, use a center drill first, then change to the jobber's length drill and drill to size. it will be quite well centered, probably within .003" or better of the table's position. The bored hole will be within .001".

Jon

Yes, I think that I will bore the next two large holes. Good idea.

While we are at it, an unrelated question: can EMC be configured with just one axis? (for testing). I want to test as much as possible with one axis (limits, estop, homing etc), so that things are well debugged. Then I would move to the remaining two axes. The question is does EMC work with only one axis?

thanks

i

Something that may make life a bit easier for you are "helical couplers"

This is one of my clients..and the source for my helical couplers used between motor and ballscrews, encoder and shafts and whatnot.

Tell em the OmniTurn guy recommened them.

Gunner

One could not be a successful Leftwinger without realizing that, in contrast to the popular conception supported by newspapers and mothers of Leftwingers, a goodly number of Leftwingers are not only narrow-minded and dull, but also just stupid. Gunner Asch

With some indicators set up you should be able to figure out what your backlash is like and get to within .01 easy enough. Might be a pain in the ass moving the indicator every time your close to travel limits, but it would at least get you there.

Ignoramus3537 fired this volley in news:C4WdnRLeE89uXaHRnZ2dnUVZ snipped-for-privacy@giganews.com:

The snide answer is, "It'll work with anything you compile it for."

I haven't started on my EMC^2 conversion of my BOSS, so I don't know for sure, but I think you can "kill" any axis independently as the package is currently configured. Don't you have to re-compile for a specific custom hardware setup, anyway?

LLoyd