Why the drill template doesn't align at 90 degrees

Fixture Tables
62

I’ve been reading through the discussions about the template not lining up when rotated 90 degrees and it seems like most of the problem is tolerance stacking from indexing hole to hole, so why not just rotate the jig 90 degrees first.

Instead of chasing the grid around the table I think there is a simpler way to run the template that keeps the error from building up.

Start by running the template sideways down the long edge of the plate using the edge dowels after carefully locating the first corner hole. The long edge does need to be straight for this to work. That establishes one long straight reference row across the entire plate. Because you are referencing the plate edge and only moving the template a couple times, there is almost no tolerance stack.

Once that long row is drilled, switch directions and start filling the plate.

When moving up the plate all you need to do is locate the template off the hole in that long reference row with a pin and verify the template orientation with a good square to the edge then make sure to clamp the far end down tight.

Since the reference row runs the entire length of the table, every new section is tied back to the same baseline instead of being indexed off a chain of previous holes.

That means the template might only get moved two or three times along the long edge, and the rest of the grid grows from a known straight row instead of accumulating error across multiple moves.

It seems like this would eliminate most of the tolerance stack people are seeing when they try to rotate the template and chase the grid across the table.

63

Many times, Jerry Garcia sang

“Once in a while
You get shown the light
In the strangest of places
If you look at it right”

Justin might just have shown that light! ;-}

–Larry

1 Like
64

We built the frame for the table without thinking we’d ever drill it for fixturing. There is 2”x3” tubing on 16” centers for support so we had to skip holes. 7 holes is what it ended up being between the supports. We bridged across the supports to maintain the pattern. We used a reference line across the front and it maintained the entire width. We are super happy with the table, so far we’re about a 1000 holes in and still have two rows to go!

65

Definitely a good idea, and to improve on it, you could also overlap a few of the holes already drilled, in order to help ensure that the “reference row” is as straight as possible. My table is small enough that I would’ve felt comfortable using a 6’ steel straight edge, had I used your method, but some people are making some big ass tables, 5 x 10, and likely even larger, so overlapping the prior row by 2 or 3 holes is just extra insurance.

The big takeaway is the use of tighter fitting pins to be used when locating the template; if you don’t, then you’re going to have to measure and adjust the template EVERY SINGLE TIME you move it, which is kinda self defeating. But yeah, you definitely have a great idea, I wish I would have thought of that, AND realized the pin tolerance issue, before I started. But I’m pretty sure that just about everyone that’s bought and used one of these drilling templates has discovered the stacking issue, unfortunately the hard way, as we watched the video, and went with the same method used in the video.

It’s the SOLE reason I even bothered to make a thread about it in the first place. If the comment in the video stating that the template could be used as a fixture had been worded differently, or a few tight fitting pins were also used in conjunction with the tacking bolts in the video, then NOBODY would have a leg to stand on.

I’m sure someone will disagree, and for them, I’ll put it like this:

Simply blaming poor results on the end user, while at the same time NOT including the appropriate parts to reliably and repeatedly locate the template, BOTH in the kit, AND in the instruction video, is NO DIFFERENT than blaming the ironworkers building an all steel bridge, providing them with TOP QUALITY beams, accurately drilled holes, and then giving them shit quality bolts to fasten the whole thing together. “Give the carpenter the finest wood, the best new tools, and crappy nails”, basically.

Anyway, yeah, you pretty much nailed it, without having to change the design of the template. Great idea. Wouldn’t hurt to use 2 tight pins and overlap a few holes though…..:smiling_face_with_sunglasses:

66

Hey Mike;

How bout I just toss a piece of 1” plate on the Bridgeport, clamp ‘er down and mill X and Y, then flip, indicate it in, and mill the third side at 45*? Then, I can use the DRO to accurately make all the proper sized holes in the correct position and orientation?:thinking::exploding_head::rofl::rofl:

Seriously though, that’s a fantastic idea; I thought of an “L” shaped template, but your idea is waaay better. I actually looked into making my own template, but there just isn’t an affordable source for quality bushings. I’m sure you could chance it and score some chineseium variety, but that’s a hard no from me. Absolutely love the idea though. But man, that’s gonna be a heavy SOB; you’d have to include band aids and some sort of heavy duty protective gloves along with that thing :rofl:

I read your comment yesterday, but apparently the words took a while to travel to penetrate :man_shrugging:t2:. In any case, still a great idea, good job man!

67

NO, IT’S NOT 45 DEGREES! iT’S SOME WIERD ANGLE! The concept depends on the Pythagorean Theorem, the observation that Asquared + Bsquared = Csquared with A and B being the legs and C being the hypotenuse of any right triangle. Cool that some small numbers 3, 4, and 5 (9, 16, and 25) work out. So set up on the two end holes… I ‘m sure I don’t need to tell you it’s best to traverse the 5 holes using a single feed direction. But maybe with the precision of CNC feeds? I don’t know, it would be interesting to compare results. But interpolation to mill a huge circle or thread a huge bore is pretty impressive…

On the weight of the triangle- Consider cutting out the middle for weight saving? Only leave enough “meat” to support the bushings? Maybe not even straight sided on the inside (But that profile would be a bitch)?

Let’s talk pins- First, will a drilled or cored hole ever be bigger than the bushing the cutter fit through to make the hole? Just a tiny chance- A drill will wander and will scrape out a diameter just a tiny bit (well sometimes not so tiny) bigger than the drill. Center drilling minimizes this, and I suspect annular cutters wander very little. But there’s still the micro mess and tearing that takes place right where the carbide digs into steel. The moral of that story is pushing a dull cutter degrades both finish AND geometry. Give the alignment pins the best holes you can muster.

OK, The pins themselves- It is easier to have one pair of pins both align the drill fixture and hold it down once it is located. That’s two different functions. To fulfill both with the same pins compromises are made, generally through undersizing pins/oversizing holes. The “tacking bolts” kind of walk this back by expanding the three balls at the bottom. How tolerant of variations in the stack thickness is this clamping system? How well do they deal with burrs on the back side? The CIRCLE that the balls engage may lead the process astray. We’re looking for a “Locational Clearance Fit: (Designed to provide positional accuracy without significant motion, allowing parts to be easily assembled and disassembled) e.g., H7/h6. So what are the size tolerances on the bushing bores and the (ground?) rod? (Yeah, chinisium bushings might be a bad idea at any price… But priced at $6-$15 a pop (x12), bushings aren’t cheap). Back to the pins…

A proper pin must be smaller than the bushing BUT if it has a precise conical end it could be bigger than the drilled/cored hole and still locate the template. It would be centering on the CIRCLE (rather than the CYLINDER though. Thats why deburring can make or break accuracy here). But then what holds the template down to the plate? Well look at all those extra holes… How about some bolts? And to make that bit of the process easier consider a block with two threaded holes (either 2 or 4 inches between centers) so you don’t have to fumble with your nuts as much. (Said with a straight face mind you). The really fancy version adjusts in length (think two nuts captive in a U channel) so you can loosen both, remove one, swing to the next location, thread in the first, remove the second, swing the block, LOCATE the template, thread in the second, and torque the bolts.

Now I’m even boring myself ZZZZZZZ

68

Whoa dude, you need to switch to decaf, for sure! :rofl::rofl::rofl:

69

Good thinking! I’d be a bit worried about the straightness of the reference edge (and maybe the “squareness” of the overall plate. It may cost more to get precision cut plate but sometimes plates are cut using really precise processes just cause that tool is where the plate is. CHECK before you drill! I’ve used a V block and a laser pointer to find problems with straightness over 20-30’ lengths…
But were you thinking of moving the template a fraction of its length and pinning two of the just drilled holes? That would be locating on the holes, would need the template to be located really straight to the just-drilled holes, and would be independent of the plate edge. (Hmmmm, there are 12 Template holes, if you pin #T1 and #T11 into #P2 and #P12 and then drill one new hole (#P13?) it will likely be more precisely located to #T1-T12 than if you pin closer together and drill more than 1 before resetting… Hmmm. But if you pin one just-drilled hole and bump the edge you’re depending on that edge to be precisely straight. Personally I’d strive for precision in the grid rather than any kind of precision to a (potentially impresice) edge. One could lay a straight edge off of a row of pins/holes, scribe a line, and grind to it if needed. How straight and located to the grid do you need it to be?

How about skirts around the sides (5 sided table)? Hmm, if you plan them in advance there’s the possibility of having 2 or 3 rows of holes the length of your table made on somebody’s big mill. The loose plates (with machine accuracy) could be used as a reference… But without bushings it is probably a bad idea to drill repeatedly through those holes. But you get to compare table flatness to skirt edge straightness!! (Better to know sooner than to find out later…)

Anyone with any ideas on how to locate skirt plates with accurate hole grids to edges of tables with hole grids? The edges matter less than the holes…

70

So if you didn’t skip you’d have a couple of rows through the plate AND the frame tube below? What would be wrong with that? Or would the holes fall along an edge of the tube?

The sheer size of your table also brings to mind whether the objective is to 1) have holes accurate enough to set weldment dimensions or 2) you expect to shim to dimension as measured, or 3) cut really accurate parts and depend on them for weldment dimensions, using the fixture table only to hold parts AFTER locating them some other way.

And do you make boat rails in Bellingham WA?

71

The main objective of our frame table is to be able to clamp items down at various locations. We square our parts up using other methods although so far the layout is pretty spot on.

We skipped the sections because it was easier than risking hitting an edge on the supports. The layout tool bridges across the skipped sections without issue so we are pretty happy.

My table isn’t perfect but for the cost of the plate, drill guide and 3 annular cutters we’re pretty happy with it!

72

I was thinking about the issues related to the hole jig & the accumulated error due to fit tolerances. I think the best solution is to as accurately a possible locate the center of the table. Using a scribe & mark up the table from the center. This requires you to “guess” where to place your first row of holes but in reality it doesn’t matter except aesthetics if the holes were slightly off in the center. You would only be making 2 or possibly 4 shorter rows of holes & your tolerance error would never rise to become an issue. If you didn’t locate your first row of holes exactly on X or Y it only means that you holes may not follow the edge of the table in parallel.

I hope this makes sense, but starting in the middle & working outward might be a better plan than to start on one edge as your datum.

Wormwood

73

@Wormwood-

I don’t mean to sound rude, but if you’ve read the entire thread, and the related threads, the solution has already been found; use precision alignment pins, and the problem solves itself, 99.9%. I’m sorry I haven’t posted up anything further, as I’ve been having some health issues, but rest assured, this issue has been beaten to death.

I’m sure micrometer Matusky will be along shortly with an overly complicated response, (no offense, Mike), but you absolutely cannot GUESS at anything when trying to build something even remotely precise; it just doesn’t work that way. The template that I used was completely within tolerance, but as further research has shown, and PROVEN, if you don’t use an accurate method (precision pins), to locate, and re-locate the template EACH AND EVERY TIME it is moved, then it will, and does have issues.

Don’t misunderstand me, I think I understand what you’re saying, but no matter the point of origin for the holes, FROM THAT POINT, they must all reference that point. Personally, I recommend to anyone trying to make a table using this template, to start with an OVERSIZED plate,(L x W), as flat as possible, and cut (preferably perfectly straight), ONE edge that will make the table top as close to perfectly square as possible. Then proceed as in the video, just make SURE to use the 2 precision diameter pins that I keep mentioning, IN ADDITION TO, the tacking bolts that are included with the template.

It really is that simple. The template is, at least the one I have, perfectly fine for the use as intended. Let me put it differently; if you tried to do the same thing, but the (whatever) method you chose to use as the basis to keep the template in alignment was (EXAGGERATION) 1/4" smaller than the intended hole size, then obviously, you would have A LOT of play (SLOP!) when you locate, and, MOST IMPORTANTLY, re-locate, the template.

I could go into a LOT more detail, using the previous paragraph as an example, but it’s late, I’m tired, and I’m sure Captain Caliper will be along shortly to (hopefully not TOO complicatedly) explain further, in the extreme…:grin:

Anyway, Welcome to the forum!

74

It was only a suggestion for you. This is your thread, so good luck to you!

Wormwood

1 Like
75

Right here MH. I’ve put some thought into the whole process and have a (potentially) patentable design for a “Tool and Method to Drill a Precise Hole Grid in Plate”. I probed for interest with Jason but haven’t heard back from him… At any rate, how about a method that doesn’t locate using pins and holes (that need clearance)? Would you consider a collaboration?

76

Mike-

You have piqued my curiosity; why not? What do you have in mind, without releasing too much information, considering it is potentially a patentable solution?

Homey