CNC Dowel Hole Drilling Accuracy

Pros help a programmer troubleshoot inconsistencies in dowel hole placement. October 1, 2005

I've recently started working for a company, programming for their Busellato. I don't have any experience on this particular machine (only on Thermwood CNC router).

They are having a problem with their dowels lining up - decks not flush with gables. Is anyone out there using a Busellato for dowel construction? Is it precise? It was suggested to set the decks and tops so that they are approximately 1/8 from the edge, but this is not acceptable for the type of projects we are working on (all commercial casework). It was also suggested to cut the parts oversized, then let the machine cut them more precise, then do the doweling, but this is adding an extra step. (We do not have a beamsaw.)

I have a few theories as to why it's not working, one being that the drill heads were crashed and damaged, then fixed and calibrated. The parts can be reversed to use the other drill heads, but some of our parts are too long. We have adjusted the heads on the machine, which works at first, then you try it again later, and it's off again. Also inconsistent material thicknesses, etc.

Any tips would be greatly appreciated. We're looking at having our supplier check out the previously damaged heads again.

Forum Responses
(CNC Forum)
From contributor D:
The machine is very accurate. You have several other things going on that might cause a problem.

Material thickness problems: always make sure the dowel joint is accurately referenced to the outside of the case. If it is not, the variations will show up as misaligned dowel joints.

Tooling: if you have an accurate drill, the hole is going to be perfect; if not, the drill will wander, or the hole will be inaccurate in size.

Support: be sure the panel is supported close to the edge, or the drilling operation could deflect it, causing a misaligned dowel joint.

Vacuum pods on the support benches: make sure there is an accurate support under the panel and that the vacuum is pulling the panel all the way down flat.

You should test drill parts to find out the cause of the problem. Then it will be easier to solve.

From the original questioner:
Thank you. Never thought to reference from the edge of the material. Something so simple! I've been using half material thickness for the code.

From contributor M:
If you reference from the face of a part, make sure you put that face down on the machine to get an accurate registration for the hole. Otherwise you will still suffer from variations in material thickness.

Aside from that... trying to do flush construction (which is what is sounds like you are doing) is the most difficult type of construction, especially with dowels, because they are so unforgiving. If you use small reveals (1/8" or less), it works very well, and the reveals absorb and hide subtle variations.

From contributor C:
You mentioned you don't have a beamsaw. Are your sides cut accurately enough? Is the problem uniform and consistent? We had a little issue like this with our wall cabinets. Older beamsaw. We invested in a new, more productive, more accurate saw and the problem went away because x and y were more accurate from part to part.

From contributor D:
When you reference from the edge of the part, if you are using a parametric variable, like T/2 for the location, that is good for production where you might make the part out of a variety of (specified) thickness materials, all consistent, but it assumes they are all the same - the face drill program is not checking the actual thickness of the top rail, for instance.

As pointed out, you need to keep track of the reference face from horizontal drilling of a part to assembly. It is too easy to reverse or flip a part if you are using a centered dowel pattern and a T/2 center in the edge. One solution is to use small reveals (1/8" or less) as suggested, and don't worry about it. Another is to offset the drill pattern in an obvious way that the assembly only goes together.

You wrote: "It was also suggested to cut the parts oversized, then let the machine cut them more precise, then do the doweling, but this is adding an extra step. (We do not have a beamsaw)."

I think what someone might have suggested is to let the Busellato machine the part using a milling past, prior to any of the drilling operations. This is not adding another step - the part is on the machine one time, but to guarantee the part is accurate and square is something you must add to your evaluation. How many poor parts produced does it take to offset the value of another line of code?

You should have some additional training on the machine programming. It is a simple machine, for sure, but Delmac has a good training program if you get a hold of them.

From contributor K:
Could be many causes, or a combination of causes. If a horizontal head is misaligned or not square to the machine axis, then you will notice that the distance will be off + on one side of the panel and - the opposite face. You can compromise by finding the correct parameters for each side of the drill output and tweaking them until you get better results. There is a work offset for each tool on Busellatos that you can alter from original spec to represent their current position from machine zero. Another thing to consider is the proper alignment of the pins along the x axis. You can check alignment with a dial indicator. I don't quite remember the setup, but there is a nice adjustment mechanism on the pins for moving them in and out. Misaligned pins will cause the panel to be out of square and not line up properly with the end panels. I have been there, done that, bought the t-shirt and already cut the sleeves off. Trust me, the better method is to have a standard boring pattern for all parts referencing from the front and outside of the case and drilling all horizontal operations on a line boring machine. It is faster and more accurate than drilling with a point 2 point.

From contributor R:
The Busellato is a great machine, but very touchy. Every calibration for all your horizontal drill bits must be dead on. Before that, all your pin and field calibrations have to be dead on. Then, and this is the hardest to control, all the pods have to be the exact same height on both sides of the machine. If all that is correct, the pods have to hold the pieces firmly. On smaller pieces with only two pods, it is going to want to slip. The Busellato is great for vertical drilling and machining, but I would suggest a horizontal drill machine with dowel insertion. I believe Delmac would love to sell you one.

The comments below were added after this Forum discussion was archived as a Knowledge Base article (add your comment).

Comment from contributor O:
I run a Jet 200 and have encountered the same issue. For me it came down to tooling. When nesting parts, and most other times, tool diameter is very important. My nesting software has a setting for tool diameter. It lays everything out according to that value. If my diameter in the software is not the same as the actual diameter of the bit being used, tolerances are off.

I make a simple calibration tool-pass to check my bit length and diameter of cut. Donít trust the bit manufacturer, check it yourself with calipers. A 1/2 inch compression bit may not be 12.7mm it may actually be 12.5 or 12.4. If the actual diameter is 3/10mm less, then your parts will be 6/10 off in finished width. The machine does not know until you tell it. I also have lost 2/10mm just from use. So check your bits occasionally.