Oval Drill holes

[cycollins]

I know there has been some discussion on this topic, and I have read it, but I believe my problem might be a little different.

I tried 1/8" drill-holes, using the 1/8" cutting bit to various depths. I observed the so-called "Wood-Pecker dance", used by the CW unit when drill depths are greater than some threshold, presumably to clear dust. Between attacks of the cutting bit, the traction belt moves back and forth. I was assuming this was an accuracy-improving maneuver on the off chance that drilling shifts things a bit (I'm only guessing and giving the unit the benefit of the doubt). Common sense says it should have to move, but there you go. So, in the end, my drill-holes were oval shaped, with a major axis at a consistent angle to the y-axis (about 30-45 degrees). Even on the first attack, which sinks to a depth of about 1/8", I'm pretty sure the oval pattern was already established, so I DON'T think the traction-shifting caused the problem (although the fatness of my 1/4" drill holes MIGHT be due to this behavior).

One explanation for this behavior might simply be that my Quick-Release chuck might be bent relative to the z-truck. Could have happened during an unpleasant discussion I had with the machine over a stuck bit. If the bit itself were bent, you would expect the whole bit to move in a conical motion (they call it "play-out", no?). It would produce oversized holes as opposed to oval ones. But if the whole chuck were at a small angle, the bit would spin straight within the chuck and the z-truck would lower for the drill attack, but the slight angle would produce an oval hole in the direction of the bend as the entire cutting bit hit the wood slightly "sideways". Just one theorey.

On the minus side for my theorey, a deeper hole would be more oval than a shallow one. I did a 1/4" depth hole and it it wasn't noticably more oval-shaped than a .8" hole. Could just be my eyes.

So if my "bent-chuck" theorey is correct, is there anything I could do about it? Maybe remove and re-place the chuck? Is there a screw adjustment somewhere to "right" things up between the z-truck and the actual chuck-bit linkage?

Let's assume my "bent-chuck" theorey is bogus (what are the odds?). Is there a common, well-understood reason for oval drill-holes that I'm overlooking? Will "calibrating" the machine address this issue? I finally have a 3/8" double-flute, straight routing bit with a flat head, 1/2" shaft and 1 1/4" depth (don't think THAT didn't take some hunting down). If I calibrate, what are the measurement characteristics that should improve?

Any thoughts?
Cycollins

[rjustice]

Cycollins,

I have been in the CNC field for 26 years, doing pretty much everything from A-Z.. including precision machine alignment.

If your machine chuck is bent, and running out it would simply give an oversized hole. It duplicates a boring bar, and cuts on the flute of the tool that is positioned furthest from center. But it does give you a fairly straight hole.

If the bearings are loose in the spindle, it is possible it would be a tapered hole that would look slightly like a cone. Fairly round at the top, but bigger, then getting smaller as it gets to the bottom. It does this due to the bit finding center as it gets deeper, and starts guiding itself and centers up.

If the "Z" axis is out of alignment by the truck guides, it would produce a round hole that looks like an oval at the top, but is truly a round hole that is put in on an angle. You could check this by taking a drill bit that fits the hole and see if it wobbles in one direction or if it fits snug top to bottom of the hole but is on an angle when left in the wood.

If the spindle is out of alignment but the "Z" guides are straight it would give you a hole that is oval at the top but more round at the bottom. Depending on the direction it is loose at the top, would tell you which way the spindle is out of alignment to the guides.

Now with all that said, you mentioned that you think the traction belt is moving as it is pecking the bit in. That sounds very abnormal if it is, and it could duplicate the spindle out of alignment condition above. I would watch your traction belts very very closely to see if they are moving, and if not also watch the rubber cog belt inside the machine "Y" axis track to verifiy that it isnt moving either. If they are i would definately contact CW....

Hope this is helpful... if i can clarify anything let me know !!!


Ron

[Digitalwoodshop]

I have good luck making holes .175 thick for the LED's to light my signs. I was wondering if you get the oval problem every time every hole or just once in a while? If the brass roller is not making contact with every movement of the board then that could cause the oval holes. Is the oval in the length direction or the width direction? Length = brass roller problem, Width = loose drive belt on the left right.

This is what I use the drill function for. Solar 911 signs.

AL

[pkunk]

Oval holes can also be caused by gear slop in the x axis gears.

[cycollins]

I appreciate all your feedback. Starting from first to last, rjustice, my terminology may be lacking, but I think what I was suggesting was your spindle-out-of-alignment case. I was suggesting that the Quick-Release chuck which represents the spindle in the case of the CW, was out of alignment on the z-truck. However, now that you mention the entire z-truck being out of alignment, I actually did the test you suggested. I was drilling the hole to run a dowel through the wood and sure enough, the dowel was not perpendicular to the surface of the wood (by an amount that was obvious to the naked eye). I guess I'm now leaning toward your out-of-alignment-z-truck theorey. So, am I screwed? (No pun intended.) Is righting the z-truck a home-repair adjustment or is it a (gasp) send-it-to-texas kind of thing? Is it the kind of thing that the calibration will address? I'm doing two-sided projects, so having a perpendicular spindle is key to getting the registration right between the front and back.

DigitalWoodshop, rockin' projects. My ovals are neither width or length-wise ovals. They are literally at something like 30-45 degree angle to the y-axis. Wierd, huh?

Pkunk, same question to you as to rjustice: which of the maintenance procedures that have been discussed will help with "gear slop in the x-axis gears"? Home repair or back-to-texas?

cycollins

[rjustice]

Cy,

I personally havent had to get into alignment on the machine yet. It probably isnt that difficult. As you stated earlier when getting rough with it to get a bit out of your stuck chuck you could have possibly knocked it out. I seriously doubt that it is out anywhere near 30 degrees. I would bet that it looks alot worse than it is.

I would contact CW and ask them for a PDF on alignment proceedures. If you get it, it would be nice to post on here.

Also, you asked about calibration. I would consider calibration a different animal than alignment. Calibration would allow you to fix something like in designer you have a carve region that measures 4" square, and when you cut it you actually get a cut that measures 3.9 x 4.1 I believe that you can tweek the machine in to get your true 4 x 4.

Someone that has run the calibration routine please chime in on that ....

Ron

[pkunk]

Pkunk, same question to you as to rjustice: which of the maintenance procedures that have been discussed will help with "gear slop in the x-axis gears"? Home repair or back-to-texas?

cycollins
It does sound like you have other issues, but gear slop is easy to fix from under the right hand panel.

[rjustice]

My ovals are neither width or length-wise ovals. They are literally at something like 30-45 degree angle to the y-axis. Wierd, huh?

cycollins

if you put the dowel through the hole is it snug from the face of the board and the back of the board?

Also, did you carefully watch to see if either the sandpaper is moving or the "Y" axis belt is moving during the pecking cycle?

Ron

[cycollins]

...the 30 degrees was in reference to the major axis of the oval with respect to the x-axis when one view the board from on top. Here's a picture to elaborate (the holes highlighted in red are 1/4" deep, the blue lines indicate the approximate orientation). There's also a picture showing a 1/8" brass dowel passing through a completely penetrating hole, taken up against some floor tiles for reference. Finally here are two pictures indicating how far a brass dowel can slop to the left and right in a 1/4" depth hole. All of the holes are nominally 1/8", cut with the 1/8" cutting bit.

[rjustice]

Cy,

Ok your pictures really helped. But, not necessarily in a good way. The fact that the holes are leaning in random directions like this takes alot of the concern of alignment out of the picture. If it was simply out of alignment in a direction, they would repeat and all be leaning the same way.

Did you watch the traction belt or the "Y" axis cog belt to see if they are moving yet?... If they are moving while it is pecking the hole in, Pkunk could be on to something with the gear issue. It would allow the workpiece to "float" around.

If no, i would be interested in trying the same tests on a piece of MDF. This would tell us if perhaps just the hard spots in the grain of the wood is influencing the bit to lead or wander off. It has been mentioned that the cutter diameter to length ratio on the 1/8" bit is a little long, perhaps the bit just doesnt have enough rigidity to stay straight, and starts wandering

The other thing would be to grab the "Z" truck, and see if it has play in any direction. My thoughts here would be that if it is loose, it is just finding a happy spot once the bit hits the wood, and stays there till the next hole...

This is an interesting problem to say the least...

Ron

[pkunk]

If I remember correctly, when I had my gear issue, I could feel the gear slop when trying to move the belts by hand.

[cycollins]

...the traction belts feel solid as does the z-truck. That's why I thought perhaps something was bent within the z-truck assembly. I understand the objections to that theorey (the overall sloppiness of the holes, the variability of the direction of the ovals, etc.), but perhaps the theorey needs a little refinement. What if in stead of being bent, something has become "soft". That is, assuming that within the chuck assembly, there are rings and/or washers of a compressible material (say vinyl or silicon or something). Let's say that during the unpleasantness with the stuck bit, one side of chuck's internal cavities became so compressed that the materials within became permanently deformed. The washer or ring or whatever get's shifted around during normal low-speed operations, like bit insertion/removal, which accounts for the variable oval direction. Once the bit is in high speed operation, the centrifugal effects tend to lodge the abnormal geometry into a single orientation, creating a single axis of rotation out of alignment with the basic z-axis. The bottom line to this theorey is simple: get a new quick-release assembly from CW (and not from Sears as I've been warned on numerous occasions). This I know to be a home-repair type service, since there's a PDF on the procedure. Probably wishful-thinking on my part, but I AM pretty sure my x-axis gears and z-truck are currently fairly solid.

[John]

I don't recall the oval hole problem coming up since it was fixed with updated software. Be sure you are running the latest software version. If I recall correctly, with the fix, if you are making 1/8" holes with a 1/8" bit the x and y axis should not move while the hole is being drilled. I haven't done any hole drilling since the fix. There are some using the hole feature for drilling shelf pin holes. Maybe one of them could verify the x and y axis should not move when drilling a hole that matches the dia. of the bit.

[rjustice]

Need to watch the belts to see if it is moving... that would verify what John is saying... if they are, but yet they are solid while the machine is at idle, then the software is moving the part. If so, again, sounds like a malfunction. are you running the latest software?...

I am going to run some holes tonight to see what i get... you really have me curious on this. I didnt go to the latest rev yet... I am running 1.119


Ron

[cycollins]

I'm running 1.120. There was definitely "pecking" and moving during the drilling. I just calibrated the machine, and I'm literally in the middle of rerunning a drill test. I'll see if that helps.

cycollins

[Digitalwoodshop]

Another test I would do is design a + on a board and make the holes 1/8 inch in the center and one step away in every direction. Then step it up in size for another set then back to 1/8 inch and see what you get.

AL

Try this file on a scrap board.

[Charles M]

It has been mentioned that the cutter diameter to length ratio on the 1/8" bit is a little long, perhaps the bit just doesnt have enough rigidity to stay straight, and starts wandering

A solid carbide bit will not flex enough for this result without snapping first so I doubt it is caused by that. However, a loose spindle could allow the bit to follow tough grain.

[cycollins]

Thanks to Digitalworkshop for the "+" idea. I might actually use that instead of a drill hole, since the idea of the drill hole was simply to get a precise location fix that I could use for a drill press to punch through to the other side (my projects are always more than an inch in thickness and the CW won't drill through.

I agree with with what Charles M said about the carbide drill. That's why I thought perhaps the QuickRelease mechanism was loose in terms of its alignment with the z-truck and might need replacing. There are also fixtures on the z-truck to which the quick-release attach and these also might be bent or wobbly. I have a profound aversion to the idea of sending the unit back to Texas for service, but I suppose it could be necessary or at least desirable.

cycollins

[rjustice]

Between the chuck, the bearings, and the bit, I think it is possible to get a lead off condition... I have seen it happen in steel.. In theory i think you are correct "carbide doesnt flex"... but again with all the variables I have seen this happen

Ron


A solid carbide bit will not flex enough for this result without snapping first so I doubt it is caused by that. However, a loose spindle could allow the bit to follow tough grain.

[cycollins]

I did another test recently and got very confusing results. I routed a rectangular path that was a little over 10" in length by a little over 4" in width. It used the 1/8" cutting bit and was 1/8" wide and 1/8" deep. It worked pretty well. The lengthwise spans of the rectangle were only slightly wider than 1/8" (~.381"). The width-wise spans were only slightly fatter (~.401"). With all the problems I was having with drill holes I was expecting much worse (although I suppose it is a ~10% difference). Anyway, I could repeat this project over and over and the results were very consistent. The drill holes on the other hand were a bit random. Finally, I overlayed a project with 1/8" drill holes at the corners. Same peculiar hole behaviors.

Then, I did a much more compact project. Precisely 6" by 4" with the same kind of routed lines and drill holes at the corner. I added some more routed lines in a circle and some crosses and a diagonal line. All the endpoints were on a .25" grid. This time, the horizontal and diagonal lines were very similar in size and (and this is the wierd part) the drill holes were pretty circular, so much so that I have trouble seeing the eliptical shape with the unaided eye. So what's the rule here? Put things on nice rational boundaries and get more precise behavior? Wierd. Maybe it was the scale of the project. I don't know. Maybe I don't need to know. I only use drill holes for reference marks (and maybe ultimately for peg holes on a sled). As long as there IS a way to get good holes, maybe I don't care. No, it's not true. This kind of thing drives me nuts. I need some sense to using this tool or I'll never feel like I can control it. Heavy sigh.

Cycollins

[cycollins]

BTW, those who said that there shouldn't be any movement of the traction belt were correct. There wasn't any. There IS the pecking behavior if the hole is deeper than about .1", but as most people assume, I think this is a dust-removal or bit-sticking-avoidance technique. There is not any movement in the x-axis direction in between pecks of the drill bit.

However, I've been using 1.12 software for quite a while now and I'm sure I've seen x-axis shifting during drilling recently.

There may be a problem with the 1/8" bit itself (or maybe the way it's mounted in the adapter). I did the same routing test I mentioned in the last post with the 3/8" straight bit. Rock solid. However, it doesn't let me drill 3/8" diameter holes with the 3/8" bit. In stead, it makes a spiral cut with the 1/8" bit to create a 3/8" drill hole, and it does a pretty good job (pretty round, but a little bigger than .375" - closer to .39").

Unless someone wants to try some deep 1/8" drill holes with the 1/8" cutting bit and then tell me otherwise, I think the conclusion is this: it doesn't work very well. It may be just some kind of limitation of the machine. Anyone want to try?

BTW, when you're routing along a path in a project, the CW machine lets you override the bit joice when the routing begins. It offers a long list of bit-choices, some of which aren't listed in the documenation. For example, I've seen references to a 1/8" carving bit as opposed to the 1/8" cutting bit. Come to think of it, the Designer software offers you interesting choices, like 1/16" ball-nose and a 1/2" straight bit that also aren't listed as accessories in the documentation or on the sales list. Obviously such things are obtainable, but it seems strange that they would offer such choices when the CW folks tend to go on a bit about how you should only use CW-approved bits. I'd love to use a 1/4" straight bit to do 1/4" holes, but oddly enough, it isn't among the bits listed. Strange inconsistencies. My guess is that the new software will come out at the end of this year and be so much better than the current Designer that this will all be moot.

cycollins

[Charles M]

I have successfully (at least to my standards) used the drill function many times. For birdhouses I typically drill a 1-1/4 to 1-1/2" hole and and a 1/2" hole for a dowel. My experience is that they have all been pretty darn accurate. The 0.500" holes have measured 0.503 to 0.508" and seem round. A 1/2" bit would have been faster but router bits aren't designed to drill. Ramping in the way the CW does is the best way. On the larger holes it will start in the middle and make concentric circles until it reaches the full diameter of the hole and makes the last pass very slowly for better finish.

[liquidguitars]

There may be a problem with the 1/8" bit itself (or maybe the way it's mounted in the adapter).

You should be able to see the run out in the bits shaft, look for extra motion
blurring on the bit shank when running, I am on my 2 chuck and change my bit adaptors when needed "look for marks and dings on the socket".
sears has new ones you can buy if needed. I have dif 1/4 bits, just make sure the DOC is the same amount as your design your working on.

Ron thanks for the run out check list its really helpful.

LG

[cycollins]

The 1/4" adapter for my 1/8" cutting bit DOES look like it fell into a garbage disposal (a lot of the black finish is gone on the upper part of the adapter). The same is true for the carving bit to a lesser extent. You think I should swap them out for new ones? I don't feel or see any obvious "play" in either one. I wonder how long it would take to get a new chuck from CW. I don't think Sears sells those.

cycollins

[liquidguitars]

I have a backup chuck if needed always a good idea and I think sears will be selling them. Regarding the bit and the adapters better to just order a few new ones and install them with new cutters. So yea you can replace them.

LG

[cycollins]

... because after looking VERY closely at the 1/4" adapter for my 1/8" cutting bit, I noticed something not-quite-right about the plate-like structure on the bottom (as it hangs from the chuck). It would be difficult to describe the shape of the deformation except to say that it was slightly-but-visibly bent into a potato-chip shape. It's pretty slight. I measured the distance between the plate and the pins that engage the side of the chuck. The distance between each pin and the plate should have been identical, but in stead it differred by about 7/1000". It doesn't sound like much, but it's getting on for a 1/100" and in the 1/16" adapter, the distances were identical out to four decimal places. I might not be THE problem, but I bet its A problem. I'll experment some

[rjustice]

Cy,

Consider that the pins sticking out really shouldnt have anything to do with the runout of the bit... the bit adapter should seat only on the taper, and there are ball bearings that seat into the groove that pull it tight into the taper. The pins are simply drilled and tapped so they arent realy intended on being extremely accurate. they simply act as the driving mechanism to twist the tool.

I am curious as to whether the wear that you see around the tapers is farily even, or if it is to one side.... if one of the pins is interfering, i would think it would have a shiny spot on it, and the worn parts on the taper would be opposite the pin....

Ron

[liquidguitars]

Ron,

I have a few 1/4 that have uneven ware and ball bearing dings along the side,
as a side note , the Sears brand OEM? bit adaptors have a little better finish including the new chuck I just installed.

LG

[rjustice]

lg,

Are the dings inside the little groove, or on the tapered area?

I also bought a spare chuck to have just in case... and just playing with it outside of the machine, it operates considerably harder than the one in my machine. I noticed that the tapers on the bit adapters from sears look nicer too, but when measuring them in a precision comparitor, they do measure the same as the ones from carvewright.

Ron

[liquidguitars]

Ron,
inside top of the groove, could be a heat treatment issue, will be off to sears for 1/4 replacents soon. The chuck shoud lock when pushed up, but
mine never did.

LG

[rjustice]

LG,

You know, the chuck that came with my machine has never locked either. The one i purchased for backup does though.

Ron

[cycollins]

This has proved the undoing of me. I tried pkunk's suggestion (blasting it with hair dryer - presumably to expand the metal and soften the loc-tight), but it won't budge. Have you discovered some magic procedure? Just to be clear, I'm trying to unscrew the pins on the side with an alan wrench, but they're as permanently sealed as they're supposed to be. Is it easier to simply pull the bit out the bottom of the adapter?

I actually considered using the metal-cutting head for my Dremel tool to saw off the old adapter (by cutting it in two halves on either side of the bit), but if the loc-tight is between the shank and the adapter, that won't help either. Add to that the fact that I would almost certainly damage the shank of the bit in such a drastic piece of surgery. Any thoughts?

cycollins

[rjustice]

Cy,

Are you talking about removing the bit from the adapter? If so, your not going to hurt the carbide bit if you get it a little hotter than what a hairdryer will get it... if you have a grill lighter, or simply a cigarrette lighter hold the shank of the bit with a small pair of vise grips if necessary, and hold the flame right at the threads of the setscrews. What you are trying to accomplish is breaking down the locktite with heat. Once it is hot enough the setscrews really arent that tight. should be able to take the allenwrench, while holding the vicegrips and back them out pretty easily. Be careful as not to get burnt.


Ron

[cycollins]

I was looking forward to using the case-hardened-steel-cutting abrasive disks on my Dremel tool. I'm kind of pathetic that way. I have a rediculously large set of Dremel attachments because a.) I sadly equate having more tools with being handier and b.) I didn't know what half of them were for and I didn't know when I might need them (like now!). I've only every used a few of the cutting attachments, a couple of the buffing bits and some sanders. But I bet I'm the only 44 year-old kid on my block that has both the router table AND the drill-press attachment. Betcha I am. I'll give the heat treatment another try, but then it's the cutters. I want to see some sparks fly!

cycollins

[liquidguitars]

I am using a press after using the heat gun if that helps.. CW has a quick change bit holder, it is way safer to work with the bits.

LG

[rjustice]

After getting the set screws loose, it is normal to have to tap the tool out by using a punch. I had one that was stuck pretty good, and yet another I pulled out by hand.

Ron

[timmcbatts]

When doing New Custom Home Builders Brisbane (https://www.lionaconstructions.com.au/services/new-custom-homes/) it is best to hire proper contractors.