Squareness problems GETTING DESPERATE!

[srd1]

I'm having a problem with my router that so far I'm not able to solve. I'll give as much information as I can and hopefully someone may have run across this. The machine is a Shop Sabre 4860 with a Colombo 3 HP 24000 RPM spindle. I am cutting 5/16 baltic birch (butter) and am using a few different styles of 1/8" bits. My problem is this. My cuts have to be at exactly 90 degrees into the wood. I run a test pattern to test the cuts. The test pattern is like a plus sign with a 4" line cutting away from me, a 4" line cutting towards me (on the same plane), a 4" line cutting from left to right, and a 4" line cutting from right to left (on the same plane). When I look down the lines, the ones that are on the same plane should be exactly in line with each other. BUT, what is happening, is every line is going out of square to the left. It will cut through the wood at an angle always to the left. I have ruled out the bits because they are used for this exact application by over 200 other businesses. I put a 1/4 bit in and ran the same test and it cut perfect. Also, if the spindle was out of square with the machine, the lines on the same plane would be out of square in the same direction no matter which way it was cutting. The faster the speed rate is, the worse the problem is. I ran the test at 100 IPM @ 24000 RPM and the lines actually curved like a banana. They curved so much that something has to be moving because the bit would have broke if it was bit deflection. BUT with the 1/4" bit at 200 IPM, it was perfectly straight. If something in the Z axis was loose or flexing, the same cut at twice the speed with a 1/4" bit would have put more pressure than the other bits at 1 10th the speed. The only thing I can seem to narrow it down to is maybe the 1/8" collet? Like I said EVERY style of bit I put in using the 1/8" collet does this exact same thing. It will even do it at a speed rate as low as 25 IPM and these bits are good for at LEAST 100 IPM @ 24000 RPM with less than .008 deflection. And I'm getting something like .032 @ even 25 to 30 IPM. But ONLY with the 1/8 bits. Any help with this would be GREATLY appreciated!!
Thanks!
Greg

[ger21]

Sounds like bit deflection to me. Are you cutting the 5/16" in one pass? What if you cut it in 2-3 passes?

Any chance of posting a pic?

[srd1]

Hi Gerry,
Thanks for the reply.
These bits are made specifically for this application and if they deflected to the point of the problem that I'm having, they plain old wouldn't work for the job. Is it possible for 200 customers to use this bit successfully at double the speeds I'm running it at and get 1 10th the deflection that I'm getting? I don't know all the variables that could cause it. Also, there is less than 1/2" of the bit below the collet. It's a tapered bit and even at the top of the cut it was off by at least 1/32 at the worst point. So that would mean the bit would have had to deflect 1/32" at it's widest point "1/8" at less than 1/4" from the collet. These are some of the reasons I have ruled out (but not completely) bit deflection. But I am certainly NOT a professional by any means and am completely open to any possibility. Is it possible to get this much deflection even at the top of the cut? Also, what would cause this type of deflection on my machine but other shops are not having any problem with them?

[lerman]

Have you tried using a carbide bit? It should have much less deflection than HSS.

Ken

[ger21]

I guess you need to try some new, good quality collets then.

[nwrepair1]

If your cutting 5/16" in one pass, that's 2.5 times the diameter of the cutter. The bit is too small and your feed is too fast. Who told you 200 other customers are using it this way? Why do you need to use so small a diameter?

[maxmachine]

It sound to me like your entire machine or gantry is deflecting. The bit would break with that much deflection. Loose rails on one of the axis? Spindle mount with loose screws? All kinds of things could cause this.

Mark

[ddgman2001]

Put a dial indicator on the bit and push it with your finger.

Move the dial up to the collet and push on the bit again with your finger.

This will tell you how much the bit is deflecting vs the spindle.

[ZipSnipe]

Sounds like deflection and loose axis guides.

[LeeWay]

It does sound like something loosing precision from the spindle down. That is pretty thick to cut in one pass with such a small bit as well. I cut 3/16" lexan in two passes with a 3/16 carbide bit. It will full cut it in one, but the cut is much nicer in two. I am feeding about 150 IPM with a PC router. I really have to slow it down on tight arcs or they don't come out right. I run those at 80 IPM.
If there are any adjustment points in the Z axis, I would look into those.
One way to see any slack or movement is to chuck up a center drill. Slow the jog rate down to about 10% and jog down into some scrap wood a little. You should be able to see any slack in the Z this way. Up and down a few times might help. Not deep, mind you, but just enough to take the spindle weight off.
If you have any movement there, then you have to fix it.

[ImanCarrot]

Are you climb cutting or conventional cutting? would be interesting to see if you experienced the same problem if you changed from one to the other...

[Denny J]

Originally Posted by ImanCarrot Are you climb cutting or conventional cutting? would be interesting to see if you experienced the same problem if you changed from one to the other...

I was thinking the same thing