Well I think I found out why my parts are not cutting right.

[knight_toolworks]

I did a bunch of test cuts to see what was going on. I cut at 1ips and 3ips with a 1/8" 1/4" and 3/8" bits up and down cut.
all the cuts came out undersized. about .02 too small. I just made rectangles of 2" size and cut outside cuts on them. the x cut sides were say 1.96 and the y side of the cuts may be 1.98.
even 1/8" and only 1ips and 1/4" deep cut was off .02 on both axis's.
I just replaced the pinions a few weeks ago. that helped with some of the problems i was having. but still parts are coming out too small.
so I checked for any play in the y carriage and x. there is a tiny bit of play in the y axis. I can move it just a bit and it springs back. all the v rollers are snug and I can't turn them by hand.
when I pushed on the carriage in the x axis and if I push in the middle it seems solid but if I push on each end one side moves with about 1/3 the effort as the other end. it snaps back in place but it does not take much effort to move it. now that has me worried that it takes so little effort to move it. and I have no idea on the y everything seems snug. and that tiny bit of play may be ok.

[Gary Campbell]

Steve...
I assume you are cutting conventional direction. What happens if you cut in the climb mill direction?
Gary

[knight_toolworks]

a climb cut gives accurate or oversized parts depending on the speed of cut. 1 ips or so will get me accurate cuts 3ips will give me about .01 to .02 larger cuts.

[Gary Campbell]

Steve...
I get the same amount of under/over sized parts, but not till I get around 6ips. Thats why I prefer to cut 1st pass climb, with an onionskin and second pass conventional thru. The climb first pass puts the combined deflection into the waste and the second pass, due to little load, cuts right on.

We are experiencing similar results as you are, but have only tested with a 3/8 bit. Since my parts are square edged, I am assuming this is the result of combined tolerances, bearing and/or gear slack, as there are a bunch of moving parts between the table and the bit. I will be replacing all the pinions to see if this will eliminate a portion of it.
Gary

[mmccue29]

just a thought have you checked the size of your bit?

[knight_toolworks]

I wish it was bit size. but most of my bits are within a few thousandths.

[knight_toolworks]

seems the pressure is about the same on the x today. when I looked at the test cuts it seems the larger the bit the smaller the part becomes.
the strange thing is the machine goes back to zero every time.

[Gary Campbell]

Steve...
The larger the bit the more lateral force it develops when cutting. This results in greater machine deflection.

When you go back to zero there is no cutting force on the bit.
Gary

[knight_toolworks]

yes it is some flex if I do a cut with offset then a cleanup pass it is right on. so where is the slop? like I said a tiny bit in the y but very little and I can't find any in the x. new pinion gears all around.

[Gary Campbell]

Steve its a combination of many things, and I assume each is a little different on each individual machine.

I am going to attack each and every motion point in the next few days with a dial indicator. I suspect after reading posts from the last few years, that the PRT gantry has more inherent flex than the PRS, but I dont think the PRS is immune. I also assume that as they are used, and especially if used hard, any machine with metal parts can "soften" and when you start to add a few thousandths slack here and there the tolerance in the finish parts seems as tho they multiplied.

Our job is to eliminate any mechanical components that are not within tight tolerances. Pinions, V roller bearings on the X, Y & Z are my places to look. I have 2 years on our machine and its time to freshen it up.
Gary

[knight_toolworks]

good idea on the dial indicator. I find about .005 play in the y axis with pressure from 2 fingers. after that it gets far harder to push. at most 001. trying to move the y in the x direction.

[daski]

Bit deflection is your problem. We have the same thing here and we just input a different size for each bit. a 250 bit is usually input as .258-.267. We do a test cut and adjust accordingly. All that matters is what the final cut is so testing and measuring when it's not cutting may help, but won't solve the problems. You need to find out what it is doing under load then compensate for THAT.

[Gary Campbell]

Dave...
To properly diagnose this problem, you must be aware that if the bit is deflecting, the cut edges of the part will be at an angle. It is impossible for the bit to deflect and not be at an angle to the part. These angles may not be easy to measure, but are noticable.

IF the part edges are square, the problem is not with the bit, it is somewhere in the moving parts of the machine.

Even tho the workarounds may be the same for both problems, (bit diamter trickery or reduced speeds) to properly diagnose and fix the problem there are, of course, two different courses of action. I prefer a fix to a workaround, but its not the easiest path.
Gary

[ron brown]

You are chasing a moving target. As Gary stated there are a LOT of things that can and will change with time. I am not sure metal softens from use, but bearings and gears do wear and it may be useful to try test cuts in a different part of the table.

IF your parts require real tight tolerances, you may be using the wrong tool or just pushing the tool too hard. Many problems solve themselves with slower feed and proper chip load.

Some bit deflection problems can be cured by simply using a larger diameter bit.

Ron





Ron

[Gary Campbell]

Ron...
Nicely stated... chasing a moving target. My machine has 29 roller wheels and 5 rack and pinion pairs. a couple thousandths wear on each and cut accuracy and quality goes out the window. Thats why maintenence is so important.

You are right, IF the bit is deflecting, installing a larger bit should cure most, if not all of it. In our case a larger bit actually increases the error at the same speed. That make me lean to either slack in motion parts or deflection of structural parts, something that I am hard pressed to believe is happening on our PRSalpha. Time to start replacing parts.
Gary

[knight_toolworks]

I can't find where the slop is except the .005 in the y. everything else seems pretty solid.

[wmcghee]

I know you said you changed your pinions, but it you may need to check the set screws in them again. I tightened mine about a month ago, and the last few days my machine would not hold a tolerence. I pushed on my x and y and the farthest x side did have play in it. I checked the pinion screws and they were loose. And the y pinion was also loose, but not enough to move when I pushed on it. But it still affected the cut. Just something you may want to check again.

[knight_toolworks]

I will but I tightened them pretty well and the problem was still there right after the new install. when I removed the old ones they were still tight too.

[Gary Campbell]

Steve...
How are you testing for the "slop"? I use a 1/2" drill rod in the spindle, and then, using a small fishing scale hooked to the rod, pull 25-30# in the X and Y direction and see what is moving using a dial indicator.
Gary

[knight_toolworks]

your method was better I used a dial Gage and magnetic base and pushed with a couple of fingers. I was wanting a scale but did not have one.

[bleeth]

Well, Gary is one of those Florida Key's guys so he tends to see things a little "fishy"!!!

[knight_toolworks]

I checked the ply in the y in different places finally and there is less when it is over to the left the least used part. so I guess it is time for a new rack. Hope I don't have to replace the x racks that would cost big time.

[bleeth]

Steve: Check the y rails themselves. It is more likely they are worn or bowed than the rack causing the issue. I found that on mine some time ago and solved it by putting hardened rails on top and shimming them in flat and level.

[knight_toolworks]

I had looked at them and they seem in better shape then the x rails. my machine only has 700 hours or so on it. the only play I found was lineal along the y. I think my lube did not work and the racks were dry too much.

[knight_toolworks]

ok I finally found a way to see if the motor is turning when I push to see the slop. a pair of visegrips slightly clamped to the pinion showed me the motor is turning.
here is a video of it. the movement is overestimated as there is only about .01 or less movement but it lets you see the movement.
http://s154.photobucket.com/albums/s266/knighttoolworks/posting/?action=view&current=movement.flv

[knight_toolworks]

I just checked the x axises and they both have that slight give. so maybe I am full of it? that this is normal?