Dimensions off between file and actual cuts

[BobA]

Papa Bear,

Was each circle and square sizes off the same amount?

My thoughts are if you have selected profile cuts outside the circles and squares you should get the dimension you program, ex 1, 2, 3, etc.

If you selected on the line your dimensions would be size minus diameter of tool, ex .75, 1.75, 2.75, etc. (using .25 diameter tool).

If you selected inside the line your dimensions would be size minus 2xdiameter of tool, ex .5, 1.5, 2.5, etc. (using .25 diameter tool).

Can it be that simple?

Bob

[jgrove255]

Sorry to hear that you're having problems papabear. I've got the same setup as yourself (HD & Bosch), but haven't ran into the dimensional problem ... However, I would check your router clamp as BillK suggested. I don't know what nextwave was thinking, but using HDPE was a poor design choice to use as a clamping material. It has a low coefficient of friction and has an interesting property, along with PTFE (Teflon), that with increasing loads the coefficient of friction declines. That's why HDPE and PTFE are often used for low friction bearings. As for me, I came out into my workshop one day and noticed mine was cracked near the clamping bolt, so I ended up making a new one out of MDF. So far I haven't had any issues with the router slipping or going cock-eyed anymore.

As far as your problem with the terracing, your tool-head is not perpendicular to your work piece. Again, this is most likely a clamping issue; the router maybe slightly askew. I would suggest removing the router and the clamp and giving it and the clamp a good cleaning before you try to reposition the router in the clamp. If that doesn't get rid of the terracing, try removing the top-half of the clamp and see if there is any debris that might have gotten in between the top and bottom half of the clamps.

Good luck!

James Grove

[4DThinker]

There are (at least) two places dimension shift can come from using the Shark. One from loose set screws on the couplers at each motor, and the other from a worn T-Nut on the threaded rods. Both will let the router shift a bit when it changes direction, or feeds a bit too fast along a line against tough material. I can move the router mount box manually left and right about 3/64", and see the threaded rod shift within that T-Nut. I've contacted NWA for replacement T-Nuts but so far have only gotten a link to their support store which does not list T-Nuts.

[wolffie]

Go via the very cheap solution to the set screws and replace them with capped allen key screws.
They are much easier to position and do not work loose so easily.
(I got the tip from Rungemach who is an engineer)
Cheers
Wolffie

[iampapabear]

Its not set screws. They are TIGHT.
I don't think its worn out, too close to new...less than 25 total hours on the whole machine.
The router is tight in the mount, no cracks.
It appears to be perpendicular, based on a large bit skim cutting a spoil board and using rediculously large stepover.

The concentric circles all start and stop at the same point of arc. They are all off at the same point and in the same way.
They are not even what could be called properly "circles", merely mostly circular.

Until I find out what is going on....either my expectations are too high, or the machine is "misbehaving" for some unexpected reason, i have no intention of spending any more money on it.

I would ask, has anyone else cut a series of concentric circles and squares on their ProPlusHD machines?

If so, would you please post a pic. I would love to compare what I am getting with what others are doing.

I started with 1 inch circles and progressed upwards in 1 inch increments.
I cut them with a 60 degree bit at 0.05 depth.
I cut them at 100 Inches per minute and another series at 50 inches per minute.

At the slower IPM rate the circles were better, but still off too much for what i THINK this machine should be capable of.

NWA says they can drive it at 200 IPM. Running it at 25% of that should pose no problems at all, but I get lousy cuts doing just that. I cut them in MDF, just about the most neutral material you can cut. No grain, even structure.

To quote.....'is a puzzlement!"

[4DThinker]

There is flex in the whole system, and it varies depending on which way you push it. When a router bit runs through solid wood there is a force that pushes the system generally 90 degrees from the direction of cut. That could lead to circles that weren't perfect. Most flex would occur if you are cutting in the center of the table. You might try cutting your circles in one corner of the table and comparing them to the ones cut in the middle. Not sure if that is your problem given how little you are removing with each circle, but could certainly be one of the contributing variables.

[wolffie]

I FINALLY got my problem solved.


The solution?

A replacement box that worked right out of the packaging

It is supposed to be from the same shipment as my old one but looks very different.

Cheers
Wolffie

[rungemach]

iampapabear,

It looks like you already covered all of the "usual suspects" and checked everything on your machine for tightness.

You mentioned your machine is an HD, and I can tell you that I was getting better cuts on my old shark pro than you are getting now,
so something is definitely wrong. There should not be any where near that amount of flex in an HD.

Also good to note is that the feed rates that vectric has as standard are usually too fast for a Shark.
200 ipm is possibly for rapids, but not good for accuracy. Your cuts at 50 ipm still look horrible. When I look at some of the squares it looks like the router overshoots in one direction, then comes back to make the side cut, overshoots again, etc. you should get a square, inside and out, (allowing for the radius on the corners of the outer edge. Does your machine "knock" or "hammer" when it makes an abrupt 90 degree direction change as higher speeds? or does it make the turn gracefully?

This may be a long shot , but: The older shark stepper motors had a spring loaded bearing setup that could exhibit symptoms resembling backlash on cuts where the direction of cut was away from the stepper motor. What would happen is during the cut, the force on the lead screw would push toward the stepper motor and push the motor shaft into the motor housing where it would compress a wave spring. Then the spring would push it back when the motor slowed down or stopped. This is very hard to detect unless you are looking for it.

I believe I have read that the motors used on the HD now do not have that type of construction, but you can check by pulling your z axis (router carriage) in the -x direction (toward the stepper) and see if the lead screw shaft moves into the stepper motor and springs back. you can do the same in the Y axis by pulling the gantry sides toward the stepper motor and see if there is any "spring" in the motor shaft. That would give you problems that increase as feed rate (and cutting force) are increased, but should give you perfect cuts at very low speeds.

If you can not wiggle the carriage or the gantry back and forth, (no freeplay), and you can not get a perfect cut at 10ipm at the .050 depth, I would start to suspect ether the Gcode or the controller.

You might want to program a simple square, and post the files here so we can see what is being generated by Vectric to be sure it is not telling the router to trace that odd path.

It may be worthwhile to send your controller in for diagnostic checks just to eliminate that as a possibility. Have NWA cut a square and circle with your controller and compare the results to another new HD.

Hope this info helps.

Bob

[wolffie]

I just found something that I thought was interesting

Instead of looking for runout, I put a digital anglefinder on the table, then on the router.
Turned out the router was nearly 3 degrees off the perpendicular to the table.
That's a lot.
It was not the mount, that one was pretty good, it was the setup of the router.
I corrected the setup, then put double sided tape between the router and the mount so that it cannot shift.
Bit is exactly where it is supposed to be and no runout.

Makes you wonder how many others think they have bad runout when the real culprit is that the router is not mounted correctly?

Cheers
Wolffie

[4DThinker]

Makes you wonder how many others think they have bad runout when the real culprit is that the router is not mounted correctly?

Cheers
Wolffie

I'm not even sure the Bosch CAN be mounted incorrectly. A cylinder in a cylindrical hole. Can see no way that router could be installed wrong.
Guessing you have a different setup than I do.

For my dimension errors the only play in my system are the feed nuts on my X and Y axis screws. They are plastic flange nuts with relatively coarse ACME thread. I've ordered replacements online. NWA responded to my request for replacements with a link to their part page, but those flange nuts aren't on their parts pages. Found them at Roton Products. I'm assuming they developed slop whenever the router bumped into the X and Y limits. You know the controller can't tell, and so the motor would keep turning until it couldn't. That would be enough to "strip" threads in those nuts at least a little each time. I'm much more cautious about keeping my jobs inside the limits now, but know I found them "accidently" several times in the past. With even a little play in those nuts the force of the router though whatever it was cutting would push the screw to one side or the other of the play inside the nut. With 1/64" play a climb cut could account for 1/32" oversize error around any shape, and a convential cut would have a 1/32 undersize error. Error in the X dimesion but not the Y suggests the X nut is defective and the Y is not. I've ordered replacements for both on mine. Anyone with similar dimensions errors on an older Shark should consider those flange nuts as the problem.

4D