Something Wacky with Spoilboard

[thecabinetconnection@earthlink.net]

Hi Everyone,

I replaced our spoilboard. I then machined it flat with the router.

So, we decided to give the bot a whirl with a test cut. I drew up for corbels and cut them out of .8125" red oak. (Used Vector to generate the cam.)

I ran the z-zero program to determine tool height from the substrate. I told Vector to cut .875 for two passes.

At 0,0 it left just a whisper of material. But as it cut down away from the x-axis, the uncut material became thicker. Over 3/16" thick.

What is going on? The board is uniform in thickness. The vacuum hold down was working and there were no visible gaps between the bottom of the board and the spoilboard.

The Y-car isn't riding up on the bearings. I checked the x-axis rails for a hump but according to the straight edge, there isn't one.

I'm a little stumped. Could my z-axis be drifting up?

Chris

[kerrazy]

Sounds Silly but was your collet clean? If not maybe your bit was slipping.
Another thing is ensure your wires a secure as it may or may not be conenecting all the information to the motors. And last but not least ensure your unit values are accurate for your machine.

Check your file in note pad and see if there is any variation that is different as this may lend some light to your situation.

Dale

[artisans@chdc.org]

Howdy Chris -

Learned an important lesson one day on a sign.

Did you surface your spoilboard with the vacuum on or off? Ours has a tendency to pull the entire spoilboard, underlayment and all, down about 1/16" from outside to center.

I was cuting a sign about six months ago and noticed the same problem (but even more scary - I thought that gravity was starting to go up - some new experiment happening here in DC)

I had recently surfaced the spoilboard, and did not do it with the vacuum on (didn't want the noise, heat, etc. in the shop that eve).

Cut the sign the next morning, wrecked the board, resurfaced it, wreched it again, reset the zero, checked the collet, checked the code, ate lunch, checked the ShopBot forum, found nada (maybe there's something there, but I was panicking)

I then ran a level on the table - the straightest, longest thing i could find - and the vacuum was still on -

There was a big ol gap at the center of the table.

Turned the vacuum off, went home.

Came back, looked at the level.

Gap was gone. Miracle of miracles, turned on the vacuum, there is is again.

From now on, I will only surface the table with the vacuum on full pull.

Other hints:
When you z zero (with the plate?), is the vacuum on? May help.

Here's a question back.

We're about to replace the spoilboard, and was wondering what the best way was to go about it without junking up the vacuum system.

Also, what gets contact cement, what gets silicone (besides the edges, nothing?) and should we "pre-surface" the MDF prior to shaving off the spoilboard? On both sides?

I look forward to your reply. Best to all, and howdy, Dale! Give us a call sometime!

Matt Barinholtz
Covenant House Washington, DC

[robinsoncr@naxs.net]

Hi All,

I've been swamped the last two days but I'm going to back and hit the ole'bot again in a few minutes.

I found nothing in the code that indicated that the z should move up. I checked the bit and it didn't appear to have shifted up 3/16". I DID NOT surface the spoilboard with the vac on. I was afraid of contaminating the system should it pull dust through.

I'm going to do a couple of experiments here in a few minutes. 1. I'm going to use my 6' straight edge and check the spoilboard's flatness. Then I'll turn on the vac and check it again and see what I get.

2. If that doesn't show anything, I'll go to 0,0 and move the bit to .015 from the deck and check with shims, then move to 12,0 and do it again. I'll do this over a 24x48 grid and plot the data should there be significant differences. That should split the universe a little.

As for zeroing the z-axis. I zeroed it with the vac on. I didn't surface both sides of the board. I couldn't get past the fact that since the table DOES have deviations in the horizontal plane, that by surfacing on one side and flipping it over and surfacing it again could only compound the problem. Even though it should improve flow.

We have a rubber seal around the perimeter, between the spoilboard and the plenum. I'm going to cut some 2" strips out of 6.5mm one sided melamine and caulk the uncoated side and apply to the edges of the spoilboard/plenum. Should help.

Thanks,

Chris R.

[robinsoncr@naxs.net]

Ok, I fired up the vac pump and then took my 6' straight edge. You couldn't see a gap anywhere on the table. You couldn't see a gap anywhere on the table with the vac off....

So, I put a bit in, zero'd the tool to the spoilboard. I then lowered my tool to .025" from the deck and used a .025" shim to check clearance.

I then moved it all around the deck. Slight interference everywhere I checked. So, I must conclude that Dale's assertion must be correct. Bit slippage?.... Tomorrow, I'll run the thing again and see what happens. Before I run it, I'll mic the distance that the bit protruded from the collet and measure afterwards.

We'll see.

Thanks,
Chris

[ckurak]

Chris,

I agree with Dale. How about checking the collet for dirt, debris, resin, etc.? Or, a bad/bent/worn collet? The cost of a new collet is probably a lot less than the material you may ruin. Or, was the bit not far enough, or too far, into the collet? Just some thoughts...

Charles

[robinsoncr@naxs.net]

Hey,

One thing that occured to me today was my plunge speed. I was using an Onsrud 1/4" compression spiral and plunging at .5"/sec. I may have been plunging too fast. What are you guys plunging at into red oak?

Chris

[artisan]

Chris....Now that you mention the material is Red Oak and your bit type, like Dale and Charles, I'm inclined to suspect your collet more than ever. Perhaps your bit wasn't fully seated or the collet was just a little dirty or both, either could cause the bit to be driven back upwards in a hard material like oak as you cut. Your plunge speed would not cause the depth of cut to gradually change over the course of the cut. I have one major question....as always....can you repeat the mistake? This will answer your question probably as to whether it's a vacuum issue or just a one time slip of the bit....D

[robinsoncr@naxs.net]

Darrell, I thought about bit depth in the collet. The bit has 1.1" of cut area and is about 2.5" long. But, if I slide the bit as far as it will go up into the chuck, it looks like there's less than 3/4" out of the chuck. It wouldn't have enough to cut all the way through.

Chris

[ckurak]

Chris,

Sliding "the bit as far as it will go up into the chuck" may not be a good idea anyway. If the fluted section is now up in the chuck, or collet, then it may not be gripping the bit well either. You mentioned that 'The bit has 1.1" of cut area and is about 2.5" long.' If, after sliding the bit up, you only have 3/4" showing, then about 0.35" on cutting edge length (CEL) is up in the collet. This will probably NOT grip well. What about exposing the entire 1.1" of CEL?

You might also take a look at http://www.axyz.com/parts_pdf/Tooling_info.pdf. They go into some explanations about router bits and the why's and why-nots in detail.

Charles

[Chris R.]

Ok, I did more cutting of the same part file today. This time I watched the spindle as it cut each pass.

Just as a recap, these are small, simple corbels. I used Vector 9.4, drew one up and then cut/pasted it to get four. I then did the offset tooling, z-depth/steps, then generated the sb code.

Machine speed = 1.0/.25
It cut the first corbel correctly and cut through. On the second corbel, it cut through on one leg then as it cut the next line, the spindle raised. So I looked at the screen and sure enough, the z-axis was raising fast as it made it's cut, it appears to be exactly .25". It completed the next leg and on to it's beginning point for that corbel, then plunged cut a second pass. It cut the first line, then at the junction where it changes direction for the next line/direction. It then raised the spindle as it cut. But, this corbel is flipped. So, it raised in the same location on the table, not the part!

On the third corbel, it did basically the same thing, except it tried to plunge farther than set in the parameters on the last 2 inches of the second pass and the first 2 inches of the third cut line; by about 1/4". In fact, it hit the collet and burned the oak.

On the fourth corbel (all in a row, oriented down the x-axis.) it did the raising thing but not the plunging too far deal.

Ok...WHAT THE HELL IS GOING ON?!!!!!

Chris R.

[srwtlc]

Chris,

After reading your recap and your problem discription, it made me think of something. You said that you used Vector to make your toolpath. Could it be possible that there is a Z-axis arc in your drawing (CG in the Z direction)? If there is you will need to "Break Interpolate" it. The ShopBot doesn't do arcs in the Z-axis. It will get lost if there is one in your file.

Just a thought because it happened to me once. If that is the problem and you need the arc to be smoother, just interpolate it with a smaller setting.

Scott

[robinsoncr@naxs.net]

Scott,

The corbel does have an arc in it. But I don't understand what you mean by "arcs in the z-axis".

The corbel is drawn in 2d and then I dictate in Vector the depth of cut. First cut was supposed to be .3885"deep, final cut was .777" deep.

Do you mean that if I drew a simple circle in Vector, generated the code and then cut it, this would happen?

Thanks,

Chris

Ps: Here's the sbp code from Vector...
'SHOPBOT PROGRAM GENERATED BY VECTOR CAD-CAM
SA
J3,0.2500,0.1250,0.8000
J3,0.2500,0.1250,0.1000
M3,0.2500,0.1250,-0.3885
M3,6.2500,0.1250,-0.3885
CG,,6.3750,0.2500,0.0000,0.1250,,-1
M3,6.3750,1.2179,-0.3885
CG,,6.2687,1.3415,-0.1250,0.0000,,-1
CG,,1.3415,6.2688,0.8813,5.8085,,1
CG,,1.2179,6.3750,-0.1236,-0.0188,,-1
M3,0.2500,6.3750,-0.3885
CG,,0.1250,6.2500,0.0000,-0.1250,,-1
M3,0.1250,0.2500,-0.3885
CG,,0.2500,0.1250,0.1250,0.0000,,-1
M3,0.2500,0.1250,-0.7770
M3,6.2500,0.1250,-0.7770
CG,,6.3750,0.2500,0.0000,0.1250,,-1
M3,6.3750,1.2179,-0.7770
CG,,6.2687,1.3415,-0.1250,0.0000,,-1
CG,,1.3415,6.2688,0.8813,5.8085,,1
CG,,1.2179,6.3750,-0.1236,-0.0188,,-1
M3,0.2500,6.3750,-0.7770
CG,,0.1250,6.2500,0.0000,-0.1250,,-1
M3,0.1250,0.2500,-0.7770
CG,,0.2500,0.1250,0.1250,0.0000,,-1
J3,0.2500,0.1250,0.4000
J3,7.7804,0.1259,0.4000
J3,7.7804,0.1259,0.1000
M3,7.7804,0.1259,-0.4785
M3,8.7481,0.1259,-0.4954
CG,,8.8732,0.2509,0.0000,0.1250,,-1
M3,8.8750,6.2500,-0.3907
CG,,8.7500,6.3750,-0.1250,0.0000,,-1
M3,2.7510,6.3750,-0.2816
CG,,2.6259,6.2500,0.0000,-0.1250,,-1
M3,2.6256,5.2823,-0.2985
CG,,2.7318,5.1587,0.1250,0.0000,,-1
CG,,7.6569,0.2322,-0.8829,-5.8076,,1
CG,,7.7804,0.1259,0.1236,0.0187,,-1
M3,7.7804,0.1259,-0.8670
M3,8.7481,0.1259,-0.8839
CG,,8.8732,0.2509,0.0000,0.1250,,-1
M3,8.8750,6.2500,-0.7792
CG,,8.7500,6.3750,-0.1250,0.0000,,-1
M3,2.7510,6.3750,-0.6701
CG,,2.6259,6.2500,0.0000,-0.1250,,-1
M3,2.6256,5.2823,-0.6870
CG,,2.7318,5.1587,0.1250,0.0000,,-1
CG,,7.6569,0.2322,-0.8829,-5.8076,,1
CG,,7.7804,0.1259,0.1236,0.0187,,-1
J3,7.7804,0.1259,0.4000
J3,9.2500,0.1250,0.4000
J3,9.2500,0.1250,0.1000
M3,9.2500,0.1250,-0.3907
M3,15.2490,0.1250,-0.4954
CG,,15.3741,0.2500,0.0000,0.1250,,-1
M3,15.3744,1.2177,-0.4785
CG,,15.2682,1.3413,-0.1250,0.0000,,-1
CG,,10.3431,6.2678,0.8829,5.8076,,1
CG,,10.2196,6.3741,-0.1236,-0.0187,,-1
M3,9.2519,6.3741,-0.2816
CG,,9.1268,6.2491,0.0000,-0.1250,,-1
M3,9.1250,0.2500,-0.3863
CG,,9.2500,0.1250,0.1250,0.0000,,-1
M3,9.2500,0.1250,-0.7792
M3,15.2490,0.1250,-0.8839
CG,,15.3741,0.2500,0.0000,0.1250,,-1
M3,15.3744,1.2177,-0.8670
CG,,15.2682,1.3413,-0.1250,0.0000,,-1
CG,,10.3431,6.2678,0.8829,5.8076,,1
CG,,10.2196,6.3741,-0.1236,-0.0187,,-1
M3,9.2519,6.3741,-0.6701
CG,,9.1268,6.2491,0.0000,-0.1250,,-1
M3,9.1250,0.2500,-0.7748
CG,,9.2500,0.1250,0.1250,0.0000,,-1
J3,9.2500,0.1250,0.4000
J3,16.7776,0.1275,0.4000
J3,16.7776,0.1275,0.1000
M3,16.7776,0.1275,-0.6120
M3,17.7449,0.1278,-0.6458
CG,,17.8700,0.2528,0.0001,0.1249 K0.0044,,-1
M3,17.8755,6.2491,-0.4365
CG,,17.7506,6.3740,-0.1249,0.0000 K0.0044,,-1
M3,11.7543,6.3722,-0.2184
CG,,11.6293,6.2472,-0.0001,-0.1249 K-0.0044,,-1
M3,11.6284,5.2799,-0.2522
CG,,11.7345,5.1564,0.1249,0.0000 K-0.0044,,-1
CG,,16.6542,0.2337,-0.8860,-5.8053 K-0.1719,,1
CG,,16.7776,0.1275,0.1235,0.0188 K-0.0037,,-1
M3,16.7776,0.1275,-1.0005
M3,17.7449,0.1278,-1.0343
CG,,17.8700,0.2528,0.0001,0.1249 K0.0044,,-1
M3,17.8755,6.2491,-0.8250
CG,,17.7506,6.3740,-0.1249,0.0000 K0.0044,,-1
M3,11.7543,6.3722,-0.6069
CG,,11.6293,6.2472,-0.0001,-0.1249 K-0.0044,,-1
M3,11.6284,5.2799,-0.6407
CG,,11.7345,5.1564,0.1249,0.0000 K-0.0044,,-1
CG,,16.6542,0.2337,-0.8860,-5.8053 K-0.1719,,1
CG,,16.7776,0.1275,0.1235,0.0188 K-0.0037,,-1
J3,16.7776,0.1275,1.6000
'MOVE TO X-Y HOME POSITION & THEN RAISE TO Z HOME
JH,
'DONE
PAUSE 20

[Mayo]

Maybe cut and paste is causing the problem since the first one cuts properly?

[srwtlc]

Hi Chris,

I took your code and backplotted it in Vector. You don't appear to have any of what I was talking about in your file. I'm not exactly sure what it is you're trying to do here, but I'm guessing that you're trying to just cut out four identical parts.? If your code backplotted right for me it looks as though your depths got messed up during the pasting process. Take another look at your Vector drawing with a default 3d view and then a front view and you can see where those Z moves are going awry.

Scott

[robinsoncr@naxs.net]

Scott,

Yes, I'm trying to cut 4 identical parts. I drew the first one, then cut and pasted the other 3. After that, I selected each line in a counter-clockwise fashion, each corbel at a time and then did the offset tooling, then z-moves, then generated the sbp code. Doing it this way, how could it get the first right and then go nuts from there?!

Before generating the code, I did take a look at the default 3d view and it looked correct. I'm not sure what's going on here...

Thanks,

Chris

[srwtlc]

Chris,

I don't wonder if when you copied and pasted you may have not reset the values in the paste dialog box. If there was a value in the Z field it would get compounded each time you pasted.

E-mail on the way.

Scott

[robinsoncr@naxs.net]

Scott,

I don't know. Today I ran the file again. I simply reopened the file and ran at the same location on the table, same bit etc.

This time it did it on the first piece as well.

This is getting a little frustrating. When I pasted the corbel drawings, I insured that there was nothing in the z-axis data field and only used x & y data fields for placement.

Chris

[artisan]

Chris... can you email me a DXF of a single Corbel. I want to duplicate your process and compare it to your code. I don't use Vector, but I can generate the code another way and I'd like to see how it compares with what you've posted....D

[grandpas@ix.netcom.com]

Could the problem be that the z-axis settings are changed. Check axis for all screws tightened.
If you changed and software, settings changed. Check varianbes and reset. I experienced the similar problem on the 2nd axis I installed and until I got the rack and pinion settings verified and corrected, that was my problem. also check to see that all comma's are in correct positions.

[srwtlc]

Chris,

If you look at the code above and walk through it you can see that the first one cut as you want, but after that your paths start going deeper and on angles in the Z plane. If you wanted them to all cut the same as the first one then something went wacky during the pasting of each successive one. Did you try the file I sent you? Does that one cut/move okay? Can you compare my .ccd with your original .ccd at various viewing angles?

Scott

[robinsoncr@naxs.net]

Darrell, Scott and All,

I was out on a jobsite most of the day and didn't check my email until tonite. I'll compare the files and give them a try tomorrow. I'll post what I find.

Thanks a million!

Chris