Accuracy Problem

[erik_f]

Is there something that I could be over looking as to why I can't get my BT48 to cut accurately? I'm too frustrated to get into it right now...but it seems to be cutting a lot of parts slightly over sized. The table is square.

[Gary Campbell]

Eric...
You dont say if you are toolpathing in a CAM software (PWiz, VCPro) or using the SB [C]ut Commands.. or yet another method. Bit diameter measured? Depth of cut? Move speed? Spindle or Router? Bit type, Upcut, Downcut, Straight?

Are any of the parts oversize in the X, Y dimension, or both? The 3 things that come to mind first would be bit diameter different than toolpathed, second, bit deflection and third, machine flex caused by a loose or out of adjustment roller on the X, Y or Z.
Gary

[dana_swift]

Erik, at post 176 since 2004, you clearly are not new to the machine, so something is amiss. But it most not be obvious or you would have figured it out from experience.

Gary's suggestions are very good as usual, and also, how about post some pictures of the results and perhaps the cut files.

What material are you working with?

Can you make some precision tests to see how far off it is? Are we talking 0.001" or 1.000" per inch?

Is it repeatable? To test repeatability: you can drill a hole, then move the carriage around without cutting using the MX and MY commands.

While the router is clear of the hole, use a sharpie to mark all around the inside of the hole. Then M2 back to the same place and drill down into the same hole. Then approach the hole from different directions and see if any material is removed. The sharpie marks will be missing on the side that has backlash or flex problems.

If it really has problems you may have a figure 8 or two holes!

Its a cheap and easy test, which can might yield a clue.

Be sure to verify the values you have in your VU settings, those must be correct for your combination of stepper motors/gearhead and rach/pinion pitch. Tech support is worth checking, my BT32 came with the wrong values, and the "calibration" data that came with the machine was wrong.

Is the machine alpha or standard?

Hope some of these questions help you discover the culprit!

D

[erik_f]

Thanks guys...I have a PRS Standard buddy. A nights sleep has helped me cool off a bit. I have made a few posts about other problems that my buddy came with. So mostly I have only been making cuts to square up the machine and test it for accuracy. The hole idea is something I have not tried yet and like you said cheap and easy. I'm going to ask about the Partworks Post processor and unit values today when I call SB. It seems to off about 1/64th when its off...I realize that is mostly within SB spec...but I was told it should be the exception not the rule. Last night I used an straight 1/2" Amana bit...and I didn't measure this bit...but in the past they have been within .001" of .5". One of the pieces came out pretty close to dead nuts. The other piece come out slightly oversize and slightly out of square...so maybe the work piece moved. I will just have to go throught it some more...but this is exactly what I didn't want to have to do with my new shopbot...test cut after test cut, every night I have free for a month to get it dialed in. Thanks for the help guys.

[erik_f]

Tech support seems to think its hold down issue...

[sawdust535]

Erik
The other day I was pocket cutting 1/2" squares with a 1/4" down cut spiral bit (Onsrud) and I was off a little more than 1/64" (pockets too small). Turned out it was the bit. The diameter of the bit and the width of a test cut I made was off. I changed the bit (used a 1/4" compression bit from Centurion Tools) and problem solved. I have a BT 32 and I am amazed at the accuracy and repeatability of the cuts.

[dray]

odds are its the bit..
I normally change all my bits out fairly often, but been busy. I recently cut a set of doors with insets and the insets didnt fit by almost 1/16""

I threw a tantrum, dragged out the table saw and recut the insets, then after carefully going through everything, I found it was my bit..

by cutting the inside of the door on the inside of the vector it was making it 1/32" too small and cutting the outside the vector on the inset caused it to be 1/32" larger!

so the bit had worn down by just under 1/64"..

Use a micrometer on your bits.

if it was a hold down issue it would be lopsided.

I would personally stick with tech questions for the shopbot tech support ppl (software etc)

[Gary Campbell]

Eric...
Hold down would have been the 4th thing that came to mind (easy to say in hindsight) but with over 4 yrs on the forum, I figured you had a handle on that by now. Good luck, and keep us posted.
Gary

[erik_f]

Part of the problem is I've only had about an hour a night to get at it. I noticed that some of my test parts were slightly out of square...I'm trying some new clamps out...so as foolish as it makes me feel to say...it could be the hold down thing. I cut some 5" squares and they were slightly out...then reclamped and cut the same squares but at 4.75" so there wouldn't be as much cutting force...and the two came out pretty good. Hard to tell on such small test pieces though. I just need more time. Once I get some more boards cut I'm going to screw down my test board instead of using the clamps and re-run the first test file I started the post about and hopefully it is something as easy as hold down...like someone else said...after all this time you would think I would have this licked. Thanks to all.

[Gary Campbell]

Eric...
To make sure its the hold down, screw down a large piece and make inside profile cuts on different size squares/rectangles. Then you can work on hold down for smaller squares. Its nice to have confidence in the machine, completing this may help.
Gary

[da_mike]

Did you make sure the gears on the stepper motor were not loose?

I had a similar issue that turned out to be one of the gears had wiggled itself slightly loose.

[erik_f]

There is a little bit of slop when the table is not in motion...but from what I understand that slop is not there when the steppers are in motion. I am going to try a few things over the next few days to eliminate any mistakes I maybe making. I will double check the pinions. I'm finding it harder to find good hold down solutions for the BT compared to the old stationary table PRT.

[dana_swift]

Erik- The ultimate hold-down on the BT's is the 1/4x20 grid that is part of the table. The best part of the grid is it is a rigid anchor for hold-down, and a position reference system all in one.

For very small parts, I have found a solution that works well:

Spray a small sheet of HDPE with 3M's Super-77. Then put the small parts on the HDPE held with the contact cement. The sheet of HDPE goes onto a vacuum hold that is bolted to the grid.

The contact cement holds both the part and the scrap solid until the cutting is finished.

Because HDPE is a high-energy surface the Super-77 will release readily. Then I put naphtha on a rag and wipe off the HDPE and the parts, that gets rid of the Super-77.

Im ready to cut the next batch of tiny parts..

Hope that helps

D