[PortlandGTS]

I'm finishing up the build of a Fine Line Automation 300 (Pics coming soon). To check the X-Y alignment, I've been cutting squares of various sizes. Turns out the alignment is great, but every square comes out 1/16" short, in both dimensions, no matter the size. At first I thought I had the steps per inch wrong, but that should give an error proportional to the distance. But I've cut squares from 17" to 46", and all come out with the same 1/16" error.

The router uses ring and pinion drives and motors from CNCRouterParts.com and a Gecko 540 controller. I'm using EMC2 on Linux for software. I'm using the steps/inch ratio provided on the CNCRouterParts site.

Any ideas what's wrong?

Thanks,
Tom D.

[ger21]

I'd adjust the steps per inch to get one size square to cut correctly, and then see what you get on the others, and go from there.

Also, try cutting CW on one, and CCW on the other, and see if you get any differences.

Check everything and make sure there is absolutely no play anywhere.

[steve123]

I wonder if you specified the wrong size cutter? If your cutter was larger than your software said it was, everything would come out smaller, no matter what size square you cut. Cheers. (I'm no expert. My 1st post on CNC zone)

[PortlandGTS]

Gerry,

Thanks for the suggestions. I now think the problem is measurement error - it's hard to measure almost 4' to an accuracy of less than 1/16".

I need to play around some more, but I think the steps/inch given on CNCRouterParts.com is probably correct.

As far as entering the wrong tool size, I haven't gotten that far yet. I'm just entering G code directly.

Thanks,
Tom

[restless1]

i built a similar machine to the fla300 but modified it to be longer and after calibrating it then comparing it to the numbers that cnc router parts supplied on there web site i found there numbers were better than my measuring devices ...mines all rack and pinion except the z axis

[flojor10]

Check your acceleration and ipm's, if it is to high when you start cutting your machine will lose steps when it starts cutting and this could be why you are always off by 1/16".

-George

[brtech]

I think your answer is "tool size".

I'm guessing you have a 1/8" bit.

You poke the tool down. You have a 1/8 hole. You move up 1", move right, move down,move left and retract. The size of the edge from that path would be 15/16" because you are losing 1/16" radius on the 1/8 diameter tool. Draw a picture.

The key information is that you lose 1/16 no matter what the size of the square. That's not steps per inch or acceleration or anything else that would vary by the length of the cut. You say it's always 1/16". I'm betting it's tool diameter.

To get a 1" square with a round bit of 1/8", you have to move 1 1/16. The inside edge will be 1". The center of the path will be 1 1/16 and the outside edge of the path will be 1 1/8.

[PortlandGTS]

I've been keeping the feed rate down, but hadn't thought about acceleration. How do you choose the proper value for acceleration?

Thanks,
Tom

Check your acceleration and ipm's, if it is to high when you start cutting your machine will lose steps when it starts cutting and this could be why you are always off by 1/16".

-George

[PortlandGTS]

Good guess, but no cigar. I've been using a 1/4" bit, and measuring from the outside of the cut on one side to the inside of the cut on the opposite side. (I'm not cutting all the way through.)

Besides, if you have a 1/8" bit, and move 1" in all 4 directions, wouldn't you wind up with a square of size 7/8", not 15/16"?

Tom D.

I think your answer is "tool size".

I'm guessing you have a 1/8" bit.

You poke the tool down. You have a 1/8 hole. You move up 1", move right, move down,move left and retract. The size of the edge from that path would be 15/16" because you are losing 1/16" radius on the 1/8 diameter tool. Draw a picture.

The key information is that you lose 1/16 no matter what the size of the square. That's not steps per inch or acceleration or anything else that would vary by the length of the cut. You say it's always 1/16". I'm betting it's tool diameter.

To get a 1" square with a round bit of 1/8", you have to move 1 1/16. The inside edge will be 1". The center of the path will be 1 1/16 and the outside edge of the path will be 1 1/8.

[p00ky]

Perhaps this won't be of much help....

In mach3 the acceleration settings are next to the no. Of steps settings.
I found this YouTube video for emc2

[ame=http://www.youtube.com/watch?v=Vm2YQVG0kdQ]emc emc2 configuration wizard - YouTube[/ame]

About 1m30 seconds in he sets the acceleration, its a box below the steps settings (he's using the inbuilt wizard). He's working in mm and sets it to 100, this would be approx. 4inches per second. This is about what I set mine too.

Sorry if this isn't useful. Hope you sort it.

[scottb08]

I'm in the process of building a FLA300 as well. I bought the kit from Fine Line. Unfortunately, Nates BOM is wrong and only specifies 1 set of thrust bearings for the Z axis instead of 2. I've tried to contact Nate several times to see if he would send me 1 more but he hasn't returned any of my emails or phone calls.

Did you assemble the pieces yourself or buy the kit from Fine Line? Did you have this problem?

Thanks,
-Scott

[OCNC]

...But I've cut squares from 17" to 46", and all come out with the same 1/16" error....Any ideas what's wrong?

Thanks,
Tom D.

I just had a wild idea about this problem. Is the sliding tongue on the end of the tape measure stuck?

Chris

[freecncplans]

Setup a g-code file to cut a square. Run it twice. If you get a difference, you have something loose or a wrong setting somewhere.

Sean

[p00ky]

I just had a wild idea about this problem. Is the sliding tongue on the end of the tape measure stuck?

Chris

I once laid a floor, all checked out great on the spirit level until I got 3/4 of the way through. All of a sudden every reading I took appeared to be wrong. I scratched my head a lot but couldn't figure it out at all. Then a friend showed up and over a cup of tea he noticed that the spirit level gave entirely different readings depending on which way round it was, he then noticed a tiny little sign on it saying "adjustable". Who ever thought of an adjustable spirit level :S

Anyways, not a bad suggestion that, all ways best to check the little things. They can lead you astray.

[PortlandGTS]

He's working in mm and sets it to 100, this would be approx. 4inches per second. This is about what I set mine too.

This may help. I had acceleration set to 6.5. I'll try setting it back to 4.

Thanks,
Tom

[PortlandGTS]

I'm in the process of building a FLA300 as well. I bought the kit from Fine Line. Unfortunately, Nates BOM is wrong and only specifies 1 set of thrust bearings for the Z axis instead of 2. I've tried to contact Nate several times to see if he would send me 1 more but he hasn't returned any of my emails or phone calls.

Did you assemble the pieces yourself or buy the kit from Fine Line? Did you have this problem?

Thanks,
-Scott

I bought pieces from the sources in the BOM. I would have bought everything from Nate, but I got spooked by his lack of response to my e-mailed questions. If you search this forum, you'll find lots of people with similar issues. The consensus seems to be that Nate eventually comes through, he's just (very) slow to respond at times.

Tom

[PortlandGTS]

I just had a wild idea about this problem. Is the sliding tongue on the end of the tape measure stuck?

Chris

Not really. But I've realized there's easily 1/16" of flex in the tongue. I'm going to start over with all the measurements soon using the 1' marker as the starting point rather than the tongue. (I've been tied up in other things since my initial post, and may not be able to do anything more for another week.)

Tom

[CarveOne]

Not really. But I've realized there's easily 1/16" of flex in the tongue. I'm going to start over with all the measurements soon using the 1' marker as the starting point rather than the tongue. (I've been tied up in other things since my initial post, and may not be able to do anything more for another week.)

Tom

Instead of measuring the machine travel with gauges and tape measures, I recommend cutting something to a length like 12", 36", 48" and then measuring the results. That's the end goal anyway, and it's easier to measure edge to edge of the cut material. As long as you don't have backlash, there may be mechanical tolerance buildup causing the 1/16" offset.

That's a lot of error for the distance you are measuring though. If there is no cure other than replacing mechanical parts at more expense and effort than you are willing to accept, then adjust the steps per inch value until you dial out the error - and get on with making interesting stuff to show us.

[Drools]

then adjust the steps per inch value until you dial out the error

This is exactly what I did with mine and now it works great. I was out .024 over 25". If you have a look at posts 146 through 153 of this thread http://www.cnczone.com/forums/cnc_wo...thk_cncrp.html you see how the guys on the form helped me fix mine.

[brtech]

...But I've realized there's easily 1/16" of flex in the tongue.

You aren't talking about the small movement of the hook along the length of the blade, right? That's the way it's designed to work. When you take an inside measurement, the hook is pressed in towards the middle of the tape, and you are reading from the outside face of the hook. When you take an outside measurement, the hook extends out the thickness of the hook and you measure from the inside face of the hook.

If it didn't do that, you would get an error for one or the other measurements. It would be off by the thickness of the hook.

Now, rivets do loosen up, and the holes in the blade can get worn so the movement of the hook is greater than the thickness of the hook and you get an error. Then it's time to toss the tape measure and get a new one.