[Weldtutor]

Your machine is coming along nicely Milhead.

Thanks for sharing the pics.

Regarding the "space" problem, have you pointed out to the GF how nicely tole paint sets up in a cooler space?

[Buzz9075]

milhead, two suggestions I have heard (heard being the big word here) on false limit triggers. 1 - Opto isolation of the Parallel port is suppose to help, 2 - There is a setting in mach 3 config->general->debounce that should be set to something like 1000.

[milhead]

It Spins! it cuts!..

I put the electronics together and wired everything up! Except for the
limit noise issues I had no suprises.

As a fellow completely new to CNC terminology I spent way too many
hours trying to understand manuals and fooling with setting limits and
otherwise learning... But this old dog apparently can be taught (spelling
aside)..

I had a entertaining series of steps using a Sharpe as a tool and trying
to draw things.. Some good some bad... Some of these were pretty entertaining, It's too bad I did not save any of the pages.. My daughter
made an honest attempt to draw a dog with the machine as an
etch-a-sketch via the jog controller... Identifiable.. Well... Kinda...

Since then I've made 5-6 cnc wood carvings including the sign below..

My task list now (with a good working weekend coming up)

1) find out why I'm getting wierd jumps in my Z-axis setting....

2) Put the machine in the garage and readjust the fine leveling...

Questions:

Since my design (not being of the box type) can have some
torstional flex. Is there a reasonable way to measure it?

Rather than the accuracy of the table being required, is there a
way to setup Mach3 with the error values for my table and have it
interpolate this into each movement? We do this sort of thing at
work for the machines I work with...

I believe that I have a fair amount of blacklash.. When I tried to mill
some small designs the slop shows pretty badly. Does anyone have
a link for a fair tutorial on how to measure it? I have some ideas but
have learned that there are few problems not already solved.

I'll post again when I get the small stuff (and big accurate stuff) to come
out to some spec, Till then I'm going to be cutting things that are
appriciated visually and not with a caliper...

[Weldtutor]

It Spins! it cuts!..

Since then I've made 5-6 cnc wood carvings including the sign below..

I'm going to be cutting things that are
appriciated visually and not with a caliper...

:wee: :banana: :wee:

Congratulations on a successful project.

Nice carvings, thanks for sharing your results.

[milhead]

Last night I had about 3 hours to layout and cut a farwell sign for
a co-worker.. One chance and I produced somthing.. Not nearly as
good as I would have liked.

It was the second large peice I've made.. I had one peice of oak..
Add a three hour running clock and it was an evening of excitement,
frustration, nerviousness, noise (don't forget noise).. and flying chips..

The savy CNC eye will notice 4 major flaws...

1) My groves are thicker at the top of the board than at the bottom...
I'm still about 0.1" off in the z axis as the Y goes from 0 to 18". I've just
not had time to adjust it out..

2) There are two unwanted gouges... Being a complete 100% CNC novice
I learned how to stop mid program adjust and start the damn z-axis
mid program... It took two bad errors to figure this out...

3) The top left of the plaque ran off the material.. (oops) but with only
one peice of wood and 1 hour remaining on the clock I had to let it slide.

4) z-Axis!... Grrrrrrrr! In the middle of the program (which was running
pretty slow) The z-axis will freak and re-zero itself a quarter or half
inch high.. I guess it's dumb luck that it's not happend to zero itself
down inside the peice (a half inch down would be a pretty spectacular
jump..) During this 50 minute excercise I had to readjust the z axis
about 5 times... #1 I think I just got lucky, #2 was the trough and #3
was the hole....

Since I'm still in the remedial CNC classroom.. I give myself a C+

Anyone want to chime in with a grade? I'm interested in your opinion..

Miller

[Weldtutor]

Anyone want to chime in with a grade? I'm interested in your opinion..

Based on the visual appreciation rule mentioned earlier in the thread I believe
you should give yourself a much higher grade, as I'm sure the recipiant of the plaque would!

For your ability to convince the GF that your router could be used in the house, a big A+ and some cheers.:cheers:

Looks great from here.

What did you end up using at the top to guide your wiring?

ALSO (What is a 2 year pumpkin rule?)

[milhead]

For your ability to convince the GF that your router could be used in the house, a big A+ and some cheers.:cheers:

What did you end up using at the top to guide your wiring?

ALSO (What is a 2 year pumpkin rule?)

Cutting MDF in the house was easy for the vacuum to capture... This was probably the last in-apartment peice I should make politically...

Currently the wires are just all kinda trailing where they want.. I used
some 4-up trailer wire for the z-axis wires but the rest are just all kinda
zip-tied in such a way as to not bind anywhere... This is still a problem
seeking a solution... I'm chasing a rumor that there may be some
scrap Igus chain around the shop... If that does not pan out I'll try to make something... It's kinda low on the ToDo: list at the moment.

The pumpkin rule requires contractractors to exit after two years, No
matter how valuable they are to the project... Corporate wisdom I guess,
we've been messed up by it before..

Thanks for the feedback!

Miller

[milhead]

Hi Guys!

I've had a few more hours on the machine now. It's been moved
to the garage.

I am still plagued by my z-axis loseing it's reference points? I set it
to 0 at the top of the material that I want to cut but after I simply start a
program it begins to cut .2 to .5 above the material... (slicing through
the air with great ease...)

In the current program I fooled with yesterday...

N10G00G20G17G20G90G40G49G80
N20G70
N30T1M06
N40G00G43Z0.7874H1
N50S16000M03
N60G94
N70X0.0000Y0.0000F20.0
N80G00X0.2142Y1.1748Z0.2362
N90G01Z-0.0335F10.0
N100G01X0.2134Y1.1821Z-0.0277F20.0
N110X0.2144Y1.1898Z-0.0239
N120X0.2152Y1.1935Z-0.0226
N130X0.2282Y1.1940Z-0.0001
(6700 more lines omitted for clarity..)

I have to set the 0, then let the program execute through line N70..
after it starts cutting through the air, halt the program, Reset the
0 and start execution at line N80....

Then it works pretty well for the remainder of the program, I have run some
big ones that have taken an hour or so....

I'm using Mach3 and quarter stepping my motors. I ahve also moved
off my laptop as a controller and have a dedicated 3GHz Pentium
as my controller machine.....

Any ideas?

[OCNC]

The photos above show a peice of 1/8 MDF on top of the base, this is NOT what I'm planning to use here... I'll have a 24x36 3/4 MDF top for the
table with two largish peices of angle iron running the length to keep it flat.

I'm not sure how to measure a torsional defection of the table yet but I'm
sure that I'm going to have a problem with it. I'm content for now to allow
the table to have some torsional flex but be rigid enough that I can adjust
it out when I set it on a hard flat surface...

A torsion box of some sort will probably be a result of my next round. I like
Joe's as it accounts for both torsional flex in the base as well as loading
flex on the large X axis rails.

Having a machine that can cut the ribs sounds like good fun...

Mil

Nice looking machine. The Purple Z is great. You might want to think about adding a stiffening rib to each of the gantry sides spanning from the x-axis rails up into the region where the gantry back attaches to the sides. Something like a 1x2 in a local hard wood or even an aluminum angle screwed and glued in place. In cross-section parallel to the x-y plane the sides should look like stubby Tees.

Chris

[ger21]

What is you're line N40 supposed to do? Do you want a tool length offset there? I think this is your problem.

You have duplicate G20's in line N10. Get rid on one of them. In N20, G70 is the same as G20 in Mach3. Get rid of it.

You can probably delete all the code before N70.

[Glidergider]

Outstanding project. Congrats on a well done build. What plans are your building too. I like the U-bolts feature. I'm currently building the Jgro but would like to adapt that U-bolt if its possible. Are your plans published?
Dave

[milhead]

Hi Dave;

Full credit for the U-Bolts goes to Joe and his long string of routers but when
I build next build I'll stick to it if the rod is solely supporting the gantry. one of the problems with the current machine is that just the weight of the gantry and tool is causing the rods to sag in the middle a few thousands....

I've thought about bending them but don't think I could get it right.. Joe's
box supports the full rod, Looks like it will work better.

Miller

[milhead]

Gotta be short because I have to get to work soon.....

First!! Thanks to every single person who has posted on their
own forums or on mine.. I've learned so much that even I'm amazed..
(old dog, new tricks.. etc). Well it's 90% to you folks that I give credit!!!
:cheers:

The G-Code help I received solved the large offset problem so zero is now
zero until I reset it!

My X and Y axis are working perfect. (Well, quite well anyway)...

I write software for a living, and we have a saying that some of the hardest
to find bugs are the most obvious...

I cut a peice of graduated steps this morning and measured them..

Hmmmm They were all about 90% of what they should be..... 0 was
0 in the four corners of my little workpeice (referenced to the top of
the board)... As I cut deeper steps they all got progressivly worse...
The Z axis was cutting about 90% of wanted depth

Well in a few brief moments.... 90% is nine-tenths... 9/10 or perhaps
I could write it as 18/20ths....

Doh! (wedge)


How many folks here think that the standard course thread on a 1/4
bolt is 18?

Well, It's one less today....

Once I told Mach that my Z-Axis leadscrew that I had 16000 steps per inch and not 14400 (the 18 to 20 TPI numbers).. Everylooks Much Much better!!

DIY Router Nervana....

Next steps... replace my zip cutter tool with my 6902VS Router motor
and I'll be able to increase my speed (at least to the point where I start
to run into leadscrew vibration problems...)

THANKS AGAIN TO ALL YOU GUYS! (and gals if any are lurking out there)

[milhead]

After fixing the Z-axis.. Here is the vacuum head and my next tool mount getting cut...

[milhead]

Hi Guys;

I replaced my tool head with my 1.5 HP Router head.. I can cut much
faster (20-30 IPM) than before

After spending 8 or so hours cutting this weeking on a project to make
end of year plaques for the soccer team I coach (U15).. I've become
sensitive to flex in the x-Rails for the Gantry. With the heaver gantry
the whole thing will wobble due to rail-flex, maximized when the gantry
is in the middle of it's travel.

This is a big incentive to move to Joe'style supported rail box when I
start anew.

However; as a patch up in the mean time..

I stupidly made my rails 39" long! So purchasing drill rod to replace
the rails would be 2 72" peices and way off budget...

How about jaming one or more small rebar peices and some concrete
into the existing rails? Is there enough cross-section to allow
the concrete/metal to stiffen the rail?

Do you think it would make any difference...

Miller

[paulC]

Don't think the rebar would do much other than make it hard to fill the rail.
Filling with concrete should make the rails stiffer. The tricky part would be allowing the concrete to cure properly and how long that would take.
Paul

[joecnc2006]

to get any structural strength, you will need a hi-yield concrete, standard concrete from out of the bag will not help at all, will just crack.

[higgrobot]

Milhead, Do you guys in the US have a source for drawn and polished Bright Steel bar? I know it's a lot cheaper than Ground Drill Rods and surely would be better than what you are using on your rails at the moment.

[Glidergider]

Miller,
I noticed back in your post #19 that you have a steel square tube running parallel to your gas pipe. Consider removing the U-bolts and welding your gas pipe directly to the square tubing. With a good side support, you might be able to transfer some load into your pipes. Limiting the warpage due to heat would be my number 1 concern. This would be the metal equivelant of a Joes2006 design.
Dave

Hi Guys;
,,. I've become sensitive to flex in the x-Rails for the Gantry. With the heaver gantry the whole thing will wobble due to rail-flex, maximized when the gantry is in the middle of it's travel.

This is a big incentive to move to Joe'style supported rail box when I
start anew.

Miller

[milhead]

Hi Folks;

This will be my final entry in this machines log.... The machine has probably 50 hours of cutting signs, parts and fun stuff on it...

My Son (14.5 years) worked with me to write a set of programs that
would carve mini-finger skateboards out of blocks of wood.. He finishes them, sticke grip tape on the top and sells them at school for $2.00 so someone has made a profit on the machine.

Not to worry, I have the CNC bug worse than ever and have started acquiring parts for machine #2... My main goal is to get rid of some of the annoying behaviors from the first machine and (of course) faster and more accurate.

My total BOM on this machine was around $500 counting electronics, bars, bolts, HDPE, bearings and Wood.. While I was pretty good at designing with CAD drawings in the beginning I don't have a complete set of drawings for the bottom end as it all kinda came together in a custom-fit sorta way.

Design #2 is in the works... I'm figuring my BOM at $1000 and am using more commrecial parts...

Sam is gonna have to sell alot of $2.00 mini-skateboards.. (grin)

Objectives for Machine #2.

#1 Optimize table footprint, My garage is a tiny place, I can't just get bigger to solve problems.
#2 Make the Gantry move front to back (This may not be wise mechanically but will keep it out of my way, I'm tired of looking around it.) It will also allow the tabletop of the machine to be used as a workbench when not in use. (again to optimize space)
#3 Need Supported rails for the gantry.. As stiff as the 1.25 bars are, There is alot of flex when the gantry is in the middle of the X axis.. There is a .02 measurable sag as well in both bars but the worst is a rotational
flex around the x-axis when the force on the two X rails is oppisite (up and down).. Joes design covers this very well, But I'm looking for a smaller way to do it.
#4 The open 1" rails for the Y axis also flex, this yealds a rotational flex around the Y axis.. Whenever the router is cutting toward the gantry it will begin to flex and hop the tool giving pretty rough cuts... I have to play with feed rates and router speeds to avoid the hopping. (Another problem I think Joe's router addresses)
#5 Stiff, Stiff, Stiff... Most of the extra $ will be dedicated to a new spindle and making the whole mechanism stiffer with less play....


My elusive goals for the #2 are..

100 ipm normal speed (Up from 40 ipm for this one)
Porter-Cable closed-loop controlled router.
.... Well many more, I'll start a new log for the second machine.....

I'll probably post another time with some final photos before I close this log for good!