CNC Beginner Fun

mdd0127

100 kW
Joined
Sep 30, 2008
Messages
1,084
This thread started as a for sale thread for my cnc mill because my shop was infected with computer demons that ravaged every computer I brought in there and frustrated me to the point of insanity. I ended up removing all computers from the shop for a week and it looks like the demons got hungry and moved on..... :evil:

I finally got a computer running normally and found some software I like so I've decided to keep the mill.

Some useful info and cool pics have developed in this thread and it might help others starting out with CNC.

I'd like to discuss e-bike parts CNC production and CNC beginner issues in this thread and move it to the proper section if possible.

More updates and video tonight.
 
I looked at ones of these a while back, isn't this sort of jeweler and engraver type mill? What cutting capabilities did you see while using it? Any photos?
 
I might have to keep the dang thing and just try to use it for a few tasks. I'm still up in the air on it. I spent a few hours on a single part on the manual machine last night and with a little over 70 parts to make, it's looking like a daunting task. If it can even get the cnc to engrave outlines for me to cut out on the manual machine, it should save a ton of time. I'm putting another new power supply in one of the old computers so I guess we'll see if I get lucky. My newest war in the quest for prototype has been the 10 degrees of backlash in the rotary table I got for the manual setup. I'm working with it but it's annoying. Someday I will find a machine tool supplier that doesn't send out stuff that isn't ready to go. :roll:

As far as the size of the machine, that's a question that I didn't get the correct answer to through research online. I just looked at the table size and travel and looked at parts others had made and didn't account for fixturing and tool size limitations......and the biggest thing, the hours of machining necessary due to machine size related feed rate limitations. It will make a 20 mm deep aluminum part with a 1/8" end mill running at 10ish ipm if you flip the stock over in the vise halfway through the profile and cut the part out from the back. It will turn bigger end mills but with a much smaller DOC. I've experimented a lot to try and speed things up and it seems to do the best with the 1/8. The drawback is the tool deflection at depth with such a small bit. I've seen all kinds of things built on these things and now that I have one, I know that complicated parts or deep cuts take HOURS on a machine this size. It will make bike sized parts and I'm a noob on the benchtop cnc stuff for sure but to be totally honest, this machine is better suited to RC, pcb's, plastic, or jewelry than bike parts prototyping. I really wish that someone would have told me that before I wrote the check but.......
 
Are you looking for help machining parts for your project? do you have any solid models or drawings of what you are trying to make?

You can see some of my projects and machining here:
http://www.grifftek.com/grifftek/Machine_Work/Recent_Parts/
the first few links are some projects mostly using rapid prototyping for my a2b metro, but starting about 6 links down are some machining projects.
regards

Bill Griffin
 
grifftek said:
Are you looking for help machining parts for your project? do you have any solid models or drawings of what you are trying to make?

You can see some of my projects and machining here:
http://www.grifftek.com/grifftek/Machine_Work/Recent_Parts/
the first few links are some projects mostly using rapid prototyping for my a2b metro, but starting about 6 links down are some machining projects.
regards

Bill Griffin

I wish that I had access to decent software and a reliable computer. I'd have some drawings. I've actually gone back to compasses, squares, pencils, and paper. It's the only way I can be 100% sure that my work doesn't evaporate into thin air. It's also really nice to lay them out and look at a few at a time instead of diddling back and forth between windows. I really appreciate your offer but I'm finalizing dimensions and revising parts as I go along, have spent my budget on tools and already spent over $400 getting machine work done on four parts. I think tomorrow is computer day again. I need to get my tools working!
 
Howdy mdd0127

I understand where your at, I have experienced the pain of using a similar tool, A MaxNC which maybe be a cousin of your machine, One of the engineers I worked for bought it for our department and within a few weeks he grew bored of it and somehow I ended up being the responsible party. Not sure how that happened it just did, I invested in a copy of Gcode 2000

http://home.roadrunner.com/~gcodemcode/gcode.htm

A simple to use DXF to Gcode converter, taught myself Gcode and went at it. I like you found any cutting tool larger than 1/8" was more than the machine could work with, I generally used 1/16" end mills and depending on the material took small (0.010") cuts, which isn't a problem when you set it and let it run in the background while you work on something else and just let it run round and round.
This worked fine for as you say, PC boards and small heat sinks, it wasn't practical for larger work. If I were in your position I'd continue to try to sell or trade it for something more appropriate, but continue to use it as a learning tool until you sell it. Just about everything thats CNC runs on GCode, perhaps a CNC plasma cutter or torch would be more like what you need to do 2D bike parts
Maybe www.cnczone.com forums might be a place to offer your mill?
Best of luck, the learning curve can be tough and low buck CNC can be buggy but the end results are worth it.

Greenerwheels
 
Quick update: It works! The the old computer runs mach like a champ and I finally found some decent software. I've been messing up parts! And even making some good ones! I downloaded Draftsight, Cambam and D2NC and absolutely hit the ground running with all three. I noticed that the code that these new programs make was using a different cutting strategy than g-simple, more like the way I'd go about it manually, so I was able to speed up the feedrate.....a bunch. Now it will hum along with a 1/8" end mill at 300ish mm/min at .5mm doc. I just need to figure out how to get odd shaped parts with no round holes or straight lines fixtured and located correctly!

Since I have finally found a way to get CNC to work for me, and there wasn't really any interest in it, I guess it's not for sale anymore.
 
Glad to hear you got your kit sorted out!

>I just need to figure out how to get odd shaped parts with no round holes or straight lines fixtured and located correctly!

Id be happy to help you with fixturing ideas if you need it. I have a lot of experience machining parts that are difficult to hold.
image288.jpg


image016.jpg


http://tinyurl.com/4qqzuco

http://tinyurl.com/4aku2ax

http://tinyurl.com/4q64beb

Let me know If I can give you any ideas, always happy to share my knowledge and experience.

Bill
 
The issues I'm having with centering are when trying to set the work offsets before running a program on parts that don't have any clear defining edges to reference.

For example, I got a split 65 tooth 219 sprocket from Azusa and wanted to open up it's bore, drill some bolt holes, and some lightening holes. The sprocket isn't split exactly through the center, and the split isn't a straight line. They cut the sprocket as one piece then just crudely shear it in half. I have a little laser alignment tool that makes a tiny dot at the center and concentric circles but my mill doesn't have enough z travel to make the circles bigger than an inch or so. For this part, optimally, I would have widened the laser circle to the tooth root diameter but it would just barely go big enough to pick up the one inch center halfish bore. I set the center the best I could this way but with the z height one part thickness above the part. I let the program run and paused it for measurement as it went around on the bottom pass, just over the part and it was off-center by a millimeter. I know at that point I could have just gone back to home, bumped the y back a hair, reset it, and run the program but I really wanted to figure out a quicker way to find center for the next seventy odd shaped parts that I have to machine from scratch or modify. My next idea, I thought was a good one. I taped a sheet of stiff paper to the sprocket half and used my drafting tools to locate a dot perfectly in the center. This process took a few minutes. Then I used the dot on the laser to pick up the center, ran the program at non cutting z height, and it was off by 2mm on the -y. Then I thought, maybe the laser isn't that accurate so I switched to a broken drill bit that I cut to a tiny point on the lathe and it lined up with the laser mark. Ran the program again, still off. So then I went back to cad, checked the dimensions, went to cam checked dimensions, regenerated code, repeated entire process above and it's still off. My next thought was that I was tired, possibly psyching myself out and over thinking things so I set z and hit run. It flew through the part but when it got to the apex of the arc, it was obviously off. I'll be cutting a fresh one from scratch out of a rectangular plate today!

So to solve the centering issue, I designed a little tool with a center degree wheel with cross hairs and a tiny hole in the center, and four graduated arms that can pivot around the center that indicate on the degree wheel. The idea is to be able to clamp the arms to a piece of stock at known measurements and angles, then just pointing the origin to the center of the tool, removing the tool, and running the program. I have it all drawn up in CAD and the profiles worked out on cambam but am stumbling around on the engraving part. Cambam imports it but doesn't recognize the origin for the engraving from the dxf. It puts it off to the side??? Anyway, what I'm doing today is trying to figure out if I'm about to waste my time building a tool that's already available,......or worse, isn't necessary at all because I'm missing something obvious. :oops:

So on that part in the second picture you posted, if you had bought that from someone else and wanted to write some code and mill some modifications into it, how would you locate the origin? That's a beautiful part by the way! What is it?

I really appreciate any pointers you may have!
 
Hey Mods,

Since there might be useful information developing in this thread, is it possible to rename it and stick it in the appropriate forum?

I can edit the first post to reflect the new nature of the thread.


Thanks,
cncnoob :oops:
 
The issues I'm having with centering are when trying to set the work offsets before running a program on parts that don't have any clear defining edges to reference

Can you drill a hole in the part, locate with a center finder, and use that as a reference?

-JD
 
On the parts I'm talking about, I'd still have to go through quite a process to figure out where to drill the locating hole in reference to the part's true center or whatever..... Maybe I'm over thinking that too though. Where I'm having the issue is getting the zero reference point of my drawing and code to correspond with a similar point on an irregular part that needs modified. For example on the part above, the reference point is the center of the sprocket which is not even on the half sprocket that needs machined. So in order to locate the center, I have to project it off into the air. It's really hard to describe...

If I'm cutting the parts from scratch, it's no problem because I can account for a centering/reference/origin point in the design and am usually cutting from a rectangular or circular shaped block.
 
Have you checked that the software has the correct TPI setting matching each axis? If the scaling factor is off then you need to correct that. If the part has an unusual shape, you could mill out a hole on a scrape plate to hold it and use the same reference to center the part.
 
Hi

The part you asked about is the tip of an exhaust pipe for an atv. if I was given that part, id figure out the dimensions of the mounting holes and then create a fixture that locattes off those holes

Here is the part I was given to reverse engineer and duplicate:
image002.jpg



More pics here: http://tinyurl.com/49kz3yl

if you could measure the geopetry of the outer shape, you coudl just machien a pocket with that shape in a block and set the part in the pocket and clamp it down. You woudl then know where the pocket you machined was located in ref to your fixture plate then you would know where your part was.

Id use this same concept for fixtureing yoru sprocket. Lets assume (usually a bad idea) that your sprocket is manufactured to some reasonable tolerance level and you know the pitch and tooth count. You know the roller diameter and can come up with a pitch diameter of the individual teeth

Make a fixture plate of suitable size to accomodate your part, then referenced from a corner add 2 or 3 holes to insert pins which will then locate the sprocket on its pitch diameter. from here you can then machien your features in reference to the pitch diameter (located off the corner of your fixture plate.
sprocket%20fixture.JPG


you can see a 3d model with integral viewer here http://tinyurl.com/4wylp4l

If you dont have the info, the manufacturer shoudl be able to give you the dimensions and possibly have it liisted on their web site.

Alot of times instead of making a fixture plate with holes drilled and reamed then inserting pins, i will just machine the top of the plate as a pocket with the pins as islands this way its usually a single piece fixture usually machined with one tool. On my equipment its very fast and easy, with a small machine this wont be the case as the volume of material to be removed will be high in comparison to drilling and reaming 2 or 3 holes.
This is a link to a simulation video showing this concept. Might not be proactical for you on this project but may be applicable for other projects.
http://tinyurl.com/45a4otx

Hope this was helpful

Bill Griffin
 
oatnet said:
The issues I'm having with centering are when trying to set the work offsets before running a program on parts that don't have any clear defining edges to reference

Can you drill a hole in the part, locate with a center finder, and use that as a reference?

-JD

Problem here is that how do you drill the holes (it will take 2) so they are in reference to the pitch diameter of the sprocket.

I havent seen the sprocket, but locating it from the pitch diameter or possibly the locating feature in the center of the spricket if it exists and is centered on the pitch diameter.

You will find that for many parts locating, and holding onto the part for the machining operation is the real challenge not the actuual machining/programing involved.

take a block of material, machine 85% of it away from 6 or more operations from every side and it gets tougher to fixture each time and you have to be careful you dont crush or distort it while clamping it:
image367.jpg
 
Moved thread to Ebike General from For Sale-Used, per new OP. ;)
 
Thanks AW, and THANKS Bill!

I'm definitely going to need to round up more material for fixturing. I need to be really smart about it though because my machine just doesn't remove very much material at once. I also haven't found a good source for scrap around here. I just spent another $100 at speedymetals for the chunk I'm machining the main swing arm out of!

I've been clamping stuff to my work holder plate(http://www.hightechsystemsllc.com/pt_plate_18_special.html) and it's really easy to get a reference from it. The holes are drilled .500 apart so it's easy to figure out where things are. I have a handy program called Sprocketeer and also have some charts from Azusa so next time I need to set one up it shouldn't be too hard. I ended up drawing the split sprocket from scratch today and machining the two halves in two operations. On the first half I had the doc at .5mm , plunge at 30 mm/min and the safe z height was at 25 for some reason so it took a couple of hours. I made changes to the code for the second half and ate dinner while it ran. The second half ran in 30 minutes or so at 1mm doc 350-500mm/min feed, 100mm/min plunge and a 1.5mm safe z due to a better clamping scheme. I got some short videos of both. I had to get creative with the feed hold, and moving some clamps around on the first part too but worked that out on the second part so I could do some drawings while it ran. They came out pretty nice.

Tomorrow I'll be turning some stuff on the lathe, possibly a little colossus motor so I'll get a short break from fixturing challenges.

I'm going to build the little 4 armed centering tool I designed over the next few days too. I got a little pointed engraver bit but haven't had a chance to try it out yet. I'll post some pics of the tool and if anyone wants to make one, I already have the cambam files done so I could just send the code.

Thanks again to everyone that's contributing here! I have a LOT to learn and definitely appreciate the pointers!
 
hey MDD,

Does your machine "Home" off from switches?
or are you touching off the table & parts to set the zeros.
I have found much frustraition is releived if you have a reliable homeing switch on at least the X & Y when it comes to hobby sized cnc stuff.
especialy if you have fixtures that get changed on & off the table.
 
Hey I have that holder plate as well, worked out a deal with the owner since he needed some of our boards. Here is a link to my CNC conversion, we kept making it bigger till it's now quite useful for most of my needs. http://www.embeddedtronics.com/micromill.html The original cost of the mill was under $300 from harbor freight, I don't see them selling it anymore. The amount of changes we did to it, I would recommend the bigger X2 mill now. I have seen that one go on sale for $500. There are lots of info on the web to convert it into a CNC. The micro mills only drawback is the column support isn't as stiff as the X2 once we added the bigger base.

If you scroll down the page on my site you will see the same plate holder, works a treat. I still rely on the vise for most of my work though.

For a good source of metal, go to your yellow pages and look for scrap metal dealers. The one I use in Mi. sells it by the lb. The cost last time for aluminum was 1.50/lb. I can most of my aluminum from them.


mdd0127 said:
Thanks AW, and THANKS Bill!

I'm definitely going to need to round up more material for fixturing. I need to be really smart about it though because my machine just doesn't remove very much material at once. I also haven't found a good source for scrap around here. I just spent another $100 at speedymetals for the chunk I'm machining the main swing arm out of!

I've been clamping stuff to my work holder plate(http://www.hightechsystemsllc.com/pt_plate_18_special.html) and it's really easy to get a reference from it. The holes are drilled .500 apart so it's easy to figure out where things are. I have a handy program called Sprocketeer and also have some charts from Azusa so next time I need to set one up it shouldn't be too hard. I ended up drawing the split sprocket from scratch today and machining the two halves in two operations. On the first half I had the doc at .5mm , plunge at 30 mm/min and the safe z height was at 25 for some reason so it took a couple of hours. I made changes to the code for the second half and ate dinner while it ran. The second half ran in 30 minutes or so at 1mm doc 350-500mm/min feed, 100mm/min plunge and a 1.5mm safe z due to a better clamping scheme. I got some short videos of both. I had to get creative with the feed hold, and moving some clamps around on the first part too but worked that out on the second part so I could do some drawings while it ran. They came out pretty nice.

Tomorrow I'll be turning some stuff on the lathe, possibly a little colossus motor so I'll get a short break from fixturing challenges.

I'm going to build the little 4 armed centering tool I designed over the next few days too. I got a little pointed engraver bit but haven't had a chance to try it out yet. I'll post some pics of the tool and if anyone wants to make one, I already have the cambam files done so I could just send the code.

Thanks again to everyone that's contributing here! I have a LOT to learn and definitely appreciate the pointers!
 
Derek at Hightechsystems is a nice guy and hassomenicesmall mill accessories not to mention his EV battery skills!
I like the trunion table with the rotary table setup for 5 axis.

I have a small mill with fixture plate that has reamed holes for dowel pins and threaded holes. I made up a poile of plates from barstock that have matching pins and c-boredholesto mountthem to the plate. you just screwthem down,machine them or your operation mount the parts and go. its really ice when the job repeats, just bolt itdown, set the offsets and go again.

You can check out this machine at http://tinyurl.com/4jz66sr

image019.jpg

You guys that are using mach3 might like to see a machine I designed and built that uses mach3 and its probing ability to assemble a dual head credit card reader to maintain a 0.003" gap between the heads.

http://tinyurl.com/4cvlfcd

For materials fo rou guys with small machines, you might find a metal cutting service and see if they will let you periodically peruse their scrap bins / barrels. The remnants they throw away are usually big enough for making parts on your small machines. I recently had some tooth belt pulley stock sawn into blanks and while at the saw shop I noticed lots of nice sized pieces of thick barstock in the scrap bin in both 6061 and 7075. The owner was nice enough to sell me material at a $1.00 per lb. I came back to the shop with 50lbs of nice plates which I have turned into several fixtures adn some parts. all for a fraction of the cost of these paterials from my normal source mc master carr

I have a friend with shop that does some sheet metal work and I have visited his scrap bin many times and have come away with nice pieces of sheet metal and peices of bar stock cutoffs from the automatic saw. they wrap a bunch of bars together with shrink wrap and saw it to length andthen there is a piece at the end that isnt long enough for the job so they just scrap it. Nice materials for cheap! At my day job I keep anexe on the scrap bin where they throw all of the obsolete parts. I get nice shafting, pulleys, gears, lead screws, steppers, stand offs, hardware, components etc. I think they get a penny a lb for these materials so they dont care what happens to them and there have been many times where they have needed stuff that I have saved.

regards
Bill
 
Well I'm glad I found this post... Cause I'm really new to the whole idea of CNC part making.... as in never done it, but might need to get some parts made in the near future.

Now I'm a graphic designer by trade, so I'm very comfortable in Adobe Illustrator -- which is what I've used in the past to make diagrams. And last time I used AutoCAD was like 15 years ago in high school. So I downloaded the trial of it... and good lord... it's a hard for me to get used to it.

What are you currently using to design your parts?


And also... for something like a triple tree fork crown, what would you recommend for the material? Probably aluminum, but what type?
 
Go download Draftsight and start drawing. I have a degree in CAD and couldn't even draw a sprocket in every other program I tried. I downloaded draftsight, drew a circle and a few layout lines, hit the array button and it was done. From download to finished sprocket in under 5 minutes. I messed with Alibre and Dolphin for days and couldn't figure out how to do anything....at all.

Once you have your drawing, you'll need CAM software to make the code. Gsimple is free but can definitely be an exercise in frustration because it has some bugs that tell you stupid things like an 1/8" end mill can't pocket an 8mm hole. Cambam is very nice and they'll be getting some of my money for sure. You can use it on a trial basis 40 times then it's $180ish. I've made fifteen drawings and used 5 of the sessions up and am ready to send them the money so I can keep it forever! Another cool one is D2NC. I haven't had time to play with it much because as soon as I figured out Cambam worked like a charm I got really busy making parts. I really need to check it out though because I think it might be better than Cambam.

If you have other questions as you're getting into it, feel free to post them here. I've been doing research and trying to get rolling with this stuff and there's a ton of info out there but it's scattered all over the place. Cnczone has been an absolute necessity too. Lots of very smart guys over there for sure!
 
I learned two important things yesterday that I thought I'd post.

1. D2NC looked really cool in the video but I opened it up last night and couldn't get it to do anything.

2. Even though my controller and mill setup is DSLS, Digital Sync Lock Servo, it will lose it's zero reference if a "limit switch" is triggered. I thought that the rotary encoders told the controller where the actual position was in relation to the commanded position and used that to sync up. All they do in reality is act like limit switches, stopping the program if something gets out of sync. In my case, the rubber flap that covers the y lead screw bound up and tripped the reset. I got it unstuck, hit goto z and it looked like it went right home. Then I restarted the program from where it stopped and the x axis was 2mm off. I caught it very quickly, reset the offsets, and was able to save the part but learned a very valuable lesson. You have to reset the zero reference point/work offsets if the program is stopped because of a crash.
 
What you are ecperiencing is one of the disadvantages of the type of system you have.

To get around this you can home the machine to the limit switches (assuming this is a repeatable home position) from this zero location you then find your fixture offset. I just write this number dowy (XYZ) then reset the home position to this location.Then when you have a problem, you can home the machine then do a MDI move to your fixture offset then clear the counters.

You can also figure out how to set your fixture offsets from the homed zero position but I always found that to be confusing on my ahha stepper systems so I just used the procedure above. I would like to think that its more straightforward with mach 3.

My only experience with mach 3 is on a machine that isnt related to a metal cutting operation using the probing routine and positioning.
http://tinyurl.com/4cvlfcd



My centroid machien dosent have this issue, it even does a encoder count check from the overtravel switch to the encoder marker pulse and if it is off compared to the last stored count it will pop up a warning letting you know there is a potential error between the home position and your fixture.

Does your servo system move to the limit switch tehn home to the marker pulse on the encoders? does it have encoders with marker pulses? if it does, you coudl write a macro that could do this automatically for you.

Bill

mdd0127 said:
I learned two important things yesterday that I thought I'd post.

1. D2NC looked really cool in the video but I opened it up last night and couldn't get it to do anything.

2. Even though my controller and mill setup is DSLS, Digital Sync Lock Servo, it will lose it's zero reference if a "limit switch" is triggered. I thought that the rotary encoders told the controller where the actual position was in relation to the commanded position and used that to sync up. All they do in reality is act like limit switches, stopping the program if something gets out of sync. In my case, the rubber flap that covers the y lead screw bound up and tripped the reset. I got it unstuck, hit goto z and it looked like it went right home. Then I restarted the program from where it stopped and the x axis was 2mm off. I caught it very quickly, reset the offsets, and was able to save the part but learned a very valuable lesson. You have to reset the zero reference point/work offsets if the program is stopped because of a crash.
 
Is there anyone out there who is serious about building or retrofitting a machine for cnc? I have a lot of surplus parts for building machines, servos, steppers, linear rails, ball screws, acme screws, lead scews with super nuts, linear actuators etc.

If your serious and want to discuss it, please email me at grifftek @ grifftek dot com and we can work something out. If your in SoCal, that would be even better.

Regards

Bill Griffin
 
Back
Top