Pumelloman Solysylva 25x37 Router Build

[thkoutsidthebox]

Make sure that your tpi for your Y axis is set correctly. This being incorrect is the most likely suspect for the Y axis not drawing your circle correctly.

[pumelloman]

Sorry for not responding for awhile, school is starting up pretty soon...

So I did go in to fix that Y axis, and I doubled the number of "steps per unit", and now it draws a circle!!!! I was so happy. Then I proceeded to use the pen attached to the CNC to write text (which worked really well) from Mastercam.

So I got a lot more confident and decided to make some parts for my CNC, namely remaking parts that I didn't like so much. Stepper motor mounts were the first thing I wanted to remake.

However, it seems that there are still some problems with the X axis....Y axis is fine, and Z is "generally" fine. The problem that I'm experiencing goes something like this... After running the machine for a while, then having it stop for a while, I re-run the same G-code, and it occasionally locks up the motors and makes a "EEEEEEEE" sound, while Mach3 says that the motors are still moving in that axis but obviously they are NOT.

This is pretty detrimental to making things, considering its hard to tell when this is going to happen (since it's not consistent). If it does happen, then I have to stop it immediately before it buggers up my work. If I try to restart my G-Code then i have to re-align everything, otherwise it's going to be moved a certain amount in whatever axis, off of what is correct.

Not sure how to fix this, but I suspect the binding of the X-axis lead screws or rails as the cause of this problem. Z axis is the Leadscrew problem, also alignment.

Unfortunately I was not skilled enough or careful enough to make this properly the first time....and I don't have very much time the second time around...
I would definitely like to solve this problem somehow so that I won't have to worry about the machine screwing up whatever I'm working on and making it No Good....


Other than that, I have to remake the part of the Z axis that holds the motor, which was MDF (NOT a good choice for material strength...I put screws in it and they are ripping it apart every time i want to loosen the collet on my router...). Would definitely like to make that out of UHMW, as well as the rest of the Z axis.

Might work on it tomorrow if the soul is willing...

[CarveOne]

pumelloman,

That is happening due to a little more binding in a particular area of travel or it's more likely in the stepper motor settings.

Try slowing down the inch per minute feed rate a little and see if that clears it up. You may find "gaps" in the IPM range where it may squeal at a lower IPM but not at a higher IPM setting. If you have not made an attempt to "tune" your machine in software yet, you should do it. By trial and error you can at least minimize the squealing. Loud resonances are not good for the motor bearings. The acceleration parameter is one of the settings that affect the resonances. Slow that down a little and you can either run at the same IPM or possibly raise the IPM before squealing starts again.

When my friend was here last week he was able to get my machine to run at 75 IPM but it would squeal like a stuck pig in one direction but not in the reverse direction. We needed to run it at around 30 IPM to stop the resonance problem. My 425 oz/in steppers will resonate unless I make and install some vibration dampeners. When I have the dampeners installed the machine should become more stable over the IPM range. The Solsylva site once had a page that described the dampener construction details but I didn't find it last weekend. I may need to contact David Steele about that if I don't find it.

CarveOne

[thkoutsidthebox]

If you dont have time to physically fix it, you can lower the speeds and reduce the acceleration until it doesn't happen anymore. This will fix it. Remember that your machine is not a $50,000 beast, so if you have to back off 10 inches per minute to be able to trust it not to miss steps thats ok.

[BobF]

Originally Posted by pumelloman However, it seems that there are still some problems with the X axis....Y axis is fine, and Z is "generally" fine. The problem that I'm experiencing goes something like this... After running the machine for a while, then having it stop for a while, I re-run the same G-code, and it occasionally locks up the motors and makes a "EEEEEEEE" sound, while Mach3 says that the motors are still moving in that axis but obviously they are NOT. ...

If you are using the hobbyCNC board, turn off current reduction. (pin 4 I think).

[pumelloman]

Thanks everyone,

I had the same thought as you Bob, because it would often lock up after coming out of current reduction. Then when it had run for a while it would no longer lock up as often. I disabled the current reduction on the X axis, but perhaps I should try the Z as well. It helped a bit on the X axis but did not completely solve the issue.

I also reduced the acceleration to 6 (in/s^2? or whatever units mach3 is set in..confusing imperial system...) which helped, since before it was at around 14. But the problem still persisted. Perhaps I have to lower it more?

CarveOne's comments on the resonance problem definitely hold true as well. I noticed that a lower speed didn't always mean that it would run "better", in fact my machine would run smoother at 50ipm than 40ipm, or something like that, once it got going. But starting out it would jam more often.

Another reason could be that I re-tapped the Y-axis by hand when I was initially installing it, (I think it's Y axis, maybe I should look over my posts again), but I didn't re-tap the X axis or the Z axis, which definitely makes them MUCH more "bindy".

I like the idea of vibration dampeners, ...
OH MY GOD!!!!
http://www.solsylva.com/cnc/damper.html
Is this the link you were talking about?! That sounds exactly like my problem! GAH!
More stuff to make now, and I don't have a lathe so drilling that hole axially might have to wait until I go into the machine shop at school. Until then I guess I will just have to tune the machine like CarveOne said and reduce the acceleration a bit more.

[BobF]

pumelloman,
Look at my thread also about the 1/2 step microsteps. I made both changes at the same time, and I don't have issues anymore. Not sure which helped the most.
http://cnczone.com/forums/showthread.php?t=43485&page=5
Post 56 is the link I found that helped.

[CarveOne]

Originally Posted by pumelloman I like the idea of vibration dampeners, ... OH MY GOD!!!! http://www.solsylva.com/cnc/damper.html Is this the link you were talking about?! That sounds exactly like my problem! GAH! More stuff to make now, and I don't have a lathe so drilling that hole axially might have to wait until I go into the machine shop at school. Until then I guess I will just have to tune the machine like CarveOne said and reduce the acceleration a bit more.

Yes, that's the one. When I tried to print that page only one page of it would print and I had to select the text and photos I wanted and paste it into an MS Word document to get it all to print correctly formatted on four sheets. Build the last example shown. I will do the same this weekend. I would be happy to modify some bolts for you on my lathe while I'm doing mine and send them to you but it would probably be more hassle to mail them to you than either of us wants to go to. I'll let you know if I find the springs at the local Lowes store. Most any hardware store should have something that will work. I think the springs should have flat ground ends to work properly, but the washers may allow plain cut end springs to work also. I expect to mail my Xylotex board back to them for repair/replacement today. CarveOne

[pumelloman]

Thanks everyone,
I will definitely try the switch down to 1/2 stepping and disable current reduction. If that fails I guess I will be in the machine shop next week attempting to make those vibrational dampeners, (I should probably make them anyway).
As was posted in BobF's thread, this appears to be related to my problem as well: http://www.cnczone.com/forums/showthread.php?t=47594

Hopefully that does the trick, won't have time to test it for a while though.

[pumelloman]

So, I got it working fairly properly today. I did what BobF suggested (turned off idle current reduction, and switched to 1/2 microstepping), and it eliminated most of the resonance at anything below 40ipm. I will make the Vibrational dampeners later when I'm in school, so I can run the machine at higher feedrates without the resonance issue.

I've been doing 0.050" passes at around 20ipm with pretty good results. I think my machine can do more but my router is not on there very securely.

I also tried melting down the file-a-wax today, but I didn't get the results that I was looking for (likely due to incompetence on my behalf). When I poured the wax into a metal container lid to make a flat piece to do some example cuts on, it ended up warping due to uneven cooling. Also for those who are interested in this, using the "Double Boiler" method to melt the wax (ie, putting a pot with the wax in it, into a pot of boiling water to control the temperature), does NOT WORK. I'm not sure why, perhaps the heat transfer was inefficient and didn't get up to the necessary temperatures (I thought it was 60 degrees C...) so it did not work at all, no melting.
However, putting the pot with wax in it on a small stove element, and turning it to around "2" (on a scale of min - 1 -....- 8 - max) worked pretty well. I heard that getting the wax too hot is very dangerous (wax shot at me today, wear GLOVES and Eye Protection), and the wax can also burn. Takes a LONG time (40+ mins) to melt down the whole block, by the time I figured out what I was doing.

The part I cut out today (for those of you who aren't building the Solsylva machine) is the Z-axis Stepper Motor Plate/Mount. I was overall very impressed with my CNC, it maintained dimensional accuracy to 0.010"-0.015" or so in the Y axis, but the X axis was around 0.025" short over 3.125", or more. What's more interesting is, when I hold the part up to some other buggered up parts that I cut out with a 1/4" bit, the other ones are slightly longer in the X direction......

Another strange thing that's happening, that you can see in the second photo, is that each cut goes a little bit farther off, so the overall result is that the cut appears to be "on an angle". Not completely sure what the reason for this is, it could be non-securely fastened CNC table, which is quick-gripped to the desk, or it could be (more likely) that the router is not on the Z-carriage securely, and can wiggle a bit and flex. This is also causing the MDF part holding the router to break, as mentioned before.

You can also see some writing I was talking about earlier, that I used my CNC to do with a Pen.

Oh, another interesting point. In the 1/4 microstepping setting, the HobbyCNC board does not do the Y axis correctly unless it's set to take 2x the steps per unit (ie, 1/8 microstepping), but in 1/2 microstepping it takes just the same as all the other axis.....maybe I have a jumper that's not soldered quite right...

Enjoy the photos! My Uncle is going to try to find some plastic for me to machine the router plate out of. Other than remaking the entire Z-carriage, I'm just starting to think of some things that I can still make with my CNC without having to worry about the 0.020" off in the X axis.

[CarveOne]

pumelloman,

Check your gantry for squareness with the x axis rails. It's possible that the x axis lead screw threads on one side is advanced by one or two threads relative to the other side which would cause the gantry to be skewed a little.

Fine tuning the machine for perfection is going to take us a while and will require occasional realignment. I did some of that this weekend to remove some minor freeplay and misalignments. I have not cut a real part so far.

At one time I wished I had drilled a hole and mounted a brass threaded insert in the bottom of each leg while I was making them so I could use 1/4-20 flathead bolts to mount the machine to a piece of MDF as a substrate. What I figure I'll do now is to do something simple like glue mitered pieces of 3/8" square oak to form "frames" on the MDF substrate for the legs to sit in This will prevent movements due to cutter side loading and will allow the MDF to be more easily replaced. The weight of the machine should prevent vertical movements of the legs.

CarveOne

[evilbetty]

that was a great read thanks for posting up al that info and pics,im about to attemp to build the same router trying to soak up as much info as i can at the moment great pics ,i hope it all finaly works well for you