Separate names with a comma.
Some features disabled for guests. Register Today.
Discussion in 'CNC Mills/Routers' started by KerryH, Feb 24, 2015.
Discussion in 'CNC Mills/Routers' started by KerryH, Feb 24, 2015.
My adventure in CNC routing.
KerryH published a new build:
Read more about this build...
First things first, due to some items being out of stock at the open builds store, I decided to hand make a few parts.
First up was my take on the 90 degree joiner plates:
Next up was making a plate to attach a nema23 to the Z-axis. I utilized the nema17 plate for the bearing recess.
Now this got my machine going, and I wanted to do some testing with a pen.
Decided to mount the spindle on it's temporary mount (Nice strong stainless steel hose clamps) and do a test cut in wood.
Made this for a friend:
Then made this for my brother's garage:
This all leads me up to my current state, I wanted to move into some acrylic and aluminum work (seeing as I need a more permanent spindle mount) but I'm having an issue with the backlash of my machine on the X-axis.
My test procedure consists of moving the axis towards the dial indicator
Setting zero on the indicator
Move axis 1 inch looks accurate
Move axis back 1 inch, note backlash
On my Y-axis (dual motors, front to back) I am seeing about .002" (0.0508mm) of backlash which I consider to be acceptable.
On my X-axis, I am seeing anywhere from .008"-.012" (0.2032-0.3048mm) of backlash which doesn't sound too terrible, but when it comes to boring a 5mm hole the left to right is about .5mm undersized because of this.
Now I have disassembled and reassembled the machine numerous times in an attempt to ensure square, level, everything is tight etc etc chasing this issue.
I even made it a point to file a flat on the stepper motor shaft for the second set screw on the pulleys in which I tightened very tight and used loctite on the screws. The pulleys are practically part of the steppers now.
I even had a piece of 1-3/4" maple planed by a woodworking shop to ensure a perfectly flat work surface in which I attached the frame to w/ angle brackets and screws.
I checked the accuracy of my electronics today by telling the motors to complete one full revolution and the results of the x-axis were as follows:
So as of this point, I believe that I am dealing with a mechanical shortcoming of my machine. I'm not sure what to address at this point though.
Have you done any motor/pulley/belt swapping? It might be a good exercise to do some substitution with all the drive components to see if the problem follows a particular component or is just always there.
This is definitely something to try, I will swap my x-axis stepper with one of the y-axis stepper just to make sure I don't have an issue there.
I highly recommend the follow very simple and inexpensive mods.
Bolt the two X axis extrusion together
Y Axis inner gantry plates to reduce flex in the wheels and bearings
Better corner brackets
Steve these are all great looking pieces. I believe I have the corner brackets covered already.
Regarding bolting the two x-axis extrusions together. I have seen this mentioned numerous times, but when I have my gantry on the dual extrusions and run the gantry end to end, it seems to prefer a very slight space between the extrusions. I thought this was odd, but I am using all spacers from openbuilds and the correct order according to Mark's videos.
Has anyone else noticed this? Did you just bolt the two extrusions together anyways?
Sorry, I did not notice your nice corner brackets. My eye was drawn down to your nice Norton motorcycle test image.
Yes, I just bolted the two extrusions together. I very tightly clamped them together and drilled taping size, then bolted them together. I found this really helped the wheels run in a consistent size groove and prevented racking and twisting.
I was using the spacers recommended, but the X axis never felt good along it's travel. After this mod was completed it solved much of the play. Sorry I cannot remember the sizes of the spacer used in the center.
Ah so you used different spacers between the wheels?
From memory yes, because the original did not fit perfectly... What I do remember was taking it apart many, many times trying to get the best fit.
What I also noticed is the two separate beams & small 5mm bolts have enough flex to smooth out inaccaracy at the cost of overall stiffness. when the beams are perfectly joined together the spacers must be a correct fit!
Thank you for that, I will add it to my "to-do" list!
Your machine looks amazing btw, very jealous of the quality of parts!
Edit: after looking through the models from your build, I can see now that when the x-axis beams are sandwiched, the distance between the wheels should be about 9mm, vs the 9.35mm of spacers the build originally calls for.
KerryH, you can also use the profile connectors you can find here http://www.openbuilds.com/resources/v-slot-connector.40/. I used them for my X axis and I'm very happy with the stiffness I achieved with it. If you have a 3D printer at hand or near you, give it a try.
Thank you Kerry, it was lots of fun building, sourcing and designing.
My computer model of the wheels & extrusion made a fundamental assumption that the bottom of the wheel runs precisely parallel with the bottom of the V-Slot. The original 3D CAD models of the extreme wheels & the extrusion have slightly different angles, so SolidWorks refuses to allow you to put a tangent constraint to both sides of the V. So I assigned a tangent constraint to one edge of the V and a co-linear constraint to the bottom of the V & the circumference of the wheel. I don't know the manufacturing tolerances of the extreme wheels, so I decided a practical approach of testing was best.
I was luck to have some spare metric & imperial spacers to play around with different combinations.
Prauk I love the profile connector approach! Funny I would have done both approaches if I had read this post earlier! It's shocking how much the two beams flex when they are not secured together as one!
Prauk that is great, I do not have access to any machine locally that I know of, but I am comfortable drilling and bolting the rails together.
When you start doing that, make 100% sure you have the v-slots perfectly aligned with each other and they can't move while you're bolting them together. Use clamps I guess?
KerryH, I have now published the inner plates I designed in the resource section.
Enjoy if you decide to do this mod...
Well it's time for an update, I've been very busy with other projects so it took a while to get back on this one!
I spent some time being EXTREMELY meticulous with the disassembly and reassembly of the machine.
@SlyClockWerkz very graciously provided a new set of gantry plates to replace the set I had due to less than favorable quality and accuracy in the machining.
I also removed ALL my v-wheels and measured the diameter of every single one so that I could pick which wheels go where for the best fit.
I also took the opportunity to bolt my x-axis extrusions together and refit different spacers to the x-axis gantry wheels so that they fit properly.
I also fitted a drag chain setup to handle the wiring.
I'm still not SUPER happy with my results, as I'm still seeing about .005" of backlash on the x-axis, but I manipulated my settings in tinyg to offset that as much as possible.
Needless to say, I'm ready to start cutting parts instead of chasing every thousandth of accuracy.
I started off w/ some tests in MDF since it's most forgiving, went to some tests in lexan, and then finally cut some aluminum today!
I seriously messed up cutting the logo in one program and attempting to cut the holes from another program and screwed the alignment all up, but the logo turned out nice for a vbit that I accidentally knocked the tip off while jogging one day, so I ground it flat and made it a flat tip vbit.
So I learned from the misalignment issue and made another attempt, successfully this time!
Unfortunately it's not without it's faults, I screwed up yet again and entered a number in the wrong field which caused the engraving which was supposed to be three passes but instead ended up being one full depth pass! it was a little too much for the machine to handle which can be seen on the bottom left of the A.
Anyways, I'm very happy and excited to have my machine in good running order so that I can start using it!
Congratulations Kerry, very pleased to see your tests.
Interesting to see the wheel size difference, this is worth checking on my machine next time it comes apart.
Ya, the wheel sizes surprised me a bit as well. I made sure to take 8 of the 23.94mm size wheels and use them as the top four on each gantry plate.
I still have many upgrades for my machine in store including inner gantry plates like you designed now that I have the ability to cut some myself.
Good to see youre up and running!