Not sure of what you mean by "the source for the holder" other than the plastic clamps in the link I gave you. I didn't test the spindle yet, I just got it, but according to @Robert Hummel , the spindle shouldn't be hotter than warm with the watercooling on while running, so plastic should handle it very well.

 

Wow, still struggling with the y stepper banging/rat-a-tatting. Tried everything recommended, accuracy not really a problem. Running 1/8, any ideas welcome.

xPro v2, 750mm x 750mm, NEMA 23s on x and y, NEMA 17 on z.

 

@Dave Hylands This is the first thing I checked.

I've since filed a flat for the second set screw so both are seated on a flat and used loctite on the set screws.

They are all solid to the steppers and the problem on the X-axis persists.

 

Have you tried adjusting the current for motors on X and Y axis? Even your x-axis seems to be a little loud - I had the same problem but when I played around with the current (decrease or increase), you get very smooth movements...

Try increasing or decreasing current by a small amount each time and run the machine for short distance (so that the motors/board doesn't burn out) one axis at a time! You should notice a difference in the motors running. If the motors run louder when you increase/decrease the current, turn the knob the other way and keep adjusting current until you hear the motors run smooth... Hope this helps..
Is your board getting hot even on the small runs?

 

Messes with the current just cannot find the sweet spot. Will continue.

Only ran one test. Gets hot when on 24 volt, so went back to 12. Not sure why voltage makes a diff. Still tuning. Thanks for the encouragement.

 

So a bit more on my build, the Y plates I got from vslot-europe are too big, big enough that even when I eventually get the eccentric spacers, the wheels won't touch! That's a pretty serious problem :/
So unless I hack these plates somehow, pending replacements from somewhere, my build is stuck. this si SO annoying as it's the sort of things I'd /check/ against a template before shipping really. How hard it is to do that? It's the /key/ component of the whole build, and I don't mind doing a little bit of filing here and there, but here theres a good 1.5mm off from my measurement. I can't even mount the motors on these plates.

 

Yes its the 1/4 fan fold foam sheet. And it is a fuselage. I set it up with tabs that hold it together until the glue dries then you break them off. Using 1/32 and 1/16 bits from drillbitsunlimited.com. Works good but sometimes the bit heats up and forms a hard ball of plastic which ruins the part. I also use this foam to test code before cutting more expensive materials. These bits work good on balsa and thin ply too.


Are you into RC planes.

 

Yes I am into RC planes too I tried cutting it with drill bits and I had the same problem of bit heating up and formation of hard foam on the bit.
I have ruined a lot of foam too until i found the tile cutting bit from rotozip (very cheap!). It has no issues of heating up and works awesome - clean cuts! I don't ruin my foam anymore but the only problem is the bit size I found was 1/8th inch. I am looking for similar bits down to size 1/16th or even 1/32 if available.. With software, I use a mac so cannot use cambium but I have been playing around with SketchuCam and looks like it cannot cut a straight line with (outside or inside cut) offsetting for the bit size - it needs a loop for doing so... Still trying to figure out whats the best option for it - any recommendations?
I also have a 9mm 445nm laser attached so that I can cut balsa and thin lite ply with it. I am still experimenting with it...

 

Let us know if it works out for you - Good Luck

 

Hey guys,
I started making aluminum ox plates for an affordable price. They are laser cut, and very accurate. The 'cut it forward' group on google plus has inspired me and I would like to do a 'group buy' to try and make the ox more attainable to the people on the fence about the build.
The target price is 70$ usd shipped (US). International buyers are welcome as well, they just need to help out with shipping.
I would like to get 6+ commitments for sets before I order. Start a conversation with me if you're interested!

http://clockwerkmech.weebly.com/ox-cnc--aluminum-plates.html

Sign ups so far: (4) to go. Once I get 6 commitments I will start another batch.
1. Dangasaur
2. Dangasaur

 

Sign me up for 2 sets. I'll also send a PM.

 

Tried cutting with 80ipm (2000mm/min) settings - works great - Thanks! Although I think the speed needs to be cut in half (or may be 60ipm?) if using dollar tree foam board with paper on both sides...
For other looking for details, I am using Dremel 3000 with a 1/8" rotozip tile cutting bit.. Soon try some 1/16" and may be 1/32" (if I find any) bits..

 

Hi mate,

I am completely new to CNC world, decided I would build a machine to build electric guitar bodies (the Les Paul is especially complicated to do by hand ;-) ). I tried designing a machine, went back to check on builds, available material in my part of the world (Switzerland), started designing again, and stumbled upon your build..... This is astonishing. I believe that one should do what he is good for, I've never designed a machine, would love to do so, but in the meantime I am just going to build yours.. I mean, it looks perfect. Thank you for sharing this with everyone, this is what makes Internet (and computers..) fantastic.

You rock

Take care
Cheers

 

xPro and speed control enclosure. Posted 123D and .stl. on Thingiverse. I'm using xPro V2 and spindle control sold by Spark-Concepts. Had to move the xPro board so I am reprinting now, but you can see pictures on my build page. USB cord interfered with the controller board (I swear I tested this...).

Speed Control -- had to recompile with option enabled. Will replace speed control knob with USB port when print is done (11 hr print). Designed to be under table. Control box next.

http://openbuilds.com/builds/ox-cable-management-electronics-case-e-stops.1810/
http://www.thingiverse.com/thing:694369

 

Looks like backlash. Check if your software has a possibility to correct for that error. I had to correct my axis just a little bit to get them "dead-on" (less than 0.02mm).

 

I'm not sure what tinyg has in this regards. I'm becoming quite frustrated though.

I have disassembled and reassembled my machine many times now in an attempt to assure things are as perfect as I can make them. I even went very far out of my way to have a 1-3/4" piece of maple planed by a woodworking shop to ensure that I was building on a perfectly flat surface.

I've attempted to utilize the dual belt setup and it doesn't seem to make a difference (and really didn't work out as well as I had hoped).

I went with larger 269 oz/in steppers assuming that larger shouldn't give me any issues but at this point I'm open to anything.

I ordered from the openbuilds but is it possible that I have low quality belts?

I'm just at a loss especially when hearing about fantastic results from others.

 

I'm thinking.... No.

 

Sorry to hear that you are still having problems... When you disassembled your machine, have you tried checking the distance travelled by motors (or revolutions) without any load on them?

I think you could just loosen the belts so that your motor is no longer connected to the belt, then try to give commands such that the motors will do 1 revolution or 5 revolutions and so on... in both directions! (calculations shown in video below)
See if you get your motors to do that multiple times accurately with no load! Try multiple numbers of revolutions in both directions...
I am not sure how would you accurately measure the revolutions (not sure if the device you are using can do it?)
Check this tutorial out (skip to 8:40):
if the motors are moving accurately then it may be the belt? - but this step will at least make sure that your electronics are working fine!

 

@KerryH, the dual belt setup will decrease the amount of backlash simply because there is less stretch due to the shorter length of belt being engaged in the process. I glued the lower belts into the v-slot profiles with power glue because it doesn't take up the space like tape would do. But even with the dual belt setup, it still is a belt driven system and some minor stretch will always be there. For that you can use the backlash option (if your software has that function) to compensate the stretch you measure.

To measure the backlash; put your dial indicator in line with the axis you want to measure. Move the axis so that it just touches the dial probe and you get a tiny reading. Then move it in the same direction (DO NOT REVERSE!) with 1" by giving the command to the software to move 1". After this move you set the dial indicator scale at zero. Then command a 1" move in the reverse direction. You should have now a move of exactly 1". (All this, provided you have set your steps/inch perfect). Anything less is backlash and you have to correct that within the software (it should have some function for that). Lets say the reverse move gives you a result of 31/32". Meaning your backlash is 1/32". That is the amount of correction you should use for the backlash correction function.

It is of paramount importance NOT too move the axis back and forth while you measure. Move in one direction only, set the dial indicator to zero and only then move back!

Checklist hardware:
1. Is the stepper motor moving? Tighten the screws that hold them.
2. Are the pulleys tight to the stepper motor shaft? Make sure they are.
3. Are the lower belts glued in smoothly, without any wrinkles? Make sure they are.
4. Are the upper belts not too tight? Make sure they aren't, but also not too loose or they will skip over the pulleys.
5. Are the wheels too tight or too lose? Correct it. Check Mark's video for how to do that.
6. Disconnect all power to the motor drivers and controller board and move the axis by hand, checking they're running/moving without binding/stalling or other interruptions.

If your software doesn't have the function of backlash correction, you might try to recalculate the steps/inch incorporating the backlash you have measured. Test the new setup and keep adjusting steps/inch until you get the desired results.

Hope this helps a bit and good luck with the tuning.

 

This is a good point and I hadn't thought about it. I will give it a shot though to make sure nothing is out of the ordinary.

These are all good points. The reason I said the dual belt setup doesn't seem so great is that the space underneath the wheel where the two belts mesh seems to be very particular. Meaning when I first used too thin of a tape the mesh was not very good, ie the top belt could move. I used thicker tape but I am still not happy with the mesh so I am running a dual belt setup but I made belt clips that cap the ends of the rail and affix the belts in place so that the upper belt is tensioned and cannot move.

I've also uploaded a video where I performed the test you describe. I believe I performed it exactly as you described:

 

Well, there you have your backlash!

The backlash will get gradually bigger and bigger over time, due to wearing of the parts of the machine. This is inevitable and part of being a machinist is checking regularly for this backlash and make corrections to have your machine operating as accurate as possible. Eventually the wear will become difficult to correct and that will be time to replace the worn parts.

I don't think (at this stage) that the problem is in your software. Try making corrections for the measured backlash and measure again. Repeat until you're within the desired tolerances.

 

I understand backlash will appear as things wear, but this is a new build, I have only made a handful of cuts in wood only besides a few holes in aluminum with very shallow passes just to see how badly they would come out.

I will be doing a lot of testing as I have to figure this problem out, if not the ox is practically worthless to me.

 

A belt driven system might always have a little bit of backlash from the start on, I guess. After all it is not a rack and pinion or ball screw driven CNC. Why not correct for the backlash either in the software or by adjusting steps/inch? If it solves your problem you can start doing all your testing and move on with it.

 

As most everyone who follows my build knows i'm not really a fan of belt drive, at least in this situation I feel it's terribly undersized.. With that being said there will always be some sort of backlash in belt drive systems, whether you can see it with the naked eye or have to measure to see it. I don't know what software your running but in Mach 3 you can compensate for the backlash. I don't think you can adjust your steps to compensate due to it will throw off your cut measurements ie; .25 hole could become .254 or .246. I agree with Prauk to try and correct it in the software, that's probably the easiest way to do it.

From reading above you say your getting within .002", that not too bad for this type of design. Keep this in mind though even if you tweak everything to death, trying to keep this machine within the tolerances your shooting for will be a full time job.. Doesn't take much to throw these small machines out of wack and have to recalibrate.

Just my .024632 cents....

Gerald

 

Yes I understand belt drive has its weaknesses but I have seen plenty of others, let alone the thousands of shapeokos that achieve great results w/ belt drive.

Also the .002" is what I get on my Y-axis, on my X-axis I am seeing .008-.012" which is unacceptable. I can't figure out why my Y-axis is so good but my X-axis is so poor. The only real difference is the fact that the Y-axis has less belt wrap and a longer section of belt exposed from wheel to wheel around the pulley.

I'm running a tinyg, mach3 is not an option and I am not aware of a way to compensate for belt drive w/ tinyg other than changing the steps per revolution in an attempt to make the best of it.

 

Did you start a build, KerryH? It might help if we know a bit more about your configuration to find out what is going on?

 

I see what Prauk meant now about changing the steps to compensate. Are your referring to the Y as the shorter span on the gantry and the X as the longer rails?

 

I will work on posting my build.

I am referring to the single motor on the gantry left to right as the X-axis. The dual motors on the side plates is my Y-axis.

 

That would be another possible problem, Gray. But it wouldn't come out by using a dial indicator, since those give hardly enough counter pressure to make that part move.