PDA

View Full Version : BUILD LOG: 3 Axis CNC router



Pages : 1 2 [3]

Neale
06-11-2017, 12:33 AM
No, don't think it would do any harm.

m_c
06-11-2017, 01:42 AM
Plugging the terminator in shouldn't do any harm, either will leaving it off, as without any expansion modules, the CAN bus won't be doing much.

The purpose of it is to stop signals from reflecting/bouncing of the end of the CAN bus wiring, and causing interference.

JoeHarris
08-11-2017, 02:22 PM
Thanks to the Legend that is Dean I got a stepper to move last night - woohoo!

https://youtu.be/59WaetUDAtU

Nr1madman
08-11-2017, 02:24 PM
Awsome!! :D

Skickat frŚn min SM-N910C via Tapatalk

Nickhofen
08-11-2017, 04:41 PM
Very nice!
In a few days you will make dust....men glitter!!!:loyal:

routercnc
08-11-2017, 08:09 PM
Anything on the machine which moves by itself counts as 'first movement' so woo-hoo indeed ! Nice one Joe and nice one Dean where ever you are ! :encouragement:

grain_r
08-11-2017, 09:11 PM
Well Done Joe.

Your thread is a constant spur on for me. I'm in the same position of young children and no time so I know how well you are doing to move the build on.
You will have to start designing those things that you dreamed of making when it was a If only I had a CNC machine I could make ??????.

JoeHarris
15-11-2017, 11:00 PM
Actual, real life, jogging movement on Y, ladies and gentlemen:
https://youtu.be/GfsmxBR0q3c


Sent from my iPhone using Tapatalk

routercnc
15-11-2017, 11:22 PM
Actual, real life, jogging movement on Y, ladies and gentlemen:
https://youtu.be/GfsmxBR0q3c


Sent from my iPhone using Tapatalk

Nice one Joe getting closer

Sounds nice and smooth too so general alignment and straightness of ballscrew is pretty good

Clive S
16-11-2017, 12:21 AM
When's the party Joe :applause::beer:

Nickhofen
18-11-2017, 08:04 PM
Awesome and as routercnc said it sounds to work smooth,(I write that to show that I know something from cnc machines), lol,you are almost there man,carry on!!!

JoeHarris
12-12-2017, 11:58 PM
Getting close now... two axis moving, spindle on and spinning.
https://uploads.tapatalk-cdn.com/20171212/a725ce51f58dfad4a16cf52de534db7f.jpghttps://uploads.tapatalk-cdn.com/20171212/ee478329df74b6385213bf7cc851becc.jpg

JoeHarris
30-12-2017, 01:11 AM
Hello all,

Bit of an explanation of my limit switch wiring and video of all axis now moving below:


https://youtu.be/YhuqM_D9lS8

https://www.youtube.com/user/joerharris

reefy86
04-01-2018, 11:11 PM
Must be proud of your self :) still watching your video's so well done mate.

JoeHarris
23-01-2018, 11:26 PM
I have been busily getting my head around fusion 360 in the background (very nice programme, can't believe they let the likes of me use it for nothing!) and sent my first programme to the machine this evening and of course things did not go quite to plan.

The intention was to cut a small pocket 55x55mm with a 1/2" router bit (12.7mm)

This was all set up correctly in fusion (I think) and it cut fine, albeit far too slow (I'll up the feed rate next time around), but the problem came when I measured the part - see pictures below... any ideas!?

After all this time building, it is dawning on me that I have no idea how to use this new tool of mine :(

Edit: Sorry - pictures make this look worse than it is as I forgot to zero the calipers! I'll take some new pictures...

JoeHarris
23-01-2018, 11:37 PM
New pics after previous post - prob need to run the calibration wizard in the settings window of MACH3 again... but being out by 0.8mm in one direction and 1.4mm in the other does not seem a great starting point.

https://uploads.tapatalk-cdn.com/20180123/2dc1dff21201779353329b5d826cdea8.jpghttps://uploads.tapatalk-cdn.com/20180123/eb6a8e9bcb80571571e2a28957793e72.jpg


Sent from my iPhone using Tapatalk

Clive S
23-01-2018, 11:52 PM
Joe don't get despondent, this is normal when first getting the machine calibrated. Check it over a greater distance say 300mm in fresh air ie. in the midi type g1 f500 x300 and check to see if it has move exactly 300mm then do the same on Y ie type Y300 and check that. you might have to amend the steps per.

JoeHarris
24-01-2018, 12:23 AM
Joe don't get despondent, this is normal when first getting the machine calibrated. Check it over a greater distance say 300mm in fresh air ie. in the midi type g1 f500 x300 and check to see if it has move exactly 300mm then do the same on Y ie type Y300 and check that. you might have to amend the steps per.

Thanks Clive, it was nice to see it cutting at last and Iím sure I will get it all sorted but I have quite a few things to fix and finish and Iím getting very impatient!! I will have another go tomorrow...

njhussey
24-01-2018, 12:52 AM
Hi Joe, good to see your machine is moving, albeit not as accurately as it should! I've been using my machine for a while...as you can see from the state of the bed (sacrificial) I still make mistakes. Every day is a learning day [emoji16] [emoji106]

Hope you get it sorted soon [emoji106]https://uploads.tapatalk-cdn.com/20180123/586b726a425d5802e0f8f788acf49b0b.jpg

Sent from my HUAWEI VNS-L31 using Tapatalk

Nickhofen
24-01-2018, 05:19 AM
No big deal.
Like Clive said you run the calibration and Mach 3 asks if the axis you calibrate,each time,moved at the distance you want,you put the correct distance in the proper box and you see the machine do the correction!
Nice to make the first cut , especially with a machine you build, right?

Neale
24-01-2018, 09:51 AM
I shudder every time I see someone suggesting using the calibration routine in Mach3! If the machine is giving errors this big, then you need to find out why, rather than hide it under the carpet. You are using ballscrews which are almost as accurate as your digital caliper. You know how you have geared (if not 1-1) your motor to the ballscrew. You know what microstep setting you have used. This gives an exact "steps per" value, and if the machine isn't achieving this, you really need to find out why. For example, it might be as simple as needing to check the adjustment of the ballscrew support bearings - and whether they were assembled correctly in the first place (don't trust the Chinese to get this right!).

Double-check everything first, including actual cutter diameter (I spent ages chasing machine errors that were down to cutter diameter errors, although that's not the case with differing errors in X and Y), and don't panic. You'll get to the bottom of it with a bit of systematic testing!

Good luck - we've all been there.

JoeHarris
24-01-2018, 10:14 AM
Thanks everyone! Neale, how tight should the ballscrew tensioning nuts be tightened? Iíve seen that people shim these but not sure how it is done tbh


Sent from my iPhone using Tapatalk

JoeHarris
24-01-2018, 10:16 AM
I shudder every time I see someone suggesting using the calibration routine in Mach3! If the machine is giving errors this big, then you need to find out why, rather than hide it under the carpet. You are using ballscrews which are almost as accurate as your digital caliper. You know how you have geared (if not 1-1) your motor to the ballscrew. You know what microstep setting you have used. This gives an exact "steps per" value, and if the machine isn't achieving this, you really need to find out why. For example, it might be as simple as needing to check the adjustment of the ballscrew support bearings - and whether they were assembled correctly in the first place (don't trust the Chinese to get this right!).

Double-check everything first, including actual cutter diameter (I spent ages chasing machine errors that were down to cutter diameter errors, although that's not the case with differing errors in X and Y), and don't panic. You'll get to the bottom of it with a bit of systematic testing!

Good luck - we've all been there.

Thanks everyone! Neale, how tight should the ballscrew tensioning nuts be tightened? Iíve seen that people shim these but not sure how it is done tbh

JoeHarris
24-01-2018, 10:22 AM
I shudder every time I see someone suggesting using the calibration routine in Mach3! If the machine is giving errors this big, then you need to find out why, rather than hide it under the carpet. You are using ballscrews which are almost as accurate as your digital caliper. You know how you have geared (if not 1-1) your motor to the ballscrew. You know what microstep setting you have used. This gives an exact "steps per" value, and if the machine isn't achieving this, you really need to find out why. For example, it might be as simple as needing to check the adjustment of the ballscrew support bearings - and whether they were assembled correctly in the first place (don't trust the Chinese to get this right!).

Double-check everything first, including actual cutter diameter (I spent ages chasing machine errors that were down to cutter diameter errors, although that's not the case with differing errors in X and Y), and don't panic. You'll get to the bottom of it with a bit of systematic testing!

Good luck - we've all been there.

Thanks everyone! Neale, how tight should the ballscrew tensioning nuts be tightened? Iíve seen that people shim these but not sure how it is done tbh

JAZZCNC
24-01-2018, 03:37 PM
You know how you have geared (if not 1-1) your motor to the ballscrew. You know what microstep setting you have used. This gives an exact "steps per" value, and if the machine isn't achieving this, you really need to find out why.

Actually neal it's not that simple and I pretty much guarantee you that if your using timing belts then the steps per won't work out per the calcs.
This is one reason why mach's calibration is best used for most people who are using belts and only have basic measuring tools.
Even then it still won't be spot on and can be improved upon. But to do this you need Linear scales and to measure over large distance.

When I build a machine I initially set it up using the calcs then I fine tune using linear scales and EVERY TIME the steps per is different to the calcs. Close but it's always bit lower or bit higher.

JoeHarris
28-01-2018, 11:04 PM
So I tried the calibration tool in MACH3 this evening. Moved a centre finder 600mm up the x axis against a steel rule. To the eye it was bang on. Moved it 500mm on y axis and it was maybe out by 0.5mm over that distance. So why was it out over 50mm when cutting??


Sent from my iPhone using Tapatalk

Clive S
29-01-2018, 12:08 AM
So I tried the calibration tool in MACH3 this evening. Moved a centre finder 600mm up the x axis against a steel rule. To the eye it was bang on. Moved it 500mm on y axis and it was maybe out by 0.5mm over that distance. So why was it out over 50mm when cutting??


Sent from my iPhone using Tapatalk

Is it missing steps? What are you setting for that axis?

JAZZCNC
29-01-2018, 10:30 AM
Is it missing steps? What are you setting for that axis?

Clives right it's missing steps but not because of anything you have done or the machine.!!

It's most likely the Step pulse is on then wrong side of the slope. It happens often with IP-M and Leadshine drives or any drive that can select the leading edge.

Easy to see and easy to fix. The test is to knock up some G-code which moves say 500 times from point A to B doesn't matter how far.
So G0 x10 then back to X3 repeated 500 times will be plenty. Then last move back to X0.
What I do is setup dial gauge and zero out against the axis. Then run code, this why only move to X3 so not smacking gauge each move then last move back to it with X0.

If out it will soon become clear.

If doing this the fix is to go into MOTOR OUTPUTS and toggle the Step Lo box. This is the only box other than Enable that IP-M uses on that screen.

Hope this fixes you.

JoeHarris
29-01-2018, 10:54 AM
Clive, Jazz, thanks that makes sense - I remember reading but not understanding this in the ipm manual! Iíll give it a go and see. Could not understand how it could be loosing steps on such a light cut.... cheers guys


Sent from my iPhone using Tapatalk

JAZZCNC
29-01-2018, 11:09 AM
Clive, Jazz, thanks that makes sense - I remember reading but not understanding this in the ipm manual! I’ll give it a go and see. Could not understand how it could be loosing steps on such a light cut.... cheers guys


Sent from my iPhone using Tapatalk

99% sure it will be that Joe. (Provided nothings lose or falling off machine which knowng you sure it won't be..lol)
Happens 50% of the time with these controller/drive combo and it's something I check for has matter routine. Can happen on any controller really but with the IP-M or S happens randomly depending on how drives setup. Esp if not used tuning software to check Rising edge setting in drive.

Edit: For anyone else reading this that doesn't have Cslabs controller but as similair problem then the same can apply.
The test and fix is the same. Or if the the drives have option to change the leading edge setting use this. Both work same.

Neale
29-01-2018, 12:09 PM
I saw exactly the same problem on my machine. X and Y were fine, but Z looked as if it were losing steps on complex 3D work. In my case, it didn't show up on single test runs; what seemed to happen was that you lost one step every time you changed direction. Fix was simple - as Jazz said, just change step polarity. I think this is more likely to happen with the CSMIO controllers as they have differential outputs - brilliant for noise rejection, but too easy to get the two wires swapped!

My only reservation about this diagnosis is that I wouldn't expect to see any significant error on a single movement, as the effect is cumulative with movement direction changes. But Jazz's test (I did exactly the same thing when I was chasing the error) will soon show the truth!

JAZZCNC
29-01-2018, 12:56 PM
what seemed to happen was that you lost one step every time you changed direction.

That's exactly what happens because the last pulse gets sent after the next direction change happens so it gets dropped.

To be honest, can happen on any system. The reason it tends to happen on these is that of the flexibility both drives and controller give. Meaning it's easy for one or other to be set wrong. Majority of older or lesser drives only allow rising edge which is ok for most sytems until you get controller that use's the falling edge.!!

JoeHarris
29-01-2018, 02:28 PM
Thanks Dean, I'll have a look and see. I don't think anything is loose or falling off!! Only one I'm not sure on is how tight the square shaped nuts on the BK bearing block should be and whether they are perfectly set up. I've done the best I can, but when jogging the machine rapidly in alternating directions on the X axis it does sometime make a noise and appears to momentarily stall. I'm not sure what that's about and it doesn't do it consistently but wondered if that could be connected to differing tension on the two ball screws? I did not get this when cutting or doing the calibration tests at the weekend. I'll try and video it so you can see what I mean...

JoeHarris
29-01-2018, 02:32 PM
Thanks Neale, on a single movement there is very little error, only when cutting a test pocket with changes in direction did i see the error so I'm hoping this is the explanation. Whilst I've taken care throughout the build and asked a lot of questions, I have made this machine with basic tools and basic skills, so I'm sure it's not just down to calibration!

Desertboy
29-01-2018, 06:46 PM
I read sometimes the seals on the ballscrew can be tight when new so when you first run the machines they can stick a little, the solution run it up and down the full length of the travel a few times. Greasing can help to.

I had a sticky bit right at the end of my travel on one of my Y ballscrews kept firing a stall detect at me, after 10 to 15 running up and down the lengths occasionally stalling the problem disapeared. The area where it stalled is now part of my homing sequence so it gets travelled over regulary never stalled again.

JAZZCNC
29-01-2018, 09:47 PM
I had a sticky bit right at the end of my travel on one of my Y ballscrews kept firing a stall detect at me, after 10 to 15 running up and down the lengths occasionally stalling the problem disapeared. The area where it stalled is now part of my homing sequence so it gets travelled over regulary never stalled again.

That's most likely to be alignment issue not sticky seals. To be honest I've fitted 100's ballscrews and never had sticky seal on the ballnut(presume you mean ballnut seal)

Now what can happen is often the spacers on the BK bearings are not long enough and the seal in BK bearing can rub on the screw.
This tends to happen mostly on cheap BK bearings. Also on cheap BK bearings, like what Chai Sells, the bearings are not always shimmed correctly or even in wrong way round so when tightening the nut they will bind.

Joe If BK's came from Fred at BST then you can tighten them fully and they won't bind. Fred BK bearings are light years better than Chai's.

JoeHarris
30-01-2018, 01:21 AM
I think I have found the culprit, there was (much to my annoyance) a loose grub screw on one of the y pulleys which I noticed whilst running Jazzís pulse test. With this tightened down and thread locked all seems to be well... The test results can be seen here:
https://youtu.be/hsT3GkAq9T8

JoeHarris
30-01-2018, 01:27 AM
That's most likely to be alignment issue not sticky seals. To be honest I've fitted 100's ballscrews and never had sticky seal on the ballnut(presume you mean ballnut seal)

Now what can happen is often the spacers on the BK bearings are not long enough and the seal in BK bearing can rub on the screw.
This tends to happen mostly on cheap BK bearings. Also on cheap BK bearings, like what Chai Sells, the bearings are not always shimmed correctly or even in wrong way round so when tightening the nut they will bind.

Joe If BK's came from Fred at BST then you can tighten them fully and they won't bind. Fred BK bearings are light years better than Chai's.

My Hiwins came from Fred but ballscrews and bk/bf bearings from chai.... not wild about taking them apart especially now!

Desertboy
30-01-2018, 09:41 AM
I hadn't squared the gantry up at this point so very possible it was alignment, if would be about the same time. It has taken me ages to get dual homing working.

Hmm interesting with the bearings will have a look, I have a spare BK12/BF12 fitting to strip down. But will make sure I buy my bearings from Fred next time I keep seeing his name mentioned.

Clive S
30-01-2018, 10:15 AM
It has taken me ages to get dual homing working. If you have this now working with 2.8 master would you mind posting the hal & ini files.

Desertboy
30-01-2018, 10:32 AM
If you have this now working with 2.8 master would you mind posting the hal & ini files.

Will do, not at work today but will go in to play in a bit I'll post them up then.

I almost went loony toons setting it up lol.

Desertboy
23-02-2018, 03:32 PM
99% sure it will be that Joe. (Provided nothings lose or falling off machine which knowng you sure it won't be..lol)
Happens 50% of the time with these controller/drive combo and it's something I check for has matter routine. Can happen on any controller really but with the IP-M or S happens randomly depending on how drives setup. Esp if not used tuning software to check Rising edge setting in drive.

Edit: For anyone else reading this that doesn't have Cslabs controller but as similair problem then the same can apply.
The test and fix is the same. Or if the the drives have option to change the leading edge setting use this. Both work same.

I had exactly the same problem and I have leadshine drivers this fixed it.

Thanks saved lots of head scratching.

Joe did it fix yours?