Separate names with a comma.
Some features disabled for guests. Register Today.
Discussion in 'CNC Mills/Routers' started by Steve hope, Feb 13, 2015.
Discussion in 'CNC Mills/Routers' started by Steve hope, Feb 13, 2015.
Routy cnc upgrade.
Steve hope published a new build:
Read more about this build...
your build looks like its rocking along. Congratulations!
With this design I would definitely look at adding the servo belt idea to help with belt stretch over those larger spans. A bit tricky to install for the first time but really feels like it makes a big difference on my smaller OX.
I am happy to help with Vcarve and Mach3 issues as I know it can be a bit tricky when you're starting out! I am NZ based so can't come to your workshop but happy to help where I can online.
What issues are you having at the moment with mach3 and v carve?
Thanks for the quick reply.
So initially, I was having dimension issues, even though I set up mach for mm it was still running in inches. I appear through persistence and **** right stubbornness to have got it working, I set the steps automatically through the settings screen and seem to have ironed out that little issue.
I know the timing belt is not ideal and would love to upgrade to rack and pinion, although I have been told that belt drive should work ok for now, I intend to put the nema 23s on to this at some point in the near future, yet another boat load of wiring issues to get over, bipolar, unipolar..... Duh, I am a little lost.
I have a feeling I have the homing and direction switches set wrong, although it travels the right direction when I load g code it appears to cut from the top left ( in plan) which in my head should be bottom left in plan!!
I know x is my shorter axis and should travel in a + direction, so probably crack that issue this weekend
ah mach3 can be a bit funny when your setting it up but stick with it as you will end up with a sweet rig!
The first issue where its swapping into inches sounds like your calling it from either the codes run after each reset or maybe from your Gcode. Have a look in your code and see if you can spot any G20 (Inch) as you want to call G21 for Millimeter unit mode
I think the timing belt should be ok for now especially if you add that second belt as in the vid I posted as this essentially acts like a rack and pinion system. Much cheaper and achievable with what you already have
Upgrading to nema 23 all round should also be done if you want to get the most from the router as I found even with a Nema 17 on the Z it struggled a little.
Sounds like you have some axis around the wrong way. Would you mind letting me jump on your setup via team viewer as it would be much faster for me to find your issues. I should be free tomorrow from around 3pm-10pm Auckland local time if that suits. This may help you work out what time that is for you http://www.timeanddate.com/worldclock/new-zealand/auckland
No can do, I don't have Internet in the garage.
I will reconfigure my homing and limit switches to another axis on the table.
Can you recommend anyone for a set of ox plates??
I intend to put a second x parrallel to the original and double up on the Z plates and put 4 or 6 wheels on that too as well as doubling up the the timing belt so I have one top and bottom of my y, what do you think, this should eliminate any wear on one and saves using the servo belt?
Wow, what a pain in the arse my machine is being.
Almost every time I tuen it back on I seem to have another minor issue.
So bought a new limit switch for the other side of my Y and wired it into the 2 existing Y limit switches either end of the table so I can try to find perfect zero for both sides so it cuts square.
Now my limit switches are not working, oh the joys.
I will persist tomorrow when I have a few hours.
It has been a long but very cool build, electronics and stepper motors are definately not my strong point.
At the risk of sounding like I am ranting, if I tell you what I have my settings at could you tell me if you think it sounds right?
X and Y dip switches are set to 1 1/6 which is 16x200 (3200).
Timing belt is gt3 2mm pitch, the diameter of the cog appears to be 15mm.
Using this I got 106.66 steps per revolution.
Although when I use the auto set on the settings page it recalculates to 89.4 steps.
This does allow for pretty precise moves in mm.
The Z axis calculations I am not so sure about, I did have it set to 1 1/6 but really struggled to get it to move without making huge noise and having to turn up the velocity to get it to move, I changed this to full step on the dip switches and have now managed to get it sounding quieter but not as quiet as I would like.
I don't understand the whole concept of steps per and how that relates to the movement, in my head if the dip switches are set right then auto tune should be as close to precise as I can find, 1 1/6 I was told by my supplier of the electronics (system 45)) was going to give me the required level of precision.
What denotes acceleration and velocity, a mach3 tutorial I watched said listen for the sound, but this makes the numerical values ridiculous.
Anyway, thanks for you help.
It's nost appreciated.
I am running late for a game today so cant help too much right now BUT I would highly recommend getting a cheap usb wireless plug in receiver for your computer out in the garage. To be honest I will find it much faster to find and solve faults if I can see you whole rig rather than expecting you ( a new Mach user) to be able to understand where all the tricky bits are. Maybe you can move just the computer inside for a bit so I can get a feel for your setup and correct any obvious issues?
Plates there a few good sellers but I highly recommend Robert Hummel (check out his AI ox build) as they are very well made, have lots of extra features that make an OX much nicer and he packs them very well for shipping.
Make sure you factor solid extreme wheels for the Z if not the whole build as they do also provide a lot more stiffness over the standard wheels.
your steps per can be a bit tricky to get your head around and to be honest I say don't bother right now. Just do each axis using the auto setting method on the settings page. This will give you a very accurate method especially if you base that number over a good size. I actually do the test over almost the entire movement for that axis to give me a good average for the whole travel.
Your Z is going to be quite different because the dive system has a very different (much higher in my case steps per) value than the belt drives.
the servo belt idea is not a special product just a name that is pulled from the original inventor of the idea. Our system uses standard openbuild GT3 belt and you only need the total length of each axis extra and a good roll of double sided tape to really help you with belt stretch over such long axis
Ok... i got to hit the road or I wont make it in time to gear up for the first match but ill add some more ranting once I get back home.
Hope life is good with you, I have taken the plunge and am now in the throws of ordering Ox plates and extra vslot for my X axis, managed to iron out the movement issues with mach 3, would you believe, operator error was the cause of the problems!!! Lol...
I am trying to find the respective code for auto tool zero which I don't fully understand but feel I need, as at the moment, I am using a lot of guess work for my Z axis and know I need the precision on that more than anything else.
Any ideas where I can get the right script from for this?
I have managed to make a probe using a crocodile clip and some PCB. It seems to work ok.
thank you for your advice and guidance.
Can you give me some more info with regard this crocodile clip probe?
If you search auto tool zero probe you will find more information than I can give you, however,
I started with Two pieces of wire, on one end I soldered a small piece of PCB (maplins) and attached the other end to my System45 (electronics).
The second wire I soldered a crocodile clip (maplins) to one end and return the other to my electronics.
The wires are connected exactly the same as your limit switches.
The principal being that you attach the crocodile clip to the shaft of your kress/router and when you lower the router the bit touches the printed circuit board and connects the signal.
You can test this by selecting ports and pins, inputs/auto select probe.
Attach the crocodile clip to the router, Just touch the PCB off the edge of the router and this will give you the required ports and pins
You need to add script to the Auto tool button to get this function to work correctly, again Google is key or ask a more seasoned veteran on open builds who may explain it better than me.
I see what you were referring to now, "zero" probe.
I thought you meant the sort of probe for measuring, recording, and creating 3d forms.
They are really expensive!
This is a great design! I'm interested in doing something similar only want to make it to take a 1/2 sheet (4x4) Were you successful in doing accurate cuts with this machine with a belt drive? What about bowing on the x-axis... were you able to solve that with thicker struts? what size did you use?
Firstly, thank you.
Secondly, here I go....
Because I used the parts I bought for the basic routy kit, this has led me to a few problems I am now in the process of rectifying......
1 new set of Ox plates waiting to be purchased to allow me the clearance required for the length of X axis.
1 new length of 80x20 for the X axis, ordered and waiting delivery.
The belt works ok though, I have managed to do a few pockets so far and the machine runs to the right dims, although and this is purely other people's opinion.......
Over time I am going to get some belt stretch, how much, I can't say right now as I have not gone into full on production mode but, I would recommend using linear rails in hindsight, if I had had the cash at the start and knew more about this I would have put on, it was only inexperience (and cost) that I used the timing belt.
Right at this moment the router struggles (deflection on the X axis) which can maybe be countered by cutting at shallower passes to keep torque down to a minimum, (obviously taking longer per cut).
As I said above, after researching other builds, I see that it would be prudent to add an extra section of vslot to the X to avoid this and have ordered that, also the routy kit has a section of 6020 which is upright, bolted to the Y axis gantry plate supporting the ends of the X axis, this is not rigid enough either, maybe works on a smaller machine, but certainly not the monster I have created.
I am now in the process of ordering Ox plates for the Y axis.....
1 to give more strength to the X and because the stepper motors connect directly to the Ox plates giving me extra clearance on the length of the X axis.
2 allowing me to ditch the s&it bit of 6020 that currently supports the X.
3 better cable management with the Ox plates.
All currently on order and waiting for delivery, will most definately post on the page as my build proceeds.
This place is a minefield of information and with the right amount of Google and general web diggin I have managed to get everything running the right direction,(speeds and velocity need fine tuning, but not going to play with this now until I have remedied the X and Y issues.
It has been very cool, very frustrating at times, but I feel like Frankenstein bringing his monster to life ( cue maniacal laughter) .
There are 4x4 machines on the market, but if you have a little bit of skill and can cut,use a tape and have a bit of a head about you, this can be achieved with parts available on the open builds store or generic suppliers, (vslot, tee nuts, solid wheels, etc) you can build one a lot cheaper than list price on many sites.
I hope this helps.
Sorry about the delay in reply, been a bit side tracked.
Did you get a zero code working?
Awesome, thanks for the info! I'm starting to think that a linear rail driven machine will be a better option... right now I'm stuck considering the following:
1 - just build the standard (large version) OX design to get a working machine in order to conceptualize, get some experience and ultimately use it to build a larger model
2 - Design and build a 4'x4' linear rail driven machine and take twice the time but end up with what I want on my first build...
I'm starting to consider the first option as this would get me something to get started with right away...
although it will be the more expensive option....
I made my own auto zero plate out of some aluminium copied and pasted some script for the button and it works a treat.
Happy day, now just waiting on Ox plates and another bit of gantry for the X axis, too much torque/deflection at the minute.
I have managed to cut some pockets fairly successfully but anything more than a mm or two and the X starts to bend on the Y axis uprights which were off the routy.
Hence the reason for the other parts.
But slowly does it, it's been six months now, a few more won't hurt me.
I would say that if you create your own cnc machine you not only have the knowledge and skills to understand it but it also is to your design and size.
A company called Robocutters have an 8x 4 machine on their site, it looks cool I think with the optional extras it was over £3k inc all electrics etc.
On average the prices I could find for a new 8x4 machine was about £4k, obviously yours is not going to be as big so just cost half in theory
I think mine has cost me about £1500 -£2000 and I understand fully how it works where I can improve it and what I can do to make my life easier.
Does the ox cut the 4x4 to the edges?
It will be interesting to see your path.
Now to learn and understand Mach3 and VCarve pro. Grrrrrrrrr
Using full length 1500mm (trimmed to be of equal lengths as factory can be off by upwards of 4 mm - from personnal experience with V-slot) it can do 4' x 4' cuts to the edge ... and a bit beyond by a few inches for the X axis. myOX has a X cutting width of about 52". On the Y, it could be a bit tight for 4' cut. I'm getting shy of 20" with 750mm length, adding 750mm of pure travel would give it around 49.5" of cutting.
From there, it's just working out if GT3 is going to stretch or if linear slides would be best from start and all of the other fun stuff of a DIY project. Tons of info and builds here at OpenBuilds to learn from ...
But please, DO NOT BUY AN 8x4 CNC FROM ROBOCUTTERS.
I HAVE BEEN LEFT £2,300 out of pocket. Missing bits, to short a gantryies, holes in wrong places, missing bits still, no electric, very rude manner. All out con,
He has now been struck off. So watch out for his new compnay
That's a shame. I had minor issues with them but they always came through. Missing items were dispatched. I had to drill a couple of pieces for my machine but then I bought individual items from them and not the whole machine.
Please bear in mind that Robocutters buy form external sources and any manufacturing defects would generally have been machined from a third party. Just a thought before you slate them online!!
Unfortunately it is what I said.
Not my place to question you but have you spoke to Colin?