My aluminium CNC Mill/Lathe build.

[S_J_H]

Digits,
Have you considered using poly V belts?
They will handle more rpm and power than a timing belt and run even smoother. They are seamless. Probably the best belt for a spindle that won't see much low rpm use. The pulleys are also very easy to make! They have a 60' profile unlike standard V belts which have 40' included angle. This means you can make them on your mill using a 60' threading tool slowly plunged in if you can chuck the part somehow in your spindle. Just put the cutting tool in your milling vice and presto you have a lathe. I have done this on my x3. Also a 60' drillmill could be used in the spindle and you could cut the groove with the part in a slow constant rotation chucked in your rotary table.
A single 5mm poly belt would be pretty good for your setup. The 5mm poly V is used on the 9x20 lathes. I made a dual 5mm poly drive on my 9x20 in low range. At 20rpm at the spindle I can twist a 3/8" bolt in half before the belts slip.
For my cnc lathe I am building I am using a HTD timing belt from motor to a jackshaft and then twin 3mm poly V's to drive the spindle for smoothest motion.

Steve

[digits]

Hi Steve - Actually, the X-2 belt drive kit that I am modifying for my X-2 head, X-3 motor monster uses V-belts of some description, so I guess I could use the original spindle pulley and then just make or buy a pulley for my motor.

I finally got the nerve to finish wiring up my 72v, 1600VA DIY PSU last night. After a bit of saftey checking with my multimeter for shorts and open circuits, I applied the AC and fired it up. My E-stop relays are powered by a 12v switch-mode PSU that will run my Grex - it's not very happy unloaded, so doesn't produce a true 12v at power up. Once I pressed the 'power-up' switch, there was a short delay, and then a click as all my main relays closed. The lights quite literally dimmed for an instant and then my two 800VA toroids hummed into life. My voltmeter read 77v unloaded, which I guess is pretty close to what I was aiming for.

I then got rather cocky and foolishly decided to connect a servo across the PSU. After powering everything down, and wiring in the servo's power lines, I plugged the PSU back in and hit 'go'. There was an even longer delay than before, and then 'bang'! At first I thought I'd toasted something, but it was actually the motor lifting itself off the top of the PSU and then crashing back down on it, as it went from 0 to 4,000 rpm in the blink of an eye. Fortunately the emmergency stop works rather well

I tested the PSU afterwards - it's fine, but I will have to wait to retest the motor. I'm going to tune a servo with a different motor first, and then I think I'll check that all 7 of my motors work.

The speed the motor accelerated has really brought home just how much I need fail-safe limit switches, so I'm going to try and order some tomorrow for Saturday delivery. I might as well get some more pulleys and belts at the same time

How do people usually deal with keeping their servos and encoders bone dry and chip-free? I can get nice IP65 boxes to seal up all my power terminal blocks, but I can't put the encoder in any water-tight enclosure as it's right next to the motor's ventilation holes.

[digits]

Woohoo! I tuned my first servo tonight - unfortunately my cheapo little 5MHz LCD oscilloscope was no use - I need two channels to be able to trigger off the change in direction and watch the damping response. I had to just tune by ear in the end, but I have at least seen a servo run under Grex/Mach 3 control. I haven't got my motor fully tuned yet, but so far I'm impressed - the acceleration is impressive, and despite being about 5x more massive than my NEMA 23 steppers, they are so, so smooth and quiet, except at standstill

With a bit of luck I might get some limit switches at the weekend, and get most of the electronics wired up. The spindle speed control will have to wait though.

[S_J_H]

Digits, I ran a direct driven belt drive on my x3 for a while. I had it pullied so it gave the spindle around 3500rpm max. Just a little underdriven at the motor. Plenty of cutting power at full power but sort of weak at rpms below 1000rpm.
I over drove the motor also and got around 5500 spindle rpm but the motor had no guts at anything less than 3/4 power. That's with the x3 motor drive. Are you using a different dc drive?
I don't know how you will use your machine but if you ever need lower spindle rpms then maybe a 2 or 3 speed belt drive might be a good option?
Anyhow, good luck on your build. It looks like it will be mighty impressive when finished!
Steve

[digits]

Thanks Steve - right now, I am just hoping it will work at all - but I have to admit, I'm getting excited now that all the major mechanicals for the mill part are done and I have motion on one motor.

As for the pulleys - I've just ordered 20 and 44 tooth HTD pulleys, so I should have the choice of 0-2500 rpm or 0-5500 rpm with a pulley change. That should be enough to get me boot-strapped, and then I can see how well it works. I am not using an X-3 controller board - I have a $50 Ebay one from BICL that is good for 2HP, so I just need to make sure I set the current limit such that I don't burn out my X-3 motor. I also have a spare servo & Gecko (assuming they all work), so I could have a Servo for a spindle motor - I'm planning to do that for my lathe/5th axis head anyway. I don't think that the Grex driver supports that quite yet, but hopefully it will by the time I'm ready!

[digits]

A busy couple of weeks at work have prevented me making as much progress as I would have liked, but I have at last got all my Geckos wired and tested

I had forgotten just how long wiring takes, and just how many minor burns soldering involves

I spent some time today fitting limit-switches to my X-axis, and a couple of hours ago, I got my first axis moving under computer and Grex control Unfortunately the limit switches aren't working and I have already managed a spectacular crash, but fortunately there doesn't seem to be much damage - a couple of the Hiwins have some damage to the plastic, but everything else seems fine.

I am very impressed at just how smoothly everything moves - and how fast - I've had it running at 20,000mm/min ~800IPM, and it's a bit scary to be honest! I think I'll limit it to about 400IPM for now.

[digits]

Well, I got my limit and home switches working last night It was a great relief to see it promptly stop on hitting the limits switches, and even better to watch it gently seek to the home switch, activate it and then roll gently back of it until it deactivated I also ran some G-code on it, but it's not that interesting with only one axis working - the rapids were fun to watch though!

Unfortunately, with the switches in place, I seem to have a lot less travel than I had designed for I think the problem is that the X-carriage was supposed to be able to glide over the motor on the X-axis, but the Hiwins actually collide with the motor mount with about 90mm of screw left to go. I can fix this by moving the X-rails out to the outer slots on the extrusion, which will also improve stability, but it will mean recutting some of the mounting plates, and they'll have to be 290mm long, which is too big for my X-1. I could make them in two goes on the X-1, but I am half tempted to just assemble the machine as is, and then have it cut them. I might also move my home switch a bit closer to my limit switch.

I guess it's high time I uploaded some more pics, but right now, I have barely enough room to move in my workshop - let alone to stand back far enough to take pics! I really need to get my new mill parts into the enclosure and take my X-1 out, but first I need to have yet another go at perfecting my coolant and chip run-off as there's no way I'll be able to move the new machine to clean under it once it is installed.

[digits]

I have been having a bit of a panic as this monster gets closer to assembly - and to be honest, I think I have been dragging my heels a bit, incase it all turns out to be a massive disappointment once it's all together!

I had a panic yesterday about the length of my servo-encoder cables. My machine is over 4ft wide, and the max length of the cables recommended by the encoder manufacturer is 6ft. I got out my tape measure last night, and 6 feet will be just about OK if I put the Gecko-drives on top of the enclosure. I won't be able to reach the front panel then, so I guess I'll have to build a proper console after all.

I also realsised as I've tried to partially assemble it all, that I really need some lifting equipment. I guess an engine hoist is the most obvious choice, but I wouldn't be able to get its front legs under my enclosure - perhaps that's something that needs fixing, or perhaps that's just another excuse for me not trying to assemble the mill

I think I'm going to buy a chain hoist instead - they're not cheap - £55/$110, but I really have far too many cuts and scraches already from trying to lift the assemblies to chest height while fighting their desire to roll over.

I did manage to bolt my X-axis to one side of my Z-axis last night. This produced a rather alarming amount of deflection of the Z-axis column - the rails and slides are solid, but the actual column is moving On further investigation, it seems that the design of my Y-axis assemblies is flawed as it is allowing the extrusion supporting the column to roll off vertical. It is under a lot of torque - I'm guessing the part-assembled X-axis is 30-40kg, so the mount point is under about 400Nm of torque, which if my geometry is correct could be putting 1000N on the column mounting points. In theory this torque will vanish when I get the other column up as the load will be purely vertical, but I'm still a bit worried that as the X-accelerates, it may cause both columns to sway.

My procrastination-free plan is to try and mount the other Z-column tonight, and then see how solid it all is. I still have some metal left over from the build, so I guess I should build some stiffners on the machine once I know where they're needed.

One other major worry is what will happen to the Z-axis during an E-stop. The last thing I want is to crash the head trough the table or to have it come crashing down on any of my body parts. The most obvious solution is a counterbalance, but I'd need 40-50kg, and it'd have to be mounted in such a way that 400IPM rapids aren't going to cause it to clatter against the machine. I am also considering an electromagnetic-brake but I can't see how I'd fit it to the screws - would it just replace my floating bearing?

One thing is certain, this project hasn't lost its desire to gobble up all of my time and money yet!

[digits]

Well, I bought a chain hoist and some straps - should arrive tomorrow. I also bought some M12 bolts so that I can add end-plates on my Y-assemblies if needed.

It was probably not a good use of time, but I did manage to temporarily mount my X-axis on both Z-columns:



The X-axis doesn't look like much without the rails or head assembly, but I could barely lift it as it is. The important thing is that it does just fit in the enclosure, and it seems pretty solid with both columns fitted My poor little X-1 is in there somewhere - once I've drilled out a few motor-couplers and milled down one part I've made slightly out of spec, it can come out and the Z-axes can move to the back of the enclosure.

Here are a couple of shots of all the wiring that's been keeping me so busy lately:


Here's the meaty 1600VA PSU - I've removed the DC-motor controller to its own box temporarily, but it will go back in soon.

And this jungle of wire is my almost complete Gecko box:


I still need to wire up the Y and Z limit switches before I can safely fire up all three axes, but I did get my spindle control relay wired up tonight, and it seems to work under Mach 3 I have plans to control the spindle speed from the Grex too, but I need to build a simple circuit to convert 0-5v to an optically isolated 0-10v for the motor controller board in the PSU box. I don't need that to run the mill or get the basic ATC working though, so it's not going to hold me up

I am not sure if those Geckos have enough heat-sinking - but I have a stack of fans I can add if needs be.

Anyone got any good ideas about spindle brakes or counterbalances for my monster Z axis?

Cheers.

[apache405]

I have seen some commercially made machines use springs to for counter balancing type operations.

[apache405]

btw: are you using XLRs to send power to your motors?

[digits]

btw: are you using XLRs to send power to your motors?

Yes, they're rated at 10A per pin, so I have used 2-pins for +ve and another 2 for -ve. The Geckos are only rated at 20A (and that's not continuous apparently), so it should be fine.

Springs might be an idea - I was thinking about gas-springs, but I don't know where I'm going to get 400mm travel dampers at sensible money.

I was in a DIY store at the weekend - breeze blocks for counterweights look like a great plan, if I can find room for them in the enclosure. I figure I could drill through them and then have them slide up and down some cheap steel tube bolted to the side of the mill.

[apache405]

I was in a DIY store at the weekend - breeze blocks for counterweights look like a great plan, if I can find room for them in the enclosure. I figure I could drill through them and then have them slide up and down some cheap steel tube bolted to the side of the mill.

if you can use some cheaper linear slides instead of cheap tubing i think your design will perform better. I have bad luck mixing low tolerance and high tolerance parts (the last time i did it, it took over 100 man hours to fix and a trip to urgent care to get my right hand fixed after my knife slipped...)

[digits]

if you can use some cheaper linear slides instead of cheap tubing i think your design will perform better. I have bad luck mixing low tolerance and high tolerance parts (the last time i did it, it took over 100 man hours to fix and a trip to urgent care to get my right hand fixed after my knife slipped...)

Ouch - hope you hand is better now! I have another dozen cuts and grazes from the sharp edges on my machine - I really must invest in a chamfering tool!

Thanks for the heads up on the mix & match problems - as it happens, I do have a pair of cheaper linear slides gathering dust - I was going to use them for something else, but they would work well for counterweights.

[handlewanker]

Hi Digits, just been looking back at some of your earliere post on this topic.
When you are turning boring or just machining anything in general, the rule for overhang, that is boring bar stickout or end mill stickout or material in the lathe chuck stickout, is no more than four times the diametre.
This means that if you are using a boring bar of 1"diam then the stick out is not more than four inches, otherwise you will get deflection big time.
The same goes for material in the lathe, a bar of steel 1" diam should not be allowed to stick out more than four inches without centre support.
Think of it this way, if you start with a bar of steel 1" diam and turn it down to 3/4" diam you are now almost 6 times stickout, and will get deflection.
The boring bar in one of your photos is about 3/8" diam going into a 1" diam hole by the look of it, yet it sticks out about 4" long.
I hope this helps as it will improve your accuracy if you can relate to the stickout ratio for all tooling.
Ian.

[apache405]

Ouch - hope you hand is better now! I have another dozen cuts and grazes from the sharp edges on my machine - I really must invest in a chamfering tool!

Thanks for the heads up on the mix & match problems - as it happens, I do have a pair of cheaper linear slides gathering dust - I was going to use them for something else, but they would work well for counterweights.

the scar on my hand is almost gone (oct. 26th is when i cut it).

Hope you cuts/grazes heal up quick.

your machine is looking nice. good luck with fitting up a counter weight system.

[digits]

Hi Digits, just been looking back at some of your earliere post on this topic.
When you are turning boring or just machining anything in general, the rule for overhang, that is boring bar stickout or end mill stickout or material in the lathe chuck stickout, is no more than four times the diametre.
This means that if you are using a boring bar of 1"diam then the stick out is not more than four inches, otherwise you will get deflection big time.
The same goes for material in the lathe, a bar of steel 1" diam should not be allowed to stick out more than four inches without centre support.
Think of it this way, if you start with a bar of steel 1" diam and turn it down to 3/4" diam you are now almost 6 times stickout, and will get deflection.
The boring bar in one of your photos is about 3/8" diam going into a 1" diam hole by the look of it, yet it sticks out about 4" long.
I hope this helps as it will improve your accuracy if you can relate to the stickout ratio for all tooling.
Ian.

Thanks Ian - as you can see, I'm just a newbie so I'm learning most things the hard way! I do appreciate the advice though - that's what's so great about this site

I have learnt my lesson about having too much of my end mill projecting from the collet, and I did use the shortest, stiffest boring bar I had in my set, but given the size of my machines, I do find that I often have no choice but to break these rules of thumb. For example that boring bar was indeed about 10mm in diameter, and about 50mm/2" long but then again, the hole I was boring was 40mm deep. I also have some very deep, narrow cuts (1/8" wide, 1" deep) that I need to make in some other parts - how do I make them if not with a 1/8" mill that's > 1.5" long? I was planning to just use tiny chip-loads to reduce the cutting force, and hence the tool deflection...

Cheers.

[digits]

the scar on my hand is almost gone (oct. 26th is when i cut it).

Hope you cuts/grazes heal up quick.

your machine is looking nice. good luck with filling up a counter weight system.

Thanks, I keep having to take my gloves on and off when switching between the machcines and lifting bits of the new one. I'd much rather get a nick from a stationary part than be pulled into my bandsaw, so I tend to wear my gloves too infrequently when lifting heavy parts of the new machine. My chain hoist helps, but it is very noisy and slow.

I actually think my machine is beginning to look rather scary! I have yet more wiring to do this week, so more burnt thumbs to come, and I'm hoping to try and assemble it on Staurday.

I really need to sort out my coolant and chip flow first though - I have bales of curly chips under my X-1. As soon as I can hoist it out of the way, I'm going to have to think of a way of re-doing it. My current plan is to use thin sheet metal to line the table, and a very large drain hole (6"x4") and then just to let all the chips flow off the table, and into a large sieve. The sheet metal is very expensive in DIY shops, but I don't think I have time to order off eBay and then wait for it to arrive. I also want to run coolant lines right round the edges of the table so that I can actively wash chips down into the drain - I have no intention of reaching under 250kg of machine!

[digits]

Well, I've not posted for a while - been too busy wiring up servos and limit and home switches. I can't believe just how long it takes, and I used to consider myself quite competant with a soldering iron!

My 'workshop' is now a proper deathtrap - I have mill assemblies everywhere, and no room to swing a cat:


I still need to put those power terminal blocks into little IP65 boxes to keep them dry, but I wanted to get all the screws and motors tested first.

On Saturday, I hoisted my X-1 off the table for hopefully the last time - I had quite mixed emotions - while it has allowed me to make all the parts for this new machine, the number of gearbox and Z-axis problems I've had has made me hate it!


I've also replaced my coolant tray with a nicer aluminium one - OK, it's not all that neat, but it's watertight, and should work well.


I've just put the mill-base on the table to see how it all fits:

I need to re-line the walls of the enclosure and then I can begin assembly

Cheers.

[digits]

Wow, that was a lot of hard work - so I'm just going to post one pic for now:



I've not tried powering anything up yet - I damaged several of my limit-switches during lifting, but the interesting thing is that it doesn't seem to need a Z-counterbalance to stay still without power.