Scratch Build – Complete Project QUAD - a Quick and Dirty PC/Spindle cooling with CNC Router attachment

So here is a first test fit...




The servo mount is not attached yet. It sits on the back side below the top plate...

 

I want my motor mounts to look like they are radioactive too.

 

freakin' awesome , it's an art piece by itself somehow

 

I finished the last part on the mill this evening...now only the brake adapter is left.

It is a "cable management" bracket.




Made recesses for the grommets on both sides, plus two M15x1 threads for the quick disconnects




I am using Koolance quick disconnects for the water cooling tubes.




So I have an organized fix in front of the drag chain for:
- Compressed air tool change
- Compressed air taper cleaning
- Compressed air bearing seal
- Spindle power line
- Spindle sensor line
- Compressed air mist cooling
- Two spindle cooling lines
- Mist coolant line

All these lines can be disconnected either at the spindle directly or at the quick disconnects.
So I can remove the spindle if necessary without spilling the coolant or having to remove any lines from the drag chain...

 

Did someone say porn? Well...here you go.
But let's leave softcore behind and go hardcore

I still need the right pulleys and belt to come in to do the brake adapter, but I could't resist another test fit, now that all milling parts are done...

 

Yesterday the pulleys came in, so I could finally finish the Z-axis.

After boring and reaming the inner hole, I added 3 threads for M4 grub screws to each pulley.
Strangely enough, the ball screw does not have a flat spot at the end stub, where the pulley sits (or a coupling would.)
What is even stranger is that I totally forgot to add one by myself
Now it is too late - no way I disassemble the whole thing again - so a press fit plus 3 grub screws will have to do for torque transmitting...




The BS pulley received also three M3 threads on the top for the brake adapter.

 

I turned the brake adapter out of some Ø35mm stock and added 3 countersunk holes.
This time I actually used the 4th axis not only for work holding, but was well for 120 deg rotation...I know, still not what Cheapskate had in mind




The BS shaft end is only 15mm in length, whereas the pulley is 26.5mm high. So this peg can slide into the pulley and center the adapter.

 

This nut slides into a mating disk at the bottom of the brake.




The brake is essentially an electro magnet.
When 24V is applied it pulls up a coil spring and the disk can rotate freely.
When the power is cut off (missing Enable signal, or at machine shut-off) the spring extends again and pushes a plate against the disk.
Then the BS is blocked and the whole sled with the spindle can not crush down into my table...

 

It seems like overkill, but it's still awesome.
-You are going to tease the F out of me with that 4th axis, aren't you?

 

I started with the pneumatics today.
First off was a little box for signal distribution to the valves.




It will also house the pump for the mist cooling.

 

The mist cooling uses a peristalsis pump on a stepper motor for coolant adjustment.




A peristalsis pump have rotating pins which knead a very soft tube and transport fluid this way.




With this system you can adjust extremely precise the amount of coolant fluid going to the nozzle.




The pump is just for flow control. The actual transport of the fluid is done quite conventional by air pressure.

The tank has two valves - three in fact.

One which pressurized the tank with 2 bar (29 PSI). This one is open, when the Enable signal is there.
So when the machine is off, there is no pressure on the tank.

One at the T-fitting at the intake. This is connected to the air line going to the nozzle.
So you can switch on air cooling.

One at the exit. This one releases the cooling fluid to the pump, which is being pushed there by the pressure in the tank...and further through the pump to the nozzle.
So you can swich on mist cooling.

 

The whole thing is controlled by a small Arduino board, which is hooked up to the CNC control.

Air and mist can be either switched on by CNC software or by external buttons.
Coolant control and an additional spit mode are controlled by external potis.

 

Not sure yet wether I use this nozzle or the cheap and chearful one from Amazon.
I actually like the flex hose on the cheap one.

 

I've dragged the air nozzle across a few parts in the past. Flex is the way to go for the filthy casual like me.

 

You name it.

The maker of this thing recommends to attach it with this type of magnetic measuring tripod ...



But first of all my axis is not magnetic and secondly I just hate to fiddle around with these arms until they finally fit.
I already do when measuring something

The flex one I have is quite sturdy and not flimsy at all (at least not yet)
If the air/coolant mix does not work well with it, I probably just use it for air, stick the long nozzle on top of it and run the coolant separately, alongside the flex hose...

 

Continue with pneumatics...

Made a small block to attach the valves. Since they all switch different pressure levels the thing is not a manifold.
The outer valve is attached from the side, the others from the back. So I had to cut the top right bolt a little bit to get the clearance...




This is a neat little unit now, which can go easily to the board.




I made a french cleat board, which I can remove quickly from the wall if necessary.
Because of my wrong original calculation of the required air flow I had to buy a second compressor

The benefit now, however, is that I can dedicate one for high pressure and one for low pressure.

High pressure is
- tool change cylinder, 6 bar / 87 PSI
- Z-height probe cleaning, 6 bar / 87 PSI
- taper cleaning, 4 bar / 58 PSI
The required air flow here is minimal and more important is that 6 bar is always available.
I added a pressure switch for monitoring, but it should be good anyway...

Low pressure is
- air/mist cooling, 2 bar / 29 PSI
- bearing sealing air, 0.7 bar / 10 PSI
They are both on, when the spindle is turning. But since I have a dedicated compressor now for this, I dialed it down to 5 bar max.
So it stops at 5 bar and switches on at 2.5 bar. It has a 35 litre tank and is powerful enough to re-fill from 2.5 bar to 5 bar within roughly 1.5 minutes when both are on.
Means when I mill it runs 1.5 minutes, is off for about 1 minute, and so on...

The downside is that I need two air filters now. That is why I have two separate units at the top of the board...

Next up is all the wiring...

 

Having a detached garage and ~200 feet of air hose would come in handy.

 

Some minor progress with the pneumatics board.
Got all the wiring and tubing done. The external tubing to the spindle will follow, once the router is there...




I changed the control logic a little bit.
The valves for pressurizing the tank and for the air to the cooling nozzle are activated by the VFD (when the spindle is on), not the CNC control board.
The valve for the nozzle air can then be manually switched on and off by an external switch. So I can choose whether I have air or not, when the spindle starts spinning.

The CNC controller has only 4 relay outputs (Tool, Flood, Mist, Aux).
"Tool" activates a relay for a 230V power socket (like for a vacuum cleaner)
"Mist" activates the coolant pump on the Arduino board for the mist cooling
"Aux" activates the relays for tool change and taper cleaning

This way I have one command left, M8 Flood (which was supposed to activate the air), as a provision.
In case I find out later I need some kind of retractable cover to protect the tool holders from chips and/or dust...

The pneumatics board is in place behind the control cabinet.
It is somewhat protected from flying chips there and I do not expect a lot of maintenance anyway.
But since it is a french cleat board I can take it of easily if necessary.
The coolant tank is, however, still in good reach.

 

I also made a few changes to the control cabinet.
Added three relays as circuit breakers for the servo power lines.
They are NC when the controller is powered.
The E-Stop button contacts two NC switches and one is in the loop to power those relays.
So when I hit the E-Stop the three relays open and cut the power from the servos - rather than just relying on the control board to cut the step/dir signals.
Safety first

The actual E-Stop loop consists now of the second NC switch and a pressure switch in series.
So if (for whatever reason why) the pressure falls below 5 bar the controller goes in E-Stop mode as well.