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

The router. A Sorotec Compact Line 0805



It will be a kit and I have to assemble everything together. Should be fun
Travel is 820mm in X and 510mm in Y.
That is alI can fit into my shop.

I debated long between 410mm or 510mm in Y.
410mm has more stability of the gantry, 510mm has more flexibility.
I went with more flexibility ("larger" woodworking projects, part nesting, blablabla...) and decided to beef up the design for improved stability.
Going to replace the casted zink profil connectors with some sturdy DIY machined aluminum angles.

I ordered it without Z-axis, because I do not like the design.
The carriers are attached to the Y-axis cross plate and the rails to the spindle mounting plate.
The benefit of this design is that the ball screw and the rails of the Z-axis are protected from any chips. But the ones on the Y-axis are not anyway - so I do not see the point.
The disadvantages (imho):
- reduced Z travel. It can travel 170mm, which is the distance from fixture plate to gantry. But you can't use this for milling. Best case your collet is flush with the bottom of the spindle mounting plate, then you still loose the endmill height.
- limited spindle height
- servo/stepper motor moves with axis - has to move its own weight
- no option for Z-axis brake

So I will do my own Z-axis.
Hopefully it will turn out like this




This has 200mm travel. A standard 57mm long endmill will hover 10mm above the fixture plate in the lowest position. In my case this is a safety feature
Do not have to go deeper, because there will be some kind of sacrificial board anyway.
And if needed one day, I can still remove the upper spindle holder and move the spindle further down (with reduced stability)

But most importantly, in the upper position I can now indicate pieces with a Haimer up to a height of 140mm above the fixture plate, as to 50mm max in the original design.
So working on pieces in a vise will be possible now.




And I added a brake for the ball screw, which hold the Z-axis in place when the servos are not powered.
The whole thing is above 10 kilos.


When finished it should look something like that.
Pneumatics and mist cooling are still missing in this screen shot. They will go in the far corner, next to the control cabinet.

 

I started to make a little bit more mess than usual in the shop...in order to put some more chipboard to the walls...




I usually stay away from this stuff in wood working, but for this purpose it is ok. Once installed you can easily bolt things to it without a drill orgy.
Like a french cleat system, for more versatility in the shop.

 

With new space gained, I started with the control cabinet for the router.
First time ever for me to build something like this.

When I converted my BF20 to CNC some 15 years ago I build the control unit myself.
There was a guy in Germany at that time who published a book about that, with explanation of all the circuits....PSU, control board, stepper driver, I/O boards, you name it...
He sold as well the naked PCBs. You had to get all the components yourself and than solder everything together.
Then put everything in a steel case and you were good to go.
Amazingly enough, it still works flawlessly today
But the wiring ("cable management" as we would say) inside is a desaster and in case of a failure these days, I would be completely at loss.

So I decided to go a little bit more "professional" this time.

Started off by putting the circuit down and created 12 pages like this, so in 10 years time (when I am old and grey) I still know what is what...

 

A box full of PSUs, relays, terminal blocks, miles of cable...




Started with the mounting board, outside of the actual cabinet




That is the basic layout.




- An RFI filter for the VFD. The VFD is still missing here, but it goes on top of the filter.
- Circuit breakers/fuses for VFD and router control. They have seperate power lines coming from the house.
- A relay for the spindle fan (I planned originally with an air cooled spindle...)
- Several PSUs for the control board, spindle fan, break, mist cooling, cabinet fans and the 48V PSU for the servos

- Resistor for spindle break
- 2 relays for air and mist cooling
- Ground terminal
- CNC control board (Eding720)

- 2x terminals for incoming main power source
- a circuit breaker/fuse for an extermal power socket controlled by the Eding (to hook up a vacuum cleaner that turn on with the spindle)
- relays for the external power socket and the ball screw brake and
- several terminals to connect the external wiring to the router (the thing will never move around, so no connectors needed)

 

Some wip of the wiring process.
This is actually worse than doing the cable management in a PC...




Preparing the cabinet




Adding the cable glands and an air intake.
This will pull the air from outside of the router housing, so in case of milling wood, the cabinet is not poluted in seconds...(maybe in minutes )




And installed at the wall



I am trying to seperate high voltage lines from low voltage lines.
Left is all the 230V area plus the well shielded cable between VFD and Motor, in the middle is the 24V area, right is the 48V area for the servo power...

 

So far so good...if only I could have stop myself at this point...
I debated back and forth, checked my budget a thousand times - and decided to go with an ATC spindle from the very beginning, rather than upgrading at a later stage.

It will be water cooled, rather than air cooled, which triggered the build of this PC in the first place...

Fortunately, the dealer is very complaisant on that.
They will take back the spindle and VFD (which are already like 6 weeks at my place) and swap it with the ATC spindle.
The VFD need to be changed from a 400Hz to a 1000Hz model.

With that, the control cabinet needed some modification.
I swapped some of the 24V 20W PSUs to a 24V 60W model and added another relay instead. The one originally for the spindle fan is now for the spindle nose pressurization valve (to keep dirt from the bearing).
The new relay is a safety release for the ATC. It is open when the spindle rotates and closed when it stopped turning. The signal to the tool release relay goes through this "safety" relay, so you can not accidently release a tool holder manually when it spins at 30000 rpm...
Both relays are controlled by the VFD, that is why they are placed close to it.

 

On the bottom of the cabinet I added a second mounting plate for the ATC control.

- a terminal block for the spindle sensor connection. There are two sensors in the spindle that monitor the ATC status (clamped, open, clamped w/o tool)
- two relays controlled by those sensors. The VFD start signal goes through the two relays and the spindle can only start when it is clamped with tool)
- the relay for the tool release/cone cleaning valve

 

The new spindle and VFD will be here in two weeks.
Until then I look into the pneumatics - another completely new field for me
Then I need to make some chips and mill the Z-axis parts...

 

Very impressive...so clean and professional

 

Yeah, it looks quite clean, once the covers are on the wiring channels...

The goal was to have everything inside one cabinet, rather to have the CNC control, the VFD and the Pneumatic control in different boxes.
I hope I can limit any interferences to an absolute minimum, if not zero by the strict seperation.

It is no rocket science though, if you can read a spec sheet and are willing to do some research.
However, I have literally no experience in this.
So please, if anybody of you guys do this for a living or did already in the past and you spot some obvious errors - please let me know

 

I made the mistake of trying to read all that while still sleepy.
Fellow extended Z height enthusiast!

 

Ohh my, a new toy is in town...




The exchange of the spindle took less than expected. Returned the old one on Monday, got the new one today.
That is a rock solid service...

It is a 2.2 kW version with 30,000 rpm.
The high rpm result in a somewhat reduced torque of 0.7 Nm (~0.5 lbf.ft)
I would have preferred a 24,000 rpm version, like the Teknomotor spindle I chose before, but you can't have it all.
Need to go a bit easier on the chip load.

When I decided to go with an ATC, I debated whether I should sell the "old" one (which was brand new, never used) on ebay and get a "cheap" ATC spindle from China.
The Chinese ATC spindles are around 50% cheaper, ddp to Germany (19% VAT + 8% duty). Taking the loss on the potential ebay deal into account, maybe 30%.
JGL spindles are currently hyped on the net, seem to have a decent run-out, but at the end of the day you are either lucky - or not.
No warranty, no returns, and appearently lately as well no customer service.


This way I have a 2 year warranty from the dealer, plus a manufacturer who answers e-mails within minutes
The visual quality is quite decent. Can't wait to make the first chips (not before August though )

 

Those are some serious spindle mounts. I likey.
-30,000 rpm... My ears are ringing already.

 

Yeap.
The router manufacturer recommended in fact to use two in a stack in order to reduce chatter... (or maybe just to boost his sales )

They are a nicely anodized extruded profile, with a reasonably clean saw cut on top and bottom.
Both sides on the flange, however, are milled. So I expect them to be dead parallel to the clamping diameter. We will see...
The two outer screws are for clamping, the middle one to widen the dia for spindle disassembly.

Again, you can get them at half the price at Ali, but I find 44 EUR reasonable for the quality and the milled sides should help a lot with tramming...

 

I am deeply satisfied with the service of the spindle manufacturer.

I have noticed these "unclean" holes on the bottom of the spindle. Since they are in the turning part which can rotate with 30,000 rpm, I was a bit concerned about balance.
The clean "5th" hole is in fact for balancing, the 4 "ugly" ones are for assembly. A simmilar tool than for fitting a disk to an angle grinder goes in them at spindle assembly.
Looks like someone has too much muscle...

I emailed this picture to the manufacturer (not the dealer) and they immediately offered a repair. 10 minutes later I got the RMA number

 

Fairly slow progress these days.

Spend the morning to figure out the air pressure consumption of the spindle plus the mist cooling.
As a result I need to buy a second compressor

Then I spend the afternoon to turn this...



...into this




And the part is still not finished...Not my speediest work, but these holes and threads have to be acurate...

 

In the meantime, however, to add at least some productivity to the day, I dusted off this little piece of equipement...




It has been a while that I printed something and I had to clean the tank first. Amazingly enough, the almost 1.5 years open resin cartridge was still ok...
Surface quality is not really important here, so I choose a 100 micron layer height, the poorest, but fastest quality setting.
Still took about 8 hours to print. Plus 20 min washing, plus 2 hours UV curing...

In fact it cures right now, so I can not show the finished part yet. In this condition it is rather soft and if you touch it, your prints are perpetuated...




Unfortunately, it will not stay that clear after curing...

 

Part is cured and support structures removed.
It needs a light sanding where the structures were attached...




Press fitting / super glueing the M3 thread inserts




Fits quite well




If this were a part with a visual requirement, I had to change the IPA bath first. There is some residue from a previous wash of black resin parts, though only on the inside...
Anyway, it came out quite nice.

Material is a so called "technical" resin, hence the fancy color. The black, white, grey, clear, ... resins are far weaker in material strength.
You can still dye it to a darker color and it keeps its tensile strength. It is a bit more effort, but then you have the best of both worlds.
Or you give it a light brush with mineral oil (like for butcher block sealing). Then it goes back to the transparency like after the IPA wash.
I debated first if I dye it black to match the motor, but decided to leave it like this...



The bow in the bottom edge is in fact designed like this, because the terminal holder of the motor is wider at the rear. It is a rather strange design and you wonder if it was designed right after Chinese New Year celebrations...

 

Finally milling the other side.

Unfortunately my mill is too small to do a 165 mm wide part in one go, despite removing all the chip covers.
So I had to re-clamp. (It's time for a router...)
Means everything related to Z-axis tramming later will have to be aligned to the first side. At least I could do all threads and holes in the first clamping.
Accuracy of this side is not so important. But I managed to get it nailed to 0.02 mm on a 400 mm length anyway.

The aluminum fixture plate (under the wooden board) I made a few weeks ago helped a lot. Greetings to Alex Banks from Bit
His YT video on that matter was a great inspiration...




Parallelism of the left linear rail to the left side is dead on.
At least I can not measure a difference top to bottom with my equipment, which has a resolution of 0.01mm, or 0.4 thou for you imperial guys out there...

This is just a test fit. Once this will be finally assembled and all the bolts are tightend down to the correct torque I will remove the dowel pins.
The right rail will self align, once the spindle mounting plate with the carriers is moved up and down a few times...

 

Good #s.
I stopped checking mine because the chinese parts were a horror show.