CNC Vacuum Table Build (WIP)

Thought I'd post my vacuum table build progress here - it'll be a slow WIP due to limited time in the shed.

I'm not a vacuum expert, so I'd welcome thoughts, opinions and discussion here. I'm working things out as I go along.

When I was thinking about playing with CNC's, I had a look at fellow forum member Pippin88's machine, and he casually mentioned that much of CNC thought-life is about work holding. At the time I thought there were so many more interesting things to think about, but it turns out he was right. There's nothing quite like cutting a perfect piece, only for it to slip loose and ping off into a cutter spinning at 20000rpm on the last cut.

Apart from the obvious issues of running the machine into clamps, and slicing up your shiny hardware or destroying your expensive cutter, there are less obvious things like cutting order to think about - cut out the insides before you do the outside, because once you do the outside, the inside is no longer clamped, and it'll move about. Or cut a channel close to a clamp and you suddenly lose clamping force as the wood bows in. Offcut box. Start again.

A solution is a vacuum table to suck those parts down. Particularly useful if you cut up a lot of dodgy cheap plywood with all those 'feature' voids and warps.

For sheer vacuum force, it seems that a rotary vain or liquid ring pump is best:

Attachment 509919 Attachment 509920
Problem is, for a given size, they make lots of vacuum, but not much flow - and we woodworkers need flow in our vacuum setup to compensate for the holes we make in the workpiece - as you make more holes, so you lose suction. A bit like occluding the end of the vacuum cleaner - completely occluded, there's a lot of suction. As you occlude less of the orifice, you loose suction. You can get around this by increasing the gas flow - so stick your vacuum hose on the end of a giant dust collector, and you'll still get suction with only a bit of the orifice occluded.

To get enough flow from rotary vain or liquid ring pump, you have to get something really big, which needs a lot of power, and a lot of cash.

So most woodwork systems use vacuum cleaner motors. They move a lot of air, but don't generate a huge amount of suction (relative to a rotary vain, for example). The solution is to use multiple vac motors. They're relatively cheap, which is good as because they're primitive, inefficient machines doing something they were never designed to do, they lose all their smoke regularly. It seems that people running these systems routinely keep a spare motor on the shelf so they can replace the inevitable failure.

Attachment 509921

Probably wise to have a fire extinguisher nearby too!

The Shopbot forums have some lengthy threads around the original blackbox vacuum concept, which eventually became the Hurricane systems so favoured by youtubers. The bloke who developed the blackbox and started the discussion eventually got fed up and left the thread in a bit of a huff - he seemed upset with people messing with his designs, changing things and building their own interpretations. Not sure what he expected of the internets! He now appears to have a financial relationship with the makers of the hurricane systems, so I think he's still involved in development, and definitely not talking about it on the open source internet.

It's a pretty simple design - mount 4 vacuum motors, connect the suction end to some PVC pipe, and stick the end of the pipe underneath a piece of MDF. MDF is porous, so the vac sucks through it and clamps your work down.

Inside it looks like this:

Attachment 509922

The baffles on the outlet side (I guess) are there to reduce the noise - these vacuum motors wail like a banshee.

A simpler variation uses silencers on the outlets:

Attachment 509923

And another flavour

Attachment 509924

So, my plan is to make something blackbox-like. Melamine-coated something for the box - MDF, ply etc are all porous to some extent, so you potentially lose vacuum. Preferably something not too flammable (possibly why the commercial Hurricane is now made out of metal).

I'll use melamine for the bottom half of the vacuum table too, and MDF for the top. I'll have to play with it a bit as you don't seem to be able to get the ultra-light (more porous) flavours of MDF here in Aus.

Attachment 509925

Im liking the idea, looks like you are going to need a 3d printer, printing some removable TPU gaskets/infills could reduce the suction requirement by reducing the vacuum area for the material size that you decide to mill.

Quote:

---

Originally Posted by Bernmc

we woodworkers need flow in our vacuum setup to compensate for the holes we make in the workpiece - as you make more holes, so you lose suction.

---

The way we got around that at work with losing pressure on small parts (even a 10hp pump still has a minimum workpiece area before it just won't hold on) is to cut the whole program 1-2mm above full depth so everything is still attached and you get full suction, then come back and take the last little bit to separate the parts with minimal tool pressure and very low chance of losing anything.

Our machine was a pod and rail type with a home made flat table conversion of the same grid style as your last pic in compact laminate sheet; it's basically the same material as laminex, but it's a full 19mm thick sheet instead of a 1mm laminate. Costs a fortune, but it's hard wearing, stable, and vacuum tight. It's also available in 13mm thickness, sometimes used for cafe table tops

You can use commercially available square section air tight foam to section off the grid if you want to.

The commercial machine I have (Multicam US) has the vacuum table made from some hard plastic. I would love to say its HDPE and that is what it looks like, but cant be 100% sure.

I dont use the vacuum table very often, rather I have an MDF spoil board bolted on top and I just screw the workpiece on. The vacuum table is really only useful when you are doing production runs of nested jobs. I have a 30hp pump and like elanjacobs said, i program it to cut 0.5mm short. Not cutting to full depth also helps stop blocking air filters.

Interested in how you go with this.

My thinking: Vac hold-down has two roles: 1) Large sheet production work / making a lot of stuff out of sheets (large spoil-board vac setups are not good for small parts, particularly if you are cutting the full perimeter). 2) Small parts / specific parts using vac hold-down blocks / pucks (something like Axiom AVK500 Vacuum Hold-Down Kit | Rockler Woodworking and Hardware)
(take with grain of salt, this is based on reading not personal experience).

I'm currently using a variety of:
MDF spoil-board
Doubled Sided tape + side blocking
Screws or nails though waste parts of sheets into spoil-board
Nails through parts into spoil-board when part finish isn't important / parts going to be painted
Side blocking - fence / blocks on one long and one short side, then opposing wedges on the other two sides

I bought a Fein shopvac years ago, partly because it has separate cooling and is meant to not overheat when there is no air flow through the hose intake. Most vacs depend on the flow through the hose to cool the motor.
Have never got around to using it for vac hold-down

I have previously tried:
Aluminium T extrusion bed - didn't like this, expensive and you don't want to cut into it. I think well placed T track with replaceable MDF bits between (that sit higher than the T track) is a better solution
Threaded holes - I actually have a bunch of T nuts in the bottom of my spoil-board in a grid. But the T nuts don't have a nice thread transition and are a pain to get a bolt into. Standard nuts in a recess with a dab of glue would have been better. In the end I just glued more spoilboard over the top and have not bothered to bore holes to uncover these T nut holes.

Some useful suggestions.

I have a 3D printer, with another large format beast on order. I may get around to making some pucks for specific tasks, but my setup would be better served with a large area table. I'll section off zones as Elan suggests. Some EPDM on order.

I'm not a fan of screws for holding - apart from leaving dirty great holes if they're in the part itself, they're just as bad for the bit as a clamp. I do have a composite nailer which works very well. The nails can be sanded and painted, so they leave less evidence of their brutality (unlike Putin).

The job that got me thinking about vacuum tables was a sheet cut into multiple drawer parts. The sheet shifted towards the end - just a little - and I ended up with a bunch of parts cut out of square. I'd need to nail each part individually to stop that happening, and I'd inevitably miss something.

I like the idea of leaving a mm or so to come back to in a final cut.

Using a down spiral is also a big help in keeping things stuck to the vacuum, but you can't plunge with them more than a few mm

WARNING: If you are Australian, your government has decided you are too dumb to work with electricity, so don't try this at home or you and yours will all die horribly. It's as bad as dividing by zero.

Motors, motors, motors. If you're a yank, then this bunch are a popular supplier of engines for vac tables. The model that seems to be used the most is a 3 stage 220V unit that sucks 148.8 inchH2O (37kPa), moves 110.3cfm (3.12m3/min) and draws 7 amps. It sells for $145 before post - a significant issue if you need to get it to Straya.

You can pick up an Ametek Lamb unit with similar spec here, but you're looking at $250 and up. I thought I might be able to find a better solution.

Conventional vac motors have brushes which they chew up in 750-1500 hours. They don't appear to be particularly efficient. I wondered whether I could find a brushless DC motor that could do what was needed.

It took a considerable amount of trawling the Chinese internets, hijacking satellites and inserting ransomware into the Chinese government servers, and although I was assassinated twice by poisoning, I eventually found a Chinese company that were an absolute joy to deal with. Responsive to my queries, appeared to have an excellent product, and were willing to sell me fewer than 4000000 units.

They had this little beast: The NXK60C122Z11

Attachment 509993

It's a 220V, 2 stage brushless DC motor. Or as the company says

Attachment 509994!

No brushes to worry about, DC efficiency, and even better, speed control - either by varying a 5V supply, or by PWM - so potential to build an arduino-based feedback speed control at some point. Decreasing vacuum -> increase the power.
Being able to control speed should mean I can adjust it to supply the required vacuum, rather than run flat out all the time, and so reduce motor wear, heat generation, fire and sudden death.

The spec sheet says it's a 1200W motor which sucks 31.5kPA and blows 3.27m3/min at 8amps. Not far off the lighthouse AC unit. And. No sparkle.

Buying from China is always a bit of a risk. It's not uncommon for a bunch of quoted specs to be way off. Not such an issue when you only end up wasting a few dollars for some little electronic bauble, but pretty painful if that turned out to be the case with these motors. They cost me around US$170 each by the time they landed on our fair shores - freight is astronomical in the post COVID era.

Attachment 510000

So here's the basics for each 'channel' of the vac:

Attachment 509999

Some closed-cell silicon mat to cut gaskets out of, the motor and its controller, switches, a vac guage, and a couple of ammeters.

I 3D-printed a quick vacuum cup to test the motor. I modelled it in Fusion 360, complete with the 1/4 BSP thread which printed surprisingly well. I just screwed the gauge in and off we went. No tapping required.

Attachment 510001

And then, the moment of truth...

Attachment 510002 Attachment 510003

That's -0.34 bar or -34kPa at full chat.

And the electrics:

Attachment 510004

I have no idea of the accuracy of my cheap Chinese gauges, but it appears that the motor does what it says on the box.

My plan is to have one of the orange displays on each channel, and the blue display measuring overall current and power.

Current draw is an issue - when I upgraded the shed electrics, we put in two 15A and two 10A circuits. Who could want more...?! I have the CNC on one 15A circuit, and the DC on the other. So for now, I only have the capacity to run two vac motors at full chat - one on each 10A circuit.

I only have a 4x4 CNC table at the moment, so I think I may start off just building a 2 motor unit and see how it goes. Most of my CNC work is in the bottom left corner of the spoilboard, so that shouldn't be an issue. I'll have to see if there's enough suction across the full 4 x 4 with two of these motors. Alternative would be to build a full 4 vac unit, and run them all at half-ish speed...

The expense and effort you are going to is impressive. I did toy with the idea myself but ultimately decided it was too much for the baby hobby machine I have.

Thinking about the power cost, 2.6kW for two pumps at 30c/kWh is 78c/h. I can buy a lot of glue and tape for that sort of money especially on a long cut and because I just play my time investment is a non issue.

Looking forward to seeing how it goes and if you think the method economical for your purposes.

Well, I've put a $60 Fe$tool dowel boring bit through a $60 Fastcap clamp, so that'd pay for 154 hours of vacuum...!

Like all of these non-commercial projects we mess about with, the value is in working it all out and making something. Realistically, there's no way the $1500-odd the system looks like costing will ever pay for itself in real money. And I can see myself getting irritated with the electrics and calling my sparky for another eye-watering day of $'s for another circuit.

But it all stops me going senile. If your semi-opposable thumbs can't make pretty furniture, you mess with machines that might!

As a qualified electrician I cannot condone the use of electric motors with a less than optimal quotient of sparkle. I often install extra sparkle in my circuits just in case.

Vac tables are just so worth the effort its not funny….no more clamps no more thinking about bridges….and the extra cost of running a good vacuum table will be a non issue once you have it sorted. The only down of a good vacuum table is really the noise.
I’m sure you are putting in heaps of effort and really the only negative is the amount of power you need to run one. If you don’t have buckets of power just run the machine a little slower to reduced the cutting force on the part your cutting.
And to be able to redirect valves for smaller isolated areas are a big advantage.