Building the ShopBot Vacuum Table

[nzbrock]

I work at the Facility for Arts Research at Florida State University as a Graduate Assistant (although I was working full time before I started graduate school).

I helped assemble out PRS Alpha 9648 and taught myself how to run it. We started off with Partworks as CAM software, but I quickly learned of its limitations. Since we primarily use Rhino 3d as design software, we decided on RhinoCAM as a perfect CAM platform for us.


After using the machine for about a year I realized the biggest drawback to our machine was hold down. Screwing sheets of wood down with screws always caused problems. A vacuum table for our ShopBot CNC router at FAR is something that will aid in the quality and efficiency of cutting flat materials. This can range from plywood to paper.

This table has been in the planning for a long time. I started by researching on this website to learn what aspects of a vacuum table were important. Since we already had a Quincy 10hp pump (two 5hp run together) with a 100 gallon reserve tank, I needed to make sure that I could adapt it to work for our system. The requirements of a CNC vacuum pump are high vacuum AND high cfm. This is something that lower end pumps cannot pull off.
We started by having FSU Facilities run a new 1.5″ copper line from the inlet of the pump on the mezzanine to our woodshop. This was a rather high expense, but it was offset by the fact that we already had a more than capable pump. The other benefit of this is that the pump will remain on the mezzanine and will keep the noise level down in the shop.

The type of plenum that I decided on is referred to as a “universal” vacuum plenum. The table consists of routed channels in the base layer, and a ultra light mdf spoil board glued directly to the base. The spoil board acts as a plenum by spreading the vacuum force over the entire surface of the table.
The existing table on the CNC router had to be dissembled in order to accommodate a vacuum plenum. It was taken down to the original plywood base sheet (.75″ oak) that is bolted to the table cross-members. A new sheet of mdf was then glued down to this oak base. Once the glue was dry, a .5″ ballnose bit was used to route the vacuum plenum and drill holes all the way through the base. Since this was an uncut bit, I ran our planing bit across the top to smooth it back out. Once the routing was finished the entire plenum sheet was sealed using thinned wood glue. This is a necessary step because mdf is surprisingly porous.


I added an inline filter in the room so that dust that enters the system can be removed in the room without turning off the pump. It is also clear cover so it can be visibly inspected on a regular basis.

From there I ran 1.5″ pvc to the front side of the table. This is where the vacuum manifold is mounted. This manifold splits the vacuum into 6 different zones across the table via the pvc ball valves. The importance of this manifold is to allow more vacuum on smaller materials without having to create a mask to cover open areas of the table. There are 2 24″x24″ zones and 4 24″x36″ zones.

[nzbrock]

From there, each pipe from each ball valve is routed to a corresponding hole that was cut in the table. Each pipe was glued and sealed to the table.
At this stage the entire plenum needed to be surfaced for two reasons; to make it perpendicular to the router and to create areas where the glue will adhere.

The spoil board of ultra light mdf was placed on top of this plenum and “skinned” to remove a denser outer layer. When this was finished wood glue was once again applied to the plenum, only this time it was rolled onto the “table tops” and not the channels. The skinned side of the spoil board was carefully lowered onto the plenum and vacuumed down to permanently adhere it. Wood and other non porous materials were placed on top to create more vacuum below the spoil board.

The reason for gluing is to limit movement of the table due to humidity and temperature changes. All of the edges of these boards were sealed as well by painting them “ShopBot Blue”.
Each zone has its own vacuum gauge so constant monitoring is possible. If vacuum were to fail while cutting it could turn into a disaster. Vacuum is measured in inches of mercury or water. Theoretical vacuum (absence of air) at sea level is 29.9 inHg because it is the equivalent of one atmosphere of pressure (14.7psi). With all zones open and nothing on the table, the pump can pull 13 inHg.

With a sheet of wood covering two zones the pump can pull 21 inHg at the table.

[nzbrock]

Once parts have been cut through this wood the remaining vacuum was 11 inHg.


Techniques such as masking completed areas can be used to increase the vacuum pressure as holes are created (cutting though the material).

[kevin]

Nice pis are you sure you cut looks like an operating room

Sometimes i wish i would spend more effort in a proper vacuum thanks for posting

[nzbrock]

Nice pis are you sure you cut looks like an operating room

Sometimes i wish i would spend more effort in a proper vacuum thanks for posting

The vacuum is sooooo much better than having to use screws and clamps.

[rb99]

Nice...

A map to what I need to build. Thanks!

Do you plug the unused vacuum holes when not using those quadrants?

Does anyone know how the "Bradyvac" compares to this vacuum?

[waterwheelman]

I am using a system similar to the Brady vac. I have two Lighthouse motors mounted on a wall in a separate room from the ShopBot. This reduces the noise level. The plumbing is 4 inch PVC run to the ShopBot with 2 inch ball valves and connections to the underside of the table. The 4 inch pipe is about 25 feet long, which gives it a little bit of a pre-vacuum chamber. I have 4 zones on a 48 x 96 table. I can hold a wide variety of parts from whole sheets, to small parts using templates. I run the two vacuum motors in a series configuration to give me about 12 to 14" of Hg. The motors are not mounted in a box. They have complete ventilation around them for cooling. I have only burned out one motor in about six years of medium to heavy usage. I am considering adding two more motors to the system to boost the vacuum and cfm when cutting difficult parts to hold.

[nzbrock]

Nice...

A map to what I need to build. Thanks!

Do you plug the unused vacuum holes when not using those quadrants?

Does anyone know how the "Bradyvac" compares to this vacuum?

Because the table is set up in 6 zones with valves I rarely have to do much masking. It usually involves throwing flat scrap pieces on the open areas of a zone.

[Brady Watson]

With all zones open and nothing on the table, the pump can pull 13 inHg....

With a sheet of wood covering two zones the pump can pull 21 inHg at the table.

Noah,
Thanks for posting pics. It looks like you put some real effort into your vac system. Bravo!

You mention that your max Hg" value is 21, but you have 13 Hg" with nothing on the table and all zones open. This tells me that your bleeder is most likely not true ultralight MDF or LDF, and/or it has a significant amount of resin content which prevents it from bleeding/breathing very well.

So...While your vacuum shows the potential of 21 Hg", in all reality you are achieving a max of 8 Hg" (delta) of usable vacuum. This is a little bit better than a good shopvac, such as a Fein. Good strong vacuums, such as yours can get away with regular MDF or LDF that has more resin than is ideal - However, you are missing out on a substantial amount of hold down power with your current bleeder board - a factor of 2.5x what you are getting now. You are getting 4hp of performance out of your 10hp setup. The rest is being wasted.

The only way to really rectify the situation is to machine off the bleeder you glued down to the grid. Extra points for thinking about them moving together. Going forward, the only way to tell how good your bleeder will be is to open up all zones with the new sheet on the grid & check the gauge. Ideally, you want 0 Hg", but since you have such a high powered system, up to 5 Hg" is acceptable.

There is a possibility that the small inlet pipe is hurting your performance as well, especially if it is on a long run with a few bends in it. This can cause the pump to cavitate and be under suction with no 'load'/valves closed because it is choking.

Also, this business about machining off the 'paper layer' on one or both sides of the bleeder is an old wives tale. Save yourself the agg and just put it on there as-is, then flatten the top to true up the surface & you're ready to work.

Hope that helps!

-B

[Chuck Keysor]

Thank you Brady for your careful reviews of posts!

I read the note that "With all zones open and nothing on the table, the pump can pull 13 inHg.", and I thought, gee, that is odd, shouldn't this go to zero? But then I figured I must be confused, and left it at that. I would guess plenty of other people read the same thing and figured they must just be confused. So I was glad to read your clarifying post on this particular point. Once again you provided a valuable service to the forum readers by providing corrective information, and educational insights!

Thank you, Chuck