[MechanoMan]

I've been reading up on this, making an MDF vacuum holddown bed. One with no holes in it, where the air infiltration through the MDF itself is enough for the vacuum effect. Which I have trouble believing, honestly, but that's what people told me and DIY info online shows working.

I'm a little unclear on a few points-

If I use "ultralight" MDF, does both sides still need to be surfaced? MDF is faced with invisibly sealed face surfaces that don't allow air to pass. I'm not sure if that applies to the "ultralight".

How deep does the vacuum bed's grid get cut, and what's width and spacing? Was that one sheet of MDF or did 2 get glued together to get the required thickness before cutting the channels?

What's the practical solution for facing MDF? I have 1/4" endmills here. I've seen some people advertising MDF fly cutters but they seemed mighty expensive. Otherwise, the math is self-evident- with 50% stepover at say 200 ipm, that's 5.76 min per sq ft. 3 hrs for a pass over 4x8. Which isn't impossible but the time sounds like a thing.

Most important:
Is it essential to glue down the MDF bleeder board? It sounds like the vacuum would hold it anyhow. I can't be sure how often this will need replacement. Cutting off the original one sounds like a big thing. I can easily picture a screw-up where someone drives the bit right through just one part of that bleeder board. It may be cost-effective to save it by gluing in a plug and having access to the bottom of the sheet would certainly help.

How much vacuum do we lose if the entire bleeder board area isn't covered?

[MechanoMan]

Huh... ok I found
http://www.shopbottools.com/files/sb...umholddown.pdf

What I'm noting here is that it takes a real serious vacuum pump.

[ger21]

The answer to most of your questions is "It depends".

What I'm noting here is that it takes a real serious vacuum pump.

Typically, yes, but a lot of Shopbot users get by with vacuum cleaner motors. Depending on table size, 2-4 motors are typically used.
Read through the threads at the shopbot forum linked at the bottom of this page.
ShopBot Vacuum Motors

If I use "ultralight" MDF, does both sides still need to be surfaced? MDF is faced with invisibly sealed face surfaces that don't allow air to pass. I'm not sure if that applies to the "ultralight".

I've never used ultralight MDF for vacuum, but I would still remove about .01" from each side. It can only help. Note that air can pass through unsurfaced faces of even regular MDF. It's not airtight.

How deep does the vacuum bed's grid get cut, and what's width and spacing?

I don't know if there are any hard and fast rules here. I think about 1/4" wide and deep slots on a 1"-1.5" grid is fairly common. The bigger the slots, the more air that can flow through them (in theory).

Was that one sheet of MDF or did 2 get glued together to get the required thickness before cutting the channels?

Again, I think this can vary, depending on how your specific machine is constructed. If you're table bed is an open frame, than I think many people bolt down a base, then glue the "grid layer" to that.

What's the practical solution for facing MDF? I have 1/4" endmills here. I've seen some people advertising MDF fly cutters but they seemed mighty expensive. Otherwise, the math is self-evident- with 50% stepover at say 200 ipm, that's 5.76 min per sq ft. 3 hrs for a pass over 4x8. Which isn't impossible but the time sounds like a thing.

What is your spindle capable of, and how much do you want to spend. If all you have is a 1/4" collet, than get the largest diameter 2 flute carbide tipped bit you can find with a 1/4" shaft. Should be about $25. Also, there's no need to use 50% stepover. Use 90% stepover. You can also go much faster than 200ipm, if your machine is capable. IF you have a 1/2" collet, you can use much larger tools. Just use the largest diameter tool you can get within your budget. Again, 2 flute carbide tipped straight bits work fine, and are inexpensive.
On an industrial machine at work, we surface with a 3" diameter cutter with carbide inserts, at 600ipm. Takes about 4-5 minutes on a 5x12 table.

Is it essential to glue down the MDF bleeder board? It sounds like the vacuum would hold it anyhow.

Depends on how much vacuum you have. If it's not glued down, you'll need a foam seal around the perimeter channel to minimize leakage.

Also not that if the spoilboard is not glued down, changes in humidity will cause it to warp. If the corners curl up, you won't be able to suck it back down.

I can't be sure how often this will need replacement.

Depends on how deep you cut into it, and how much cutting you do.
We typically try to cut about .003"-.005" into the spoilboard. With each cut you make, you'll start to see a decrease in holding power, as these .003" deep channels allow air to escape.

You'll need to resurface when you no longer have enough holding power. This can vary greatly depending on part size and the amount of vacuum you have.

I can easily picture a screw-up where someone drives the bit right through just one part of that bleeder board. It may be cost-effective to save it by gluing in a plug and having access to the bottom of the sheet would certainly help.

I've patched the spoilboard with MDF plugs on occasion. You only want to use a few drops of glue at the edges, though. You don't want the glue to block airflow. If you accidentally cut all the way through, I'd replace the board.

How much vacuum do we lose if the entire bleeder board area isn't covered?

Depends on how much vacuum you have, and how porous the spoilboard is. The simple answer is, you'll lose a lot. Probably an amount proportional to the overall spoilboard size. If half the table is uncovered, you'll probably lose at least half of your vacuum. That's why most people use several zones, to turn off unused portions of the table, and concentrate the vacuum where it's needed.

Also, as the spoilboard gets thinner, more air will leak through the unused portions, so you'll get more leakage.

[MechanoMan]

Have 18,000 RPM 3KW spindle. Yeah 50% stepover was excessive. I don't have a frame of reference for routing thin layers off MDF, except if it's only 1/4" the tooth velocity sounds like it's an issue, a 3" cutter would have its teeth moving 12x faster for the same RPM. It's a 5x8 bed.

It's an open-community shop. Everyone here's pretty good-natured but we're talking about a pack of monkeys here. I expect someone failing to set their tool height would lead to plunging into the bleeder board periodically. I wouldn't be surprised if I had to deal with serious board damage every couple of weeks. Hmm I could maybe wrap a tape line around the boundary? But I've not known tape to stick to MDF very well.

I think I've got a line on a used regenerative blower in town for cheap. So that would be great! I will need a filter.

Currently the 8x5 bed's 4x support pieces of 8020. So there's 3x areas. Which now that I think about it may be a problem. It sounds more efficient to take 2x pieces of 8x4, cut them to 5x4 and make 2 zones.

Initially I planned to do the plastic Raptor nails for a holddown solution. I have an 8x4x3/4" sheet of MDF where I cut 9x pairs of holes for double-hole nutplates and used nylon bolts countersunk into the surface to hold it down from the top. Well not "countersink", it's a flat hex bolt head, the hole's welled out so the bolthead drops flush with the surface and there's just enough clearance for a socket to turn the hex head. Nylon so in case of serious error it wouldn't destroy a bit, although it could certainly still find the aluminum rail. I was just gonna nail the second sheet of MDF on top of that.

Now I'm a bit puzzled how to attach the box to the frame in a way that can be disassembled without having to cut off the bleeder board. I mean my first thought was "ok I could just do basically the same thing" and cut the plenum channels around these attachment holes. I guess that's still possible because I could route out a hole large enough for a socket through the bleeder board each time I add a new one. But I'd have to make sure to index so I can find the underling holes because they're blind at that point. 'Cause otherwise I'd have to do something like make a lot of L-brackets all down the 8020 so it can bolt from below, and embed some captive T-nuts inside the bleeder board layer.

I was thinking "hey just bolt on the bleeder board with more nylon bolts welled out to be flush with the surface" but no, because if they're placed within the cutting areas there has be be a bare min of about 1/8" under the bolthead to take hold and about 1/4" depth for the bolthead's thickness. So starting with a 3/4" piece of MDF I could only use about 50% of the surface during resurfacings until I reach the bolt heads and then I'd have to toss it. Actually IIRC the way those holes actually went in practice these was only like 3/16" of clearance above the installed bolt head. Doesn't sound cost-effective.

LOL another reality just struck me. If I had a damaged bleeder board that was glued and had to be milled off, that means the ENTIRE mass of the entire sheet is gonna end up in the dust collector. I can only imagine the bulk and weight of the whole sheet turned to dust.

[MechanoMan]

OK I think the motor that this guy has in town is this:

AMETEK Technical and Industrial Products

This seems pretty limited performance, compared to your ShopBot Vacuum Motors link. Despite being 2hp (expensive to run that for hours) the free-air CFM is only marginally higher and the peak sealed vac is significantly less than the 512w motor there.

[seesoe]

check out the end of my build thread, I think you will find it very informative! Best of luck!
http://www.cnczone.com/forums/cnc-ro...build-log.html

[GLCarlson]

Huh... ok I found
http://www.shopbottools.com/files/sb...umholddown.pdf

What I'm noting here is that it takes a real serious vacuum pump.

75 bucks and a motor will get you that. Got mine from buy_our_stuff_please on Fleabay. Welch 2 stage Duoseal, rebuilt. Seller was dead easy to deal with, and honest about it- no guarantees, but "sounds OK". And it is OK. Absolute steal when you look at new price. Couple pulleys, a base, and done.

'Course, ya gotta know a bit about vac systems to be trusted loose with these- or you'll ruin it real quick. Dry-ice cooled cold finger vapor trap -or, at least, something to keep the big rocks out of the pump chamber, don't pull a lot of air through it, change the oil regularly (and use vac oil, not 30 weight!), etc. These are lab-grade pumps -not high vac, but standard rough pumps. Definitely qualify as serious vacuum pump for this use though.

[MechanoMan]

Welsh Duoseal?? I'm confused, that's capable of exceptionally high vacuum but not under flow. Datasheet says 0.2 cfm-0.62cfm depending on the model. I'd expect bleeder board flow to be much higher unless the board's very very small.

Did you actually try this out, or did you just get the pump?

I'm not clear on the flowrates involved. I did gather 7" of static pressure should be "pretty good" for hold down force. The actual CFM flow rate would depend on board area and board type and I'm not finding any info to go on here. Actually if someone specified the model # and the static pressure achieved, from the datasheet's static pressure chart that would give a guess as to CFM (but that performance chart isn't really exact). If I did have a flowrate I'd also need zone size and bleeder board type to understand CFM requirement.

[ger21]

You probably want at least 150 CFM.

[MechanoMan]

150 CFM for how much zone area? Now when you say "150 CFM", are you just referring to a blower's free-air rate? Because that's a very different number than its actual rate under static pressure.

Shopbot says 2"-4" Hg, 4" Hg is 53" H2O. Now this guy:

Black Box Vac Source - Let's Talk ShopBot

Brags about 7" Hg, which is 95" H2O from " 4 Lighthouse Brand LH-7123 240V motors. These are 10.4 inHg/110 cfm units each. Combined they should pull close to 10 inHg and have close to 400 cfm.."

But that's NOT the datasheet. Datasheet shows at 95" H2O that motor moves only 35 cfm per motor. If he's actually experiencing "holds over 9 inHg when applied to less than the full 4 zones" that's 122" H2O, so the datasheet says that's about 13 cfm per motor or 52 cfm for all 4x.

[ger21]

Ideally, you'd have 0 CFM when holding down full sheets.
The only airflow should be from leakage and uncovered area. A lot of air can flow through MDF, so you need a lot of CFM (free air rate) to maintain vacuum.


CFM is only used to maintain the vacuum, and doesn't really provide any holding power itself.
If you are holding your part with 10"HG and the pumps are drawing 100CFM through the spoilboard, it'll hold the same as if you had 10"Hg and 5CFM. The issue is that you won't be able to maintain 10"Hg at 5CFM, because you'll lose suction due to leakage.
As you start cutting parts out of the sheet, you'll get a lot of leakage just from the width of the router bit.

On our 5x12 table at work, we have two 10HP Becker's, rated at about 25"Hg and 175CFM. The gauges will show about a 15%-20% drop in vacuum from the start of cutting to the end of the sheet, due to leakage through the kerf.

How much vacuum you need depends on a lot of things. Cutting forces are a huge part, and so is the slipperiness of the material. The larger the tool, and the higher the feedrate, the more vacuum you'll need.
I doubt that 4"Hg will do much good.

[MechanoMan]

Yeah I see that you get basically 0 CFM under full coverage. Of course if you can't develop a pressure drop at significant CFM load you can't leave anything uncovered, and can't even get the full-size board to stick initially, if it's not able to draw flow through the uncovered board.

ShopBot actually says 2"-4" Hg (last page):
http://www.shopbottools.com/files/sb...umholddown.pdf

And all those "Lighthouse" motors that people are using, the spec sheets show a MAX pressure (zero airflow) of ~122 in H2O = 9" Hg, and the work product is max at about 5.15" Hg. So, 2" -4" is pretty likely to be common for people doing it.

[ger21]

I think that they're saying that you lose 2"-4" through the spoilboard. Not that you have 2"-4" of holding power.
Or, that you pull 2"-4" with an uncovered spoilboard. The vacuum goes up when you cover the spoilboard with your workpiece.

[MechanoMan]

I am unsure what they were saying when they worded it, yes. But that's just it. The 4" "loss" is the initial holding pressure and also the final low-coverage pressure.

Yeah with *complete* coverage you can approach the max vac. But that may just be excess. I don't think it can be "necessary". The whole scheme is that the "conventional" requires unused area to be covered but this "universal" undrilled bleeder board system does NOT require coverage.

I'm saying if you had a zone with 4" of static on it, you're supposed to be able to hold a small object on it. The static pressure has to be enough to hold it in place and it won't increase for small objects. Well maybe it'll develop 10" static when fully covered but if a 3" circle doesn't move under 4" static while being cut, a full sheet of plywood sure as hell isn't gonna move with 4" static on it. So it won't matter if it increases with larger pieces. The only important number would be static pressure when uncovered. Does that make sense?

[ger21]

I think what they're saying is that if your pumps(vacuum motors) can pull 9"Hg, than you'll actually have 5"-7"Hg at the table.

Yeah with *complete* coverage you can approach the max vac. But that may just be excess. I don't think it can be "necessary". The whole scheme is that the "conventional" requires unused area to be covered but this "universal" undrilled bleeder board system does NOT require coverage.

There's no such thing as excess with a vacuum table. You want as much vacuum as you can possible get, because no matter how much you have, you will have parts move on you.
I use one of these systems every day, with $20,000 in vacuum pumps. You most definitely need to cover unused portions of the table. A huge amount of air will be flowing right through the uncovered spoilboard, and while this airflow is decreasing your vacuum. And the more times you surface the spoilboard, the less resistance it has to this airflow. We use a 3/4" MDF spoilboard. When it gets below 1/2" thick, our vacuum level has decreased by about 30% of what it was with a full sheet.

I'm saying if you had a zone with 4" of static on it, you're supposed to be able to hold a small object on it.

Who says your "supposed" to be able to? The smaller the object, the less holding power you'll have. 29"Hg is about 14psi. 4"Hg is about 1.9psi. On a 3" circle, you'd have about 13lbs of holding force. So, it takes 13lbs to lift it up. However, the force required to slide it along the table is much less than 13lbs, so if you try to cut it too aggressively, it'll move. And once it starts to slide, it's much easier to lift up. Especially if you use upcut router bits, which will more than likely just pull the part up off the table.

This also assumes you have a freshly surfaced, pristine spoilboard. And grooves in it from previous cuts will greatly reduce vacuum on very small parts.
Small parts can be very difficult to hold down on this type of table. Onion skinning is often employed to help keep the parts from moving. You'll want to cut all the parts on your sheet, leaving about a 1/32" skin on the bottom, then go back and cut them all the way through, starting with the smallest parts first, because your vacuum will start decreasing with each part you cut out.

[MechanoMan]

Excess holding force doesn't hurt anything. Excess or unclear specification is a problem. The extra cost of an unnecessarily high spec is both in $$$ and project completion. It's a problem to pass up hardware that I can find and I can afford out of some misguided understand of the actual requirements. In fact if I say 10HP, our facilities manager is gonna nix it outright because we don't have panel capacity for another 10HP..

Hmm 3" circle may not be a realistic picture. I mean if you were using Raptor nails, you probably wouldn't stop the system to fire a nail into that circle, you'd leave it fixed with tabs.

[ger21]

If you want someone to tell you the bare minimum you need, it's not going to happen. Too many variable involved.

Many people get decent results with 2 vacuum cleaner motors. If you asked those people if they wished they had more suction, they'd tell you yes 100% of the time.

No matter how much vacuum you have, more would always be better.

[MechanoMan]

Yeah but it's always good to go questing for numbers. When you say "vacuum cleaner motors", is that the 3-stage Lighthouse stuff?

I mean the regen blower I saw originally, that was very big. And heavy. And expensive. But I don't think it'd perform well here, because numbers. A bigger motor may produce less static pressure in the end.

Actually I don't think there ARE that many variables. Starting with the assumption of a surfaced Trupan board and minimal leakage through the inactive sides, the CFM per sq ft under vacuum for a given static pressure should be constant. So it's just a matter knowing what that constant is, and what static pressures give minimum/good/great holding force and you could calculate zone size and motor selection from there.

[beachcnc]

I just purchased a nice used Sutorbilt (Roots Type) positive displacement blower...Rated 160 cu/ft min @ 15 in/hg using 7.5 HP @ 3,600 rpm...,,.I am mounting it outside my shop in a sound reducing enclosure and powering it with a 16 hp electric start gasoline engine...through a #6 LoveJoy coupling.....
I can pull higher Vacuum if I use water spray injection into the inlet of the pump...with an increase in the HP required.....