Need help with Vac setup.

[dray]

I have been pondering this since I recieved my cnc 6 months ago and am now ready to make the change to a vacuum table setup.

I cut 90% 3/4" mdf and lately have been doing alot of rather small parts (12"x12")

I am at sea level (there about) I am wanting to go with the 3 fein setup but not sure if its strong enough to hold down my pieces.

I have been avoiding the vacuum threads because I knew I would eventually have to do it so here I am..

I cut 3/4 UL MDF in 2 passes at 600 inches per minute, can I stick with a down spiral bit?

Whats the best design?

Will it be strong enough to hold my material down?

All help appreciated!

[myxpykalix]

Here is one drawback i have found with the vac table setup (which is GREAT). When i cut mdf for my wainscoting i found that if i cut into the table (which you will do) if you leave a cut of the part in the table and then put something smaller on the table to cut it acts as channels to let the vacumn escape around the part and won't hold down worth a darn. So you might be surfacing your table more often but by far it is the best method of holddown. I just cut thru a metal clamp and destroyed a crappy bit in the process holding down a 1.5" thick piece of maple. So its not the end-all-be-all but for what you do i think it is a good solution.
I use 2 feins with a 4 zone pvc pipe setup.

[beacon14]

Jack is right about the accumulated grooves affecting the vacuum but if you are mainly cutting full sheets it's not as much as a problem since the grooves rarely reach the edges of a full sheet. One trick when cutting small parts by themselves is to cover the rest of the table with scraps, holding the scraps tightly against the workpiece when you turn the vacuum on. That effectively increases the square footage under vacuum and helps keep small parts from shifting - sometimes it's faster than turning zones off and on if you only have one small part to cut.

The down spiral works fine with a vac table as the downward pressure of the bit helps keep the sheet down. Also since it tends to keep the kerfs packed with sawdust that helps maintain the vacuum as well.

Worst case is you may have to experiment with toolpathing strategies when cutting small parts but the bottom line is if you make yourself a decent vacuum table you will not regret it - you may kick your self for waiting so long to do it though.

[myxpykalix]

Also one tip after you have glued your spoilboard down and surfaced the face off take a tube of silicone caulk and coat the edges of your mdf all the way around as that is a big vacumn leak point.

[henrik_o]

Danny,
I’m a novice when it comes to vacuum (and so much else), but for your application I would seriously consider going right for an industrial blower instead of the Feins.

You are cutting right through 19mm MDF at very high speeds, you need both strong hold down and a LOT of flow.

My Swedish sales rep has quoted me a price on an URAI 45 PD blower at $1650. That’s including shipment to Sweden and to my door. That’s for the bare blower; it needs to be kitted up with motor starter, filters, silencer, relief pressure valve etc, but these are standard parts that do not cost much. Say $ 1900 (I forget the exact quote but it’s in that region) – and that’s my price here in Sweden for the whole kit pre-assembled at my door. For you in the US, it should be a couple hundred dollars less due to shorter shipping.

I understand that a Fein III Turbo can be had for about $350. Three of them adds up to $1050. You’ll need to add shipping and probably some items to kit it, let’s say $1200.

As such, the price difference 3xFeins/1xURAI is somewhere in the region of say $500-700.
For that money, instead of some daisy-chained shopvacs you are getting an industrial grade vacuum system. The 15hp URAI 45 pulls (approx.) 16”Hg @ 325 CFM. Want to move up a bit? Add about $700 and get the URAI 56 which pulls (approx.) 16”Hg @ 500 CFM.

There are some downsides. The URAI series is power hungry and VERY loud (over 90dB). You’d need to put it in an adjacent room or sound insulate it.

My budget for a competent vacuum system was in the region of $6000. Given the savings I have realized by doing some research, I’ll either pocket the change or do some more kitting. One thing I’m looking into is electric actuated valves and reducers and an industrial grade control system. It might be a squeeze to get it inside budget, but man, how nice wouldn’t it be to control vacuum zones via touchpad and also be able to direct stronger vacuum to a certain zone or just let the system take care of it: if one zone starts to lose vacuum it can automatically compensate.

In a certain sense, I curse our trusty old ELMO regenerative; it’s working so well right now the motivation to go through and finalize a truly competent system is somewhat lacking. I mean, looking at the local used market there’s an Altendorf on sale now for about that money … or that Holz-Her edgebander I’ve wanted for so long.

Anyway, there are better minds here that can guide you, but do your research and you might find industrial systems are not as expensive as you would expect. And while I may well be wrong, I find it a bit hard to believe you can throughcut small parts in 19mm MDF at 600”/m without an industrial system.

[dray]

Thank you henrik, Ill start looking into industrial models.. I was under the impression I would pay upward of $5k then still have to do plumbing which would be a whole day of my time along with parts cost.

I do cut 600"/m but 2 passes, I am not sure if this matters. But I really do not want to slow down production for feins if they wont do the job

[dray]

is this large enough?
http://cgi.ebay.com/5-HP-ROOTS-BLOWER-MODEL-33-URAI-4634_W0QQitemZ120129510173QQihZ002QQcategoryZ26227 QQrdZ1QQssPageNameZWD1VQQcmdZViewItem

[beacon14]

I think you can still cut in two passes at that speed, but maybe not equal depth passes; by leaving a very thin skin with the first pass and cutting through with the second you should be able to cut pretty small parts with a multiple Fein system.

[dray]

I also just found this but it closes soon
http://cgi.ebay.com/REPUBLIC-MANUFACTURING-HRB-900-REGENERATIVE-BLOWER-3-PH_W0QQitemZ150159285279QQihZ005QQcategoryZ46548QQ ssPageNameZWDVWQQrdZ1QQcmdZViewItem

[thewoodcrafter]

Danny,
I bought a 15HP regen blower for mine about a year ago and glad I did. I have an old PR with a PRT upgrade so I cut at a fraction of your speeds and with some small cabinets parts I need to add tabs or they will still move around. It was one that ShopBot sold with there machines about 3 -4 years back.
Anything smaller and I think you will have problems with pieces sliding around.

[dray]

Thx Roger..

How do you cut the tabs off? By hand?

[thewoodcrafter]

I have an oscillating edge sander I clean the edges with.

[henrik_o]

Danny, you can check out specs of URAI blowers as well as several other brands at www.republicsales.com (http://www.republicsales.com)

Cutting in two passes definetly changes things. With a good industrial system you shouldn't have to, though, unless we're talking small parts, for them two passes is still the best strategy. You might not get away with 600IPM throughcuts in any case, but if you scale back the speed somewhat you'll still have a time saving overall. Only way to know is to experiment, I think.

[fleinbach]

Danny,


I purchased the 7.5 HP blower from Shopbot and so far it has worked great for me. I could not go any larger because I do not have 3 phase power. I cut primarily MDF and my preferred method of cutting is to leave a thin skin. If cutting .75 in material I set my depth of cut to .747 This leaves a very thin skin but sometimes there is places where it cuts through. So far I have not had any problem with movement cutting even small parts from 1/2 to full size sheets. I have even cut circles as small as 2" with a .25" up spiral bit in 3 passes at 5ips.

The thin skin is easy to remove after cutting. If the edge is not critical I simply push the part right out of the sheet. If I need a clean edge I turn the sheet upside down and use the corner of a piece of scrap. Find a place to start and press the corner into the cut. It requires very little pressure to follow the profile all the way around and very light sanding cleans the edge nicely.

Although I have not used tabs since I set up the vacuum system I did use them on everything previously. To remove the tabs I picked up one of the Royobi cordless routers at Home Depot. With a straight bit with bearing you can clean up the tabs very quickly. Being cordless is really nice and you would be surprised how long the battery lasts when you are only using it to clean up tabs.

[dray]

Heya guys thanks for all your replies..

I have not made a decision yet, but I am thinking that unless I can get a vacuum system that will hold down what Im cutting, let me cut all the way through for under $3k for now I think I will continue to brad nail my stuff down with 1 1/4" 18G brad nails.

I have not hit a nail in about a month and the extra time sanding or cutting tabs to me anyways just does not seem worth it.

We have a good production system now, I was just thinking if I could get one guy to be doing something else while the sheet is being cut for 10-15 mins it would be beneficial, but the change over seems like it kind of evens out as far as labor goes or at least not enough to justify buying a low end vacuum table.

You guys are great thx!

[henrik_o]

I did some experimenting. With our ELMO, about 400IPM seems to be the preliminary limit for one-passing 19mm MDF parts that are 500x500mm.

Actually, the parts did not shift, but above that edge quality was unacceptable. Danny, are you getting clean edges with 600IPM two-passing, and if so what is your setup? (Bit, rpm, depth of passes etc.) How do you stop the cut out parts from shifting as they are cut out, brads on them as well?

Part shifting began at approximately 300x300mm.

Now, the ELMO is a good unit but severly underpowered when compared to for example a URAI 45.

*IF* it would be possible to get good quality edges doing one-passes at 600IPM, then I do not doubt such a system could provide solid holddown even for fairly small parts.

[bleeth]

If anyone does project work in which you have the chance of bits and blades coming into contact with nails you might consider these:

www.raptornails.com (http://www.raptornails.com)

[dray]

yes Henrik I also tack the pieces that I am cutting in 2 corners. while they are being cut.

But keep in mind I cut mdf cabinetry it all gets sanded and painted

8241

[henrik_o]

Danny,

Could you post pictures of the edges you get at 600IPM and also what bits and rpm (on which spindle) plus depths of each pass?

I have done some more tests and some calculations, and I believe for our setup with an URAI 45 we should be able to get acceptable edges at 500IPM one-pass. However, this relies on no crossovers into already routed channels, since there is considerable distortion at that speed when crossing over. Realistically, we are probably looking at 400IPM. At that level of force exerted on the blank, the URAI 45 should be more than enough and the major limitation will be size of pieces. I think about 200x200mm should be fine, below that it's harder to say.

Do note that this is on a machine with a 14" Z axis. The chatter/distortions in combined moves and crossovers are likely much more pronounced for us than for 8" machines.

With the right bit, spindle and proportions of speeds (move/rpm), I believe 500IPM one-passing with acceptable edges and relatively small parts is quite possible with an URAI 45.

So, yeah, tentatively, you could invest about $2000-2500 and get a system that would reduce labour time and also cut your machining time to almost half of how you're cutting now. It would need more research, but I believe that's the general ballpark.

[beacon14]

Danny,

Did we make a leap when we started talking about having to cut 3/4" MDF in one pass? Your initial post said you are using two passes now and it sounded like you just want to not have to go any slower with the vacuum than you are going now. If you change from two equal depth passes to the skin method it's still two passes and I think you will still be able to do both passes at 10IPS. I'm willing to try it with my 2 Feins but not until I install my spindle and new motors which I hope will be in the next week or so.

There is another factor not mentioned yet. Cutting a full sheet of 10" x 10" parts is different than cutting a sheet full of cabinet parts with a few small parts here and there. When cutting sheets like this I make sure the small parts get cut first while they are still attached to the rest of the sheet, then the large parts can take care of themselves. This combined with the skin method works for me down to 4" x 12" or so (I'm only cutting at 6IPS now with my Porter-Cable but like I said I'll give it a go at 10IPS with the spindle). I also leave a skeleton around the small parts but cut large parts as close together as I can - check out my last column on cutting strategies on the Shop Website - it's all detailed there. The other advantage is it eliminates the problem of "crossover" that Henrik mentions since on the last pass it's the same amount of material being removed everywhere.

Anyway the labor savings not having to use the brad nails will add up quickly since you are doing a lot of sheets. You could start with a couple of Feins for $700 and see how it works. Then if you need something larger you can use the Feins around the shop - they're great for that also.

[henrik_o]

Dave, Ive been trying to go over this in my head and doing some light experimentation.

The way I see it, if one takes the time and money to set up a competent holddown system, one should be targeting two basic scenarios.

First, one pass, I'll butcher English and call it onepassing. This should be our overall goal for any scenario, since it radically reduces maching time per sheet. If we have a sheet with typical cabinet parts, rafix and shelf holes, let's assume we have 5 minutes boring per sheet. Let's further assume that we are onepassing at 500IPM, which takes 4 minutes. If we did two passes at 600IPM, we'd gain a little per pass, say time per pass 3.5 minutes (we don't realize the whole speed gain due to jogs), but we have to do two so it's 7 minutes. The time per panel is 9 minutes for onepassing and 12 for twopassing. Add 3 minutes evacuation cycle per sheet, 12 minutes against 15 minutes. That doesn't sound like much of a speed gain. However, if we're machining for six straight hours, the difference is six panels (30 v. 24), or one per hour. If we have 2.5 cabinets per panel, the difference in cabinets is 15 cabinets. That's a small kitchen right there.

Now, most of you are going to say, ok, but is that extra speed really worth it? For those of us who don't just chew out parts all day long but rather have as our business model the whole envelope from design to final install, this time saving is going to be trivial spread out over a whole job. This is true. But I think there's a catch here: as long as the machine is working, one man is pretty much tied to it. You can do some work in between evacuation cycles, but it's not going to be much. As such, in my preliminary thinking (out loud) the primary reason for increasing cycle times per panel is not mainly related to time savings over an entire job, but rather to minimize the time of a dedicated operator during the machining stage. Yes, one can still argue that the savings are mundane, but one also has to relate that to the cost of installing a system allowing for this kind of operation. If it comes down to plunking out a couple of grand, one day for install, that ain't a lot, either. Instead of asking 'why', why not?

(All of the above assumes the machining process is not a bottleneck, if it is, then that changes the whole calculation. If there's ever hands waiting for something to come off the machine, recouping the investment is going to be one fast transition.)

The second scenario concerns twopassing. This is the target for runs where edge quality is highly important, for example if edges need to be sanded by hand rather than just edgebanded. To save on labour for sanding (or etc) we need a really good edge. In this scenario, we'd likely want to do a very deep and fast first pass, and then for our second pass we will cut through and also do a very slight lateral change in toolpaths cutting to exact dimension and with a very clean edge. So, the first pass in 19mm MDF would be run at say 18mm depth and lateral final dimension+1mm. The second pass would be at 19mm+clearance depth and lateral final dimension. Due to the small volume of material to be machined, we should be able to do a very fast second pass, and since the first pass is not final in any dimension, it too could be very fast as long as chatter/deviations from the first pass do not dent the material more than 1mm inwards. The latter observation goes to my nest post so I'll leave it at that.

In parting, yeah, this is pretty much just thinking out loud as a newbie. I have a lot to learn and will in time (hopefully) realize what matters more and what matters less.

[henrik_o]

With all that blathering out of the way, on to concrete experiments.

I've done some light testing, and snapped some pics.

The arrows in the following images indicate direction of cut. The images have been manipulated to enhance chatter/deviations. Some cuts did not go right through and I used an upcut bit so edge quality on top was not super, but most of the debris you see on top is chips from the cutting. Also, please note that this was done on a PRS with a 14" Z axis: it would be very interesting to see results on an 8" PRS (yes, we are thinking about making a raised table that could be evacuated for cuts requiring 14".)

First;


8242

This is the result of the specs as per the text in the image when climb cutting through straight material. We see well defined tool marks and some very slight chatter, but this edge is overall perfectly suitable for edgebanding (caveat: it did not go all the way through and the upper edge is not 100% perfect). The manipulation of the image makes it seem much worse than it is: it feels almost smooth to the touch. I can cut like this with the system I have today as long as the edge will be post-processed in another manner than manual sanding. With a better bit and a higher rpm, it should look even better. (Makes mental note to self)


8243

Here, the toolpath intersects a pre-drilled hole. This is important for us Rafix (etc) guys. As we can see, before the bit enters the zone of the pre-drilled hole, it compares to the edge in the image above. However, as it exits the hole, there is a very clear series of lateral oscillations, most pronounced at the beginning and then reduced but at a tighter interval until it steadies up some 60mm from the point of impact. These marks are clearly defined by the touch and all extend outwards from the part being cut. Depending on tolerances, this would probably not require sanding before edgebanding.

Then, mayhem;


8244

This is what I meant by 'crossover cut' in a post above. This image shows the 6.35mm toolpath intersecting a series of 6.35mm toolpaths cut 50% through. As you can see, there is a very violent lateral oscillation at the point of impact, with steadying oscillations after POI as per the image before. All deviations extend outward from the part being cut. The highest point of the waveform is a full 1.5mm. This part would require machine post-processing of the edge. Also very interesting is that as you can see in the image, before the first impact we have a full throughcut, but at the end of the series we do not. The vertical shift is very slight, less than 0.5mm, but it was enough to make a throughcut an onion skinning, needing postprocessing, and as you can see there is no oscillation so this shift is definitive and would afaict affect the extent of the toolpath beyond it.

Ok, this is what I have so far. Comments are very much welcomed. I will try to do this set at 24k RPM and see what difference it makes. I am very pleased that so far all deviations have been extending outwards of the part being cut.

[dray]

I will get some info later tonight or tomorrow when Im at the shop. For now the bite is on so I'm going fishing!



Friends help you move.
Real friends help you move bodies.

[henrik_o]

... and the supa' friend knows the phone number to Winston Wolfe, right?

Happy fishing, we'll get to drill further into this in due time.

[harold_weber]

Henrik, interesting photos of what happens at your "crossover" points. I wonder if this behavior has something to do with the "lateral stiffness" of the gantry? You might get different behavior depending on whether you are moving in the X or the Y directions if the "lateral stiffnesses" are different in these two directions.

[bcammack]

Looks like bit deflection to me.

[Brady Watson]

I second the bit deflection. You might want to try lightening the up the chipload by either slowing down the MS, increasing RPM, or preferably using a larger diameter cutter (like a 10mm) to increase stiffness. When you are running full-tilt @ 8+ IPS, you really have to make sure that everything is up to the task. If you are stuck with a 6.35mm bit for production efficiency, it may help to get a tool with more flutes to reduce chipload & bit deflection. This is debatable since the cross-section of a 2 flute tool is meatier than a 4-fluter...but the 4-fluter has 2 more cutting edges.

-B

[henrik_o]

We're not set on a 6.35; I used it because the 2-flute 8 mm downspirals we ordered have not yet arrived. I would expect results to be slightly better using this bit and a higher RPM: it would be even better with a 10mm since what causes this oscillation is the crossover into zero material spaces and it gets worse the larger these spaces are proportionally to the cutting geometry-- a 10mm bit would display much less oscillation when crossing a 5mm space, etc.

However, it has become quite clear that the real culprit here is the 14" Z. Re-running these tests we can very clearly see the oscillation as a pendular movement along the x axis in the whole gantry. For now, we will make do with twopassing at drastically reduced speed which we've tested and works very well quality-wise. It's not a problem since Embla (my PRS) is not into full production yet.

I expected this, there's just no way the 14" gantry could be as rigid as an 8" version without a different construction.

The current plan is to continue as is with modified speeds, and as we ramp Embla into full production towards the winter we will build a raised table, design target being changeover from raised to normal in ~1 hour.

If anyone has done this or if you have ideas or advice, I'm all ears. Preliminary thinking is to use tubular steel or aluminium extrusions as spacers beneath the table in the raised position, and using a hoist to hold the table as changeover is performed. Maybe raise the table a bit extra, say a 6" Z in the raised position, thus gaining a little added rigidity.

What do you think?

[harold_weber]

Right, I would not expect to see the oscillation or "ringing" if the problem was only bit deflection.

You say you see "pendular movement along the X-axis in the whole gantry", do you see this behavior while cutting in the X direction, or just when cutting in the Y direction? If it is only in the Y direction, you may be seeing a case where the torsional stiffness of the gantry is too low, especially when you are near the bottom of a 14 inch Z travel. If this is the case, then yes, I believe the higher you can have the Z position of the cutter, the more reduction in the pendular movement you should observe.

You can verify this by simply building up a 6 inch tall stiff box, attach it to your current table, then attach a piece of your MDF to the top of it, and try the test cuts in this piece.

[henrik_o]

Yes, it is more or less confined to cuts along the Y -- the rigidity of the gantry can not cope with rapid variations in sideforces when cutting along the Y at high speeds resulting in oscillation across the Y / along the X. This loss of rigidity is easily testable simply by exerting manual force along the X axis when the gantry is at a standstill but the motors engaged: it takes only moderate force to move it +/- 2mm, after that rigidity kicks in.

Exerting manual force along the Y axis produces very little movement even at high load, but I said 'more or less' and yes, when running into resonance during a cut along the X axis, it does result in Y oscillation. This would afaict happen with any Z height, but the results are certainly more clearly visible with a 14" Z.

Good thinking about the raised box, I had an idea to test it with a torsion construction, but finally I scrapped that since the problem here is so easy to detect simply by exerting force along the X with the Z at -10"; -8"; -6" etc. I personally think about 6" is the sweet spot, in order to produce clear motion under heavy force at that height you are starting to move not just the gantry but the chariot as well: such loads are unlikely to occur even at high speeds unless you're doing something really wrong and are, afaict, about to detrack.

[henrik_o]

OK, upon rereading the above post I realize there's an awful lot of "along" and "across" this and that and it's easy to get lost in it unless you by experience and/or intuition know the problems under debate here; I'll try to post some pics of the axises/oscillations/forces at work some time tomorrow.

In a sense, I think I'm 'blessed' (so to speak) to have a 14" Z since it means deviations whose causes may not be entirely clear on 8" systems are VERY clear on mine, but even with that reduction in Z I can say right away that you are affected by it, even if it may not always be really clear. For those that haven't looked into this much, maybe we can get some more discussion and solutions going right here in this thread?

[wardsa]

Danny,
How many nails do you put in your panels? Are you using PW? I'm going to use VCP for toolpaths. Do you mark or index where to nails before you start cutting?
Thanks,
Terry

[dray]

Danny,
How many nails do you put in your panels? Are you using PW? I'm going to use VCP for toolpaths. Do you mark or index where to nails before you start cutting?
Thanks,
Terry

I just use 1 in each corner of the sheet. Then as its cutting I tack 2 corners of pieces being cut. It takes about 10 mins to cut a sheet, so I have one guy assembling/sanding and one guy loading sheets, tacking, removing, stacking.

Then when they are cought up on cutting they both assemble while I make custom parts.

It usually works out pretty good and everyone stays busy. I was thinking of vacuum table for a little more time assembling, but I think at 10 mins +/- to cut a sheet it may be more chaotic, having a guy assembling for 5 mins then back to loading, removing etc etc.

I use 1 1/2" 18g

Yes I use PW

[henrik_o]

Danny, you need a vacuum system, no ifs and buts


Seriously though, if your system works for you that's fine and all, I just can't see how you do that. Consider the following pic;


8245

This is a 1530x3050 sheet of oak veneered blockboard machined for cabinet tops and bottoms using Rafix SE32 joinery hardware (the Rafixes will be invisible in the installed piece). Time to cut this whole sheet was almost an hour, but with a larger bit and a raised table we should have it down to half of that.

The way it worked is I put the panel on the table myself using our Häfele panel cart, engaged the vacuum and DC and started the file, all of which took about two minutes. I then left the 'bot and started staining pieces already cut. After about an hour (which should be half an hour soon) I heard the 'bot was not cutting anymore, went out, evacuated the sheet, swept the table, loaded another sheet, vacuum, DC, run file, off with me. Total work time per sheet, say five minutes.

If I had needed to be on hand with the nailgun, not only would I have needed to stay right there all the time, costing me 1 (or 1/2) hour of labour, I would consider that an accident waiting to happen. Also, to add insult to injury, evacuating the table surely must take some time if you have to pull out nails, and then there's the issue of the holes these nails make. At least some of these holes would have to be puttied, no? That adds time to do that, maybe waiting time for it to set, then it needs to be sanded. Even if you don't do this now, this being acceptable to your customers, it would seem to effectively prevent you from taking on high end jobs with veneered panels.

What am I missing? How do you build your cabinets? I don't want to come off as arrogant, it is quite possible that it is I whom is just not seeing a better model.

In parting, here's a pic of a stained sidepiece in one of the drying racks,


8246

[dray]

Hi Henrik, I build 99% all MDf. I rarely do any veneers, MDF cuts fast and allows me more freedom to build things as I see them in my minds eye. If I want to round over an edge I do it. If I want to carve fluting I do it and I do not change materials.

It only takes me 10 minutes to 15 minutes to cut out a sheet not an hour or more.

This His/Her desk unit was built in a day, the next day We cut, installed doors drawers added softclose etc in about 4 hours.

My painters then did a Shading laquer (3 coats)to get the color depth then 4 coats of finish and we installed it in a day.

I do not do any staining/finishing myself.
It is hard enough to organise jobs, design, be a good carpenter and give 100% to my customers. If I get diluted by another trade I cannot give my customers 100%


Day 1

8247

Day 2

8248

Install Day

8249

That last picture was poor from an employees phone, but the cabinet has the same finish as this one


8250

8251

[harryball]

This is a recent update I did. The base is 3/4" ply followed by 1" MDF with a 4 zone grid cut in that has 1" wide exterior boarders on the zones, all very well sealed.

The top sheet is 3/4" Ultralight Trupan and this time I did a full glue down. I also added squared registration slots that I insert 1/4" pieces of hardboard that stick up 1/4" (I got this idea from the forum). My plan is when I surface to the point the slots can't be used I'll surface down leaving about 1/4" of the original board then lay down glue around each zone and several spots inside and simply apply another sheet of Trupan on top.

So far, this has worked very well, just toss a sheet up and go. The 0,0 mark is at .75,.75 from the 0,0 of the top itself. When I want to run using the registration blocks I zero the machine, move to .75,.75 and zero X,Y and I'm good good to go. The registered area is 48 1/4" x 96 1/4". I've found leaving a 1/4" lip works well, I can put a small strip of round foam weather stripping (or caulk saver cord) if I find a sheet that needs it to help suck it down. So far I'm very pleased with the performance.

And some photos...


8252

8253

8254

8255

Robert

[beacon14]

Do you use a Custom Cut to zero then re-zero at .75, .75? If not try it. That type of always-used short routine is perfect for a custom cut - you can do what you described with two keystrokes.

Nice job on the table. Simple and effective.

[dray]

Ok, now that everyone has convinced me to go with a vac system, how exactly do they work?

Let me make some illistrations of my basic understanding so that I can be corrected

8256

[dray]

Sorry, I am a visual kind of person but is the above diagram fairly accurate?

[myxpykalix]

That pretty much is the way i did my table. I drilled holes in my plywood plenum board and went and bought several toilet lid bolts and nuts (the nylon kind) screwed the plywood down, cut my plenum in it, skinned top and bottom of mdf then glued it down. Then surfaced the top.

3 feins might be overkill if you ask me, but others more knowledgable might know more.

Make sure to use plenty of silicone around the connections to the table of the pvc pipe.

[fleinbach]

Danny,

Your illistration looks pretty good but you don't need to screw or nail the bleader sheet down. I ran beads of latex caluk around the parimeter and on every high point of the grid. Then after placing the bleeder board down I put another 1" sheet of MDF on top and and turned on the vacuum and left it on for 24 hours. This worked very well for me.

As for my opinion of using multiple Feins I will say I purchased and experimented with 2 Feins before I broke down and purchased the 7.5 blower from Shopbot. What I learned when using the 2 Feins is than unless you set them up to independent zones you get absolutely no benefit whatsoever. I made many tests when connected to the same line and my results showed that the increase in holding power could barely be detected on my vacuum gauge. The same was true for air flow. The 2 vac hook up recorded the same drop in vacuum after cutting the same parts. So to me the only way to benefit from multiple Feins would be to put them on separate zones that way when the first zone is cut through and you lose holding power another zone is still holding the sheet solidly in place. You will still need to tab small parts or leave a thin skin to keep them from popping loose.


I personally like the security of the extra holding power of the 7.5hp blower. I do not tab anything anymore. I just leave a very thin skin to hold the parts from shifting. and have cut parts as small as 2" in diameter with an uncut spiral with no problem. Removing the thin skin is very easy.

[harryball]

You may want to put your 4 zones arranged in the four corners...
________
|___|___|
|___|___|

If that makes sense to you.

Then you can work with 1/4" sheets at the 0,0 point.

As for one or two feins, I use 2. I'm not sure what Franks doing but I did notice a difference between a single and twin setup. On the entire table I ran a single fein to cut a full sheet and my parts moved around. I agree that the vac gauge said about the same thing but when running the twins on the full sheet the parts don't shift on me. I don't use tabs or a skin except on small parts.

Robert

[Brady Watson]

1 Fein = about 7" Hg & 125 CFM
2 Feins = about 7" Hg & 250 CFM
4 Feins = about 7" Hg & 500 CFM

The additional Feins will give you no more holding power, but they will give you more CFM - allowing the system to soak up leaks and kerf loss.

-B

[brian_h]

Frank & Brady & Robert,

I'm really close to starting my vacuum table. I want to double-check some details...

You started with a 'base' sheet of plywood that lays on top of the metal frame.

Next you glued a 'plenum' board on top of the base board and surfaced and cut the zones into the plenum.

Then you sealed the plenum. (Frank: Like you I have epoxy paint from 2003 that I'm thinking about using. It was going to be my shop floor until I decided to tile it instead. Hopefully it's still good. It was a lot of money for nothing.)

Next you glued a piece of Trupan on top of the plenum, using silicone (Frank) on all of the high spots, both around the edges of the zones and on top of all the little squares cut in the plenum. Then seal the edges of the Trupan.

Now you can surface the Trupan and it's ready. Danny's illustration has the bottom of the bleeder board surfaced, presumably to break the 'seal' of the MDF, but is that necessary with Trupan?

As far as terminology, the top sheet of trupan is both the bleeder board and the spoil board, right??? You continue to resurface the trupan as needed until it's time to glue a new spoilboard (trupan) on top.

It seems to me you'd have to be careful not to put too much glue between the old and new spoilboards or you'd risk hampering the airflow through the trupan.

I'm going to be using Feins, probably two to start but I'll design the plumbing to accept one or two more later. Each Fein will have it's own zone(s).

Hybrid Table: I've decided to use a baltic birch jig that is clamped on top of the spoilboard for my T-slots and cams. From what I've learned with my temporary T-slot surface, it should work well. I'll share pictures when it's done.

Sooo.... Do I finally have it right???

[richards]

That's just about the way that I built my setup (before I switched over to Brady's combined plenum/spoil board, which works better for me). However, I have never glued the spoil/bleeder board to the plenum board. Instead, I lay some All*Star tape around each zone (1/16th inch thick, 3/8 or 1/2 inch wide). The tape gives a really good seal between the two sheets with none of the hassel of glue or silicone. Also, I surface both the top and bottom of the spoil/bleeder board; however, there is no trupan to be had in Salt Lake, so I use ultra-light MDF. Trupan may have different air-flow characteristics than ultra-light MDF.

[harryball]

Sounds about right, some things to consider...

You do need to surface both sides of the Trupan, as I surfaced the my Trupan the vac was running and the pressure went from 1.5" to 1" meaning more airflow and less vac just from taking off the skin.

I really like having the ability to apply both vacs to a single zone. Many of my projects fit well on 1/4" sheet and having both vacs pulling on a single zone really gives me a lot of airflow to deal with leaks from cutting through. I've also make "flow through masks" to fit a single zone. I can drop the mask over the zone and both vacs are pulling through the openings, then placing the material on top and the hold down is remarkable. I can cut the parts free without tabs. I tried this trick with a single vac and had difficulting getting it to hold.

When I glue down my next sheet of Trupan I plan to lay down the glue over the zone grid and outer edge, then in the middle of each zone I figure 6 or 8 globs that should press out to about 1.5" circles spaced regularly should do the job.

Robert

[henrik_o]

Brady,

quote:
1 Fein = about 7" Hg & 125 CFM
2 Feins = about 7" Hg & 250 CFM
4 Feins = about 7" Hg & 500 CFM

That doesn't sound right: it would mean three 1.2kW shopvacs could pull more air than a 15hp roots blower.

Isn't the Fein rating for unrestricted airflow with zero compression, or somesuch?

Also, does it really scale proportionally like that in a vacuum system?

On another note, is it really a good idea to use anything containing silicone for an application like this (sealing edges, etc)? If you cut through an area where this sealant has been applied, wouldn't you effectively spray a fine mist of silicone over your material (well, what the DC doesn't catch)?

[thewoodcrafter]

Henrik,
That is at 1/2 the holding force no matter how much air is moving.
My blower is rated at 14".
That is twice as strong.

[Brady Watson]

Zactly.

CFM is always inversely proportional to Hg"/vacuum suction. At full vacuum, there is zero CFM - or at least this is what you should aspire to achieve with a vac system. Otherwise you've got a leak!


-B

[henrik_o]

Ok, that makes sense, but those flow ratings for the Feins still sound very high. At 500CFM, wouldn't it be almost impossible to lose pressure, a lower pressure than on an industrial system, admittedly, but still?

Or 250 CFM with 2 Feins: let's say you're cutting a 4x8 panel in half then machining out the respective blanks, and you remove the first when it is done and then cut the second, at that CFM you should have almost the same pressure on the second one as the first.

What am I missing here?

[harryball]

Oh, I used yellow wood glue, worked just fine. I don't like the idea of using stuff that would gum up the bit if I cut that far down.

Each fein is rated at 125 cfm, so four of them have 500 cfm. YES, this is a higher rating than the industrial systems on cfm but not vac.

The example (if I understand it right) of cutting a 4x8 sheet in half and removing one half doesn't work because you'll find that with that much open table you'll have a much lower vac but you'll be moving a lot of air. Higher cfm means you can have more leaks... but not that much. Air flow and vac loss is not linearly proprotional to open area. The higher cfm does mean that you'll maintain more vaccum while cutting kerfs open.

Case in point, I have a full sheet project that I mentioned earlier that I tried with one fein. The parts would shift as I was finishing up the sheet. However, with both feins running all the parts stick and stay. The gauge shows a little more than 1" of additional vac than when running a single unit. With this observation I've often wondered if a 15" system that only pulled 120 cfm would do as well as a 7" system pulling 250 cfm in that specific case.

Robert

[fleinbach]

Brian


In answer to your concern about your Epoxy. My Epozy paint was on my shelf from 1999 until this past January and I got it when I help a friend clean out his old shed and I have no idea how long it was in there.

I used a latex caulk not silicon. The reason was silicon seems to skin over in a short time and I needed enough time to cover a large area.It took around 15 to 30 minuites to hit every area.

As for 2 Feins on the same zone you could get more CFM of air flow but if you cut through your material with a 1/4" bit you loose pressure very quickly. I have not made any tests with the Feins since January but I recall with one or two Feins I was getting a little over 6 inches before cutting. After cutting through all my parts on a 5' X 8' sheet vacuum was down to 3 inches. The parts where large so they still held ok. My vacuum gauge is graduated in inches and difficult to differenciate fractions of an inch but I was get the same readings with 1 or 2 Feins.

Though it is true that more CFM should allow you to mantain more pressue in the end the remaining pressure is all that matters. In other words say you cut away enough area for a Fien to reach say 4"of mercury. Now on another system you have a blower capable of 15" but you cut away more area and now have ony 4" of pressure. Both systems at this point have equal holding power. Of course in this scenario the larger system allowed more cut area. The same cut area on the Fein would have most likely been reduced even lower and possibky not held.

[brian_h]

Mike: I was originally thinking about doing an upside-down plenum/spoilboard (Brady-vac), but decided to go all-the-way after thinking about the different ways I could use the 'bot.

Robert: Looks like I'll surface both sides. Thanks.

Frank: Yea! I don't have to feel guilty about the unopened, light sage-green epoxy paint! Latex, sounds like a good reason. I do use silicone everyday in two pneumatic calk guns. The guns should come in handy.

Robert's two vacuum, one zone idea has merit. Most of my projects seem to happen right around the x=0, y=0 point. It may be a good idea to have an option of either running the vacs separately or in series, depending on the need.

I also need to research that special valve Brady was experimenting with for incorporating a low-CFM, higher-vacuum connection that over-rides the vacs. I don't know what its called, and I don't know if Menards carries it, but it may be a good thing to consider installing the valve, since I already do a lot of vacuum-bagging type work. It came up at Camp Shopbot and I haven't had time to research it yet.

Thanks for all the info., guys. I'm really looking forward to this project.

[Brady Watson]

Brian,
You don't need to surface both sides of the Trupan as long as it is really Trupan. Just flatten the top & go to town.

I'm post more on the 'special valve' either here or on the Tricks section in the next column.

-B