J. Phil Thien's Projects

Another possibility is to connect a vacuum line to the collection canister, to create a vacuum in the space between the canister and the outer surface if the bag. That should resist having the bag being sucked in.

Grizzly has designed this: http://cdn0.grizzly.com/manuals/g0440_m.pdf so apparently it works.

Okay- it looks as though nobody replied to your questions, so I'll give it a shot. However, I'm a newbie myself, and in the process of building my own separator, so do not consider the following any sort of expert advice, rather a reflection of my understanding of posts on this forum.

First, the orientation of your baffle is correct- the "solid" section (180 degrees without a slot) should be right in front of the intake.

Second, for me at least the answer of the height of the center pipe (outlet) is confusing. My interpretation of the original description by Phil is that it should stick down about half the diameter of the pipe from the top. Other posters have place it half the diameter above the bottom (the baffle). For the standard separator, the thickness of the chamber is about the diameter of the outlet pipe, so it's the same distance either way. Once you get into more elaborate designs, where the chamber is a lot taller, as far as I can tell the distance is best at about half the pipe diameter above the baffle. In the picture of your build I can't see the outlet port, but if it's the same diameter as your intake port, say 3", I'd start at about 1-1/2" above the baffle, but make it adjustable to start with. That way, you can adjust for optimal placement. Higher will give better airflow (suction), but allow greater bypass (more dust will go on to your DC or vac. Lower will improve separation, but if too low airflow will suffer.

Third, things you might do to improve performance:
1) chamfer (bevel) the bottom of your slot. The top of the baffle should be flat, but a sharper edge  will help separation.
2) Anything contributing to airflow turbulence will decrease separation performance. Ideally the airflow coming into the separator should be tangent to the wall, so it has a smooth circular path. I can't tell whether your elbow inlet is pointed outward a bit- if so, the air will tend to bounce off the wall.
3) If you can make the baffle supports thinner, you will get less turbulence. Threaded rod or wooden dowels should be plenty strong to support the baffle; a thinner baffle (say 1/4" Masonite or plywood) would be lighter, and also have a thinner edge at the slot. Also, from designs I've seen I recall that the supports are closer to the wall- yours seem pretty close to the center, which will interfere with smooth airflow into the outlet port.
4) Finally, the basic Thien design seems amazingly robust (forgiving): you're going to get very good dust separation with just about any build that approximates Phil's original design. A lot of the discussion seems to aimed at fine tuning modifications and optimizing, but with diminishing returns.
Just my .02  :)

Thanks for the clarification!  :)

FWIW, my simple minded picture of what's going on is that the work the impeller is doing is spinning blades against air resistance.

If the intake is restricted (or esp. closed off), there is a partial vacuum, so less air, less resistance to the blades, so less work- and lower amps. [In a total vacuum, the only work the impeller would have is due to bearing friction.]

If the outlet is restricted, pressure builds up in the impeller, so the air is denser, so more resistance, harder for the blades to turn, so more work, and more amps.

Okay, WoodCzar, since nobody else has replied yet, I'll jump in. I'm a newbie myself, and had some of the same questions myself, before getting a handle on things. Hopefully, someone with more expertise will add later, but in the meantime, here goes....

First, there is no real "best" design- the design depends on multiple factors- capability (size) of your DC- shop vac or multi-HP; cost (scrap bin inventory); time/ease of construction; available tools; shop space; degree of attention to detail/fussiness/compulsion; etc.

That being said, here are some general "rules" I have gleaned from the collected wisdom:

Overall, top hat designs offer better performance, but are a bit more complicated to build, in particular the separator wall, and the side inlet junction. Making airtight seals between the lid, wall, baffle, support ring and waste container is vital. With "drop-ins" (for lack of a better word), the inlet is through the lid, and the walls of your waste container are the separator walls; you only need a good seal between the lid and container lip.

Separation is best with straight (non-turbulent) air flow, so a side inlet, coming in tangent to the separator wall, is better than from the top, with a right angle bend under the lid. Also, a rectangular inlet, instead of round, reduces turbulence as the incoming flow hits the separator wall. Side inlets are easier to incorporate into a top hat design. If you're going to go to the effort of making a top hat, definitely incorporate a side inlet.

Also, a straight air flow exit port is better for DC performance, esp. if the DC impeller is near the exit port, such as when the impeller is located just above the separator, usually with top hat designs. Check out retired2's postings about air straighteners.

Generally, the larger the separator, the better the separation, esp. with fine dust (the hardest to separate). Unless you want to go with an elaborate funnel shape below the baffle, the maximum diameter of your separator is determined by the diameter of your waste container.

A tall separator (increased height between the lid and baffle) allows more time for air to circulate, and better separation. The inlet is best just located just beneath the lid.

One complicating factor is that a larger separators need more powerful collectors for good airflow, so generally shop vac couple separators are coupled with smaller separators. I don't know of any specific guidelines, though.

A narrow slot width (7/8", of even 3/4") offers better separation of fine particles, but clogs more easily with large chips (planer or jointer, e.g.). The 1-1/8" slot width seems to be the best compromise. Note that slot width doesn't scale in relation to separator size: the slot for a 24" wide chamber is NOT twice as wide as one for a 12" wide one.

Bottom line: As long as you follow Phil's general design, you should get very good separation. Top hats with side, rectangular inlet offer best performance, but are more involved to build. Phil's original design is cheap, fast, and easy to build. Like a lot of things, performance improves with complexity and expense, but with diminishing returns; where on that performance/price curve you sit depends on personal preference.

Just my .02  :)

Okay, first some disclaimers: I'm not any sort of qualified engineer, nor have I had any personal experience with such a monstrous blower, so don't say I didn't warn you!

You might want to check out Cincinnati Fan http://www.cincinnatifan.com/index.htm Their catalog section (and maybe elsewhere) has tables of impeller type, size, cfm, hp, etc., to give you some idea of what routine motor hp requirements would be.

Since your impeller is such overkill for the needs of most any woodworker (except maybe Bill Pentz  ;) ), one thing to consider is to run at a lower rpm by using a small pulley on your motor, with a v-belt driving a larger pulley on the impeller shaft.

According to fan laws (who knew?) ftp://www.nyb.com/Letters/EL-03.pdf cfm is directly related to rpm. Static pressure is proportional to rpm squared, and hp is proportional to rpm cubed.

So, if I am not mistaken, if you reduce impeller rpm by half, cfm would be one half, static pressure would be one fourth, and hp requirement would be one eighth. It seems that you ought to be able to use your 5 hp motor, and still get acceptable performance for your needs.

What sort of container is the "final resting place" for your chips? If it can be made reasonably airtight, why not make it into a simple chip collector (inlet port in lid, with hose to suck contents of Thien separator, and outlet port to shop-vac?

If you're like me, I dispose of chips with my residential trash in wheeled Toter bins (in yard waste bin if all chips are "clean"). Not airtight, and local requirements are that trash must be bagged. If so with you, maybe something like this would work to empty your separator into: http://lumberjocks.com/wdkits1/blog/8004  Just a thought

retired2: You're right- I just needed to read further into your thread for the CFM data.

Thanks, John

Thanks, Phil, for the explanation and clarification. If I understand yours and retired2's comments, a tall separator with a bellmouth outlet placed low (close to the baffle) should decrease bypass, while maintaining (or perhaps improving!) air flow.

And retired2,

QuoteSorry to the OP for high-jacking your thread.

absolutely no apology is warranted. The information  you provide is, as Mastercard says, priceless.

I went back to your rectangular inlet/bellmouth thread, and dug down to your table showing amperage (and dB) relative to various "loads" and outlet position. Have you collected any further data regarding outlet position with a bellmouth, to determine whether there is a maximum air flow position? That is, there must be a point, as you further lower the bellmouth, where the gap between it and the baffle is so small that it impedes air flow. Also, I assume that you are relating air flow to amperage, or have you done any direct measurements of air flow. If so, how do you measure it (and do you know of any DIY (i.e. cheap) methods of direct air flow measurement?

Thanks to both of you, and others, for sharing your wisdom, and the patience and willingness to share it with newbies such as myself.

Okay, I'm circling in, so to speak) on my separator design, and am thinking of a tall separator (like the one Bulldog8 has built), with a rectangular inlet (like retired2's).

My question is whether the vertical angle of the inlet has any bearing on separation efficiency. It's my understanding of the postings here that the more rotations of air in the separation chamber, the better collection of fine particles; a tall separator allows more rotations. In that case, it makes sense to me that the best placement of the inlet port is right under the top lid of the separator, and parallel to the lid (horizontal). That way, the air flow, as I understand it, takes a shallow spiral path.

On the other hand, Bill Pentz makes a big deal about having an inlet to his cyclone plan that is angled downward, with the downward flow reinforced by an internal air ramp. The air flow takes a much steeper spiral flow.

I realize that a cyclone with a cone is not directly comparable to a Thien separator, but does angling airflow downward, as Pentz recommends, make any sense in a Thein baffle separator? My sense is that while it might improve air flow through the separator, separation efficiency, esp. fine particles, would suffer, so I'm leaning toward going with a horizontal inlet. Any thoughts?

Caveat: I do not have any formal training in air flow technology, nor am I an engineer of any sort, so as they say, "my advice is free, and worth every penny."

That being said, by putting the entire collection system upstream of the impeller, you're basically turning it into a vacuum pump. With the conventional location of the impeller, it works two ways: 1) as a vacuum pump, drawing air (and dust, etc.) through the inlet, and 2) as a blower, forcing air through a filter bag/canister, or venting outside.

My gut feeling- again, not a professional!  :P  is that dust collector impellers work more efficiently, i.e. achieve more air flow, when they act more as blowers (pushing air against a resistance- such as a filter) than when acting purely as vacuum pumps, "sucking" not only through the ducts, separator, etc., but also through the filter.

Somebody with an air flow meter should be able to tell pretty easily  in a system, identical in every way (ducts, separator, filter, impeller),  except impeller location (traditional, or at far end), which has better flow.

Okay, I'll give it a shot, but I must emphasize that I am not in any way an expert- don't say I didn't warn you!  :)

As far as having a rough (sandpaper,etc.) wall, instead of smooth at the bottom of the cylinder, some small particles might be caught there, but I think that after a short time that area would be caked over with dust. Since the dust isn't physically attached (stuck) to the wall, it would tend to be pulled off the wall by the airstream. My guess is that small irregularities (sandpaper particles) would also cause microscopic turbulence and disruption of laminar air flow, which could actually be detrimental, by "kicking up the dust" that's on the wall- I picture miniature dust devils being created.

As I see it, the whole idea is to get dust out to the wall of the separator, where it can eventually fall by gravity through the slot, while the circulating air eventually goes to the center of the separation chamber, and exits. The baffle separates the rapidly moving air above from a relatively stagnant (less air flow) area below. If I understand correctly, your idea is of having side ports in the lower part of the wall, so dust particles can escape out the side, instead of dropping out the bottom. The problem I see is that air would go through those slots as well, and end up causing a lot of turbulence in the waste area below the baffle. This would stir up dust already collected, and since the air has to go someplace, it would go up through the baffle and exit, carrying dust with it. So, I don't think it would work.

Of course, there's a pretty good chance that I'm completely wrong! But this is what I like about this forum- lots of interesting ideas!

I had a similar question in a recent thread, which was addressed by retired2; see this thread: http://www.jpthien.com/smf/index.php?topic=1046.msg5639#msg5639.

Like your plan, I'm planning using a ring beneath the baffle, with a goove to seal the waste can. I thought of extending a "tongue" of the ring beneath the baffle- like yours, but with a essentially a really wide slot. Anyway, check retired2's comments. I've followed his threads, and he sounds pretty smart. Nonetheless, I might try my idea regardless. Something about fools rushing in, I guess.

Good luck with your top hat. I just need more time to get mine underway.

First, apologies to nj_mike, if I've hijacked the link- the questions you had were the same as mine. Second, thanks to all for replies and info. Thanks esp. to retired2. I'm planning on wholesale borrowing of your build.

As of now, I'm planning on using Formica for the cylinder wall (it's actually a scrap of Wilsonart I picked up), without any substrate. I figure that I'll add more vertical ribs (?8) between the upper and lower rings, to provide more support. Construction adhesive is a great idea- the though of using contact cement gave me the willies in this application. Bending the laminate to ~20 inch diameter shouldn't be a problem, since it's already wrapped in a 14" roll.

My collector is an older model (blue paint) Jet DC 1200. Specs sheet says static pressure is 11.44 inches of water. FWIW, that translates to about 0.4 psi. I'm hoping that with bracing the Wilsonart will be strong enough.

My earlier question about dust particles eating away at the wall was based on some comments that Bill Pentz made on his site, saying that wood, esp. species high in silica, effectively sandblasted commercial cyclones, causing rust through. He suggested adding a rubberized coating at the impact area. In hindsight, for my application it seems like overkill.

One final (for the moment, anyway) question/thought about retired2's design of a "sandwich" baffle. You have a picture showing a chamfered mdf disc attached to the under side of your baffle, I assume to improve rigidity. What do you think of cutting the bottom ring (between the baffle and the waste container), so that a "tongue" projects into the center, below the baffle, in the sector where there's no slot. That is, instead of an mdf "island" beneath the baffle, have a "peninsula". I figure it would give better support, and might eliminate the need for a baffle support in the separation chamber. Just a thought... see attachment of a very crude drawing.

I'm yet another neo-builder with the dilemma of what material to use for the separator wall, so let me jump into this thread, too. I'm planning on ~20" diameter, to fit on top of a Brute can, based in large part on retired2's design.

I have scrounged both a good length of 1/8" hardboard, and one of Formica. It seems that the hardboard would be easier to fasten, but does anyone have experience with hardboard durability: would incoming chips scrubbing the wall erode the surface.

On the other hand, I figure Formica would be more durable than hardboard, but I'm not sure of how best to use. My only experience with Formica/laminates has been the traditional countertop installs, cemented to a solid substrate. Can Formica "stand alone", that is, can it work as a separator wall, supported only by mdf rings at the edges, with several vertical ribs (a la retired2) spaced between the rings. Or, does Formica need a continuous substrate for adequate strength?

Probably the strongest install would be a hardboard outer ring (substrate) with a Formica lining, but thinking of bonding Formica to the inside of a cylinder with contact cement gives me a headache- how to get a good bond with matching seams, without disaster. Does anyone have experience with partially supported Formica, or with more forgiving adhesives? I can't find any info on Formica walled separators, either in this forum, or a general Google search, so any advice would be greatly appreciated. This is a great forum :D