J. Phil Thien's Projects

Quote from: Ken45140 on November 29, 2007, 06:44:43 AM
Thanks much for the new picture.  Very clear and the profile of the elbow is helpful.  The detail is so good I just have to ask: what is the wingnut (and its bolt) doing?  I can see nothing in the side photo on your main web page.

;D

The wingnut (and machine screw it is connected to) cover the 1/4" pivot hole my circle making jig left behind.  I don't want to leave any open holes directly under the vortex finder (output) tube.  It is unlikely that a 1/4" hole would pass much dust, but If you examine the picture you'll see there is plenty of sawdust on the bottom of the baffle in that spot.  Without it, some of that dust may sneak back to the shop vac.

That would work fine.  Someone else suggested all-thread or carraige bolts through thinner PVC pipe.  Whatever gets the job done w/o being too much wider would be just fine.

Someone E-Mailed today and asked what the baffle itself was made of.  It is 1/8" tempered hardboard.  This is smooth both sides, and is quite slick.

Due to the nature of the material (dark color, produces lots of fines when machined) hardboard also works great for testing separation efficiency.   ;D

Quote from: siliconbauhaus on November 26, 2007, 07:52:53 PM
This looks good so far and I'm interested in it.

At the moment I run a shop vac with a separator lid on top of a 5 gal bucket. It's not bad but I'm sure it could be better.

I'm always interested in hearing about other user's levels of success w/ commercial tops.  Do you mind if I ask a few questions?

(1) What brand/size shop vac do you use?
(2) What type of top do you use?
   (a) http://www.rockler.com/product.cfm?page=1515&filter=separator
   (b) http://www.woodcraft.com/family.aspx?familyid=4310
(3) How much (if any) debris makes it back to the shop vac?  Is it just fines or is coarse debris, too?
(4) Does the level to which you fill the bucket before emptying affect the debris making it back to the shop vac?

This is a photograph shot from the bottom.  I've cut an index card (to scale) showing the profile of the PVC elbow, and placed it on the bottom of the baffle so you can see its position relative to the baffle and the side of the can.

There is a small corner cut out of the elbow (reflected on my mock-up) to allow additional clearance between the side of the can and the elbow.

I'm working on a SketchUp model which will hopefully help more.

Got the following E-Mail form a guy named Rich:

*****
In a magazine test some time ago they used chalk to test filter efficiency.  You could do the same thing with your cyclone separator.  Add a couple tablespoons of chalk to your intake and run the vac for a minute or two.  Then see how much dust stays in the cyclone can and how much gets to the vac.  Don't use too much, it can clog filters badly.  -Rich

*****

I thought it was a neat idea so I took some bright pink chalk and sanded it with 150 sandpaper.  I created approx. 2 tablespoons of powder and then poured it into the cyclone's intake.

I've included a few pics below w/ results.

After performing the test and removing the hoses, I noticed a little chalk in the exhaust port.  I figured most of it had snuck through.  But upon further examination I was pleasantly surprised to see that most of it stayed in the cyclone (garbage) can.  The walls of the can, below the baffle, had a thin pink coating.  And the debris that was already in the bottom of the can had a thicker pink coating.

I don't know that this test tells me much, but I feel better with the chalk in the can than I would if the chalk ended-up in the vac. :D

Q: I read about the idea of expanding your design to 4" hose sizes, implying use with a fullsize dust collector.  I read about the problems you were trying to solve - but all of these were manifested with shop vac -2.5" size implementations. I do not think the carrryover problem you solved exists with larger scale dust collector solutions.  The traditional trashcan solution for a fullsize dust collector uses 4" hoses and a 30 gallon trash can, which is considerably taller than the 10 gallon can you used.  I have used a 30 gallon trashcan with a homemade lid and a 2 hp Jet dust collector for years. I never experienced the carryover problem you mentioned.  My lid looked just like yours, except it has no baffle.  I suspect it works as well as it does because its so much deeper and a larger diameter.

A:  Thanks Bob.  Interesting observations.  I have personally witnessed horrific scrubbing in shops I've visited, even with the 30-gallon cans.  Many people I know only use the cans with their planers because they can't keep anything smaller than very large chips in their cans.  But I will do testing w/ and w/o the baffle when I finish building my larger model.

*****
Q:  That looks great for a shopvac to keep the drum from filling up as fast. That sort of design typically doesn't scale well to larger collectors if the goal is to separate out the fine dust. Once you start moving more air the physical size required makes this design impractical. This is the reason for the popularity of the cyclone--done right it can separate out a large percentage of even the very fine particles.

A:  My tests indicate that my design scales infinitely better than designs w/ no baffle. I'm trying to apply the 90-10 rule. That is, I'm trying to provide 90% of the performance of a true cyclone for 10% (actually, a lot less) of the cost and complexity (and size). I make no claims of outperforming or even being the equal of a large Torit cyclone.

Quote from: wrobertdavis on November 25, 2007, 05:31:21 PM
One of the links on your webpage apparently points to your old forum software URL. You might want to correct that.

Thank you!

The version of phpbb that I was using needed updating anyhow.  Switching was a no-brainer.  Thanks again.

We lost a few messages when I decided to change forum software (to make registering easier and photo uploads possible).  So here is a recap:

*****
Rick:  The baffel has a 240 degree by 1.125 relief.  What is the relationship between the elbow outlet and the 240 degree cut out?

JPT:  The elbow is under the portion of the baffle that is not relieved.  My tests have shown that the baffle should be placed so the back of the elbow is aligned with the beginning of the larger radius.

*****
fishnskiguy:  Neat design, I'm impressed!

My question is similar to Rick's.  I'm having a very hard time orienting the inlet elbow with the baffle.  As I understand your reply to Rick, a particle of dust would exit the elbow and then travels along the entire 120 degree portion of the baffle that is tight to the can, and then reaches the beginning of the 240 degree 1.25 inch relief, and then drops into the can somewhere along this 240 degree section.  Do I have this correct?  A drawing or photo showing the location and direction of the elbow relative to the baffle would be a big help.

JPT:  You are nearly correct, except the 120-degree portion starts back at the bend of the elbow.  So the particle travels along all but approx. 2" of the 120-degree portion.  I found this design worked best to reduce turbulence at this critical point.

*****
fshagan:  Love the design of this...simple and elegant.

How large have you scaled it up?  I'm wondering if the limiting factor to using it with a shop vac is the diameter of the container rather than the overall size...if you could find a taller can that is the same diameter you might be able to increase the capacity without losing any performance.

JPT:  I don't have a single tool with a 4" port on it.  I have a Ryobi BT3K table saw (heavily modded) w/ a 2.5" port, a home-built router table with down-draft box (2.5" port), and Inca jointer/planer (2.5"), and an Inca bandsaw (1.5" port).  All in a tiny basement.  The shop vac is my friend.  :)  Perhaps a dust collector would work better w/ the table saw (even through the 2.5" port), but it would certainly work worse with the router table's down-draft box.

I did make a 20-gallon version w/ 4" in and out.  I tested it a little bit with someone else's blower for a couple of days in my shop and it worked great.  I rigged it so the blower sucked the can, and blew a cloth (1-micron) filter bag.

That was approx. six months ago.  At that time I thought a nice system would be a 30-gallon can with two 4" inputs (on opposite sides of the top) and a single 6" output.  That output would connect to a 1.5-2-HP blower motor, which would push into a large cart. filter like you see on the tops of nicer Jet and Delta single stage collectors.  I would rig the filter in such a way that there would be a clean-out on the bottom (much like large cyclones have on their filter stacks.  You could probably assemble an entire system like that for under $250 if you carefully assembled the parts of eBay/craigslist.

To answer your other question, the taller the can (provided you have the room) the better.  Or at least, the taller the can the more it will hold (someone has to lift it when it is time to empty.   :o)

The diameter of the can does affect the performance.  Too large and the baffle will never be cleared of the debris.  That is the worst case scenario.  Too small is a problem, too, as things can spin too fast to settle from the stream.  HOWEVER, in spending the last couple of years observing these things working, I'm amazed how quickly even the finger material will escape.

Many people have asked questions in the forums where I make my initial announcements. That is fine, but I thought I would recap answers here to avoid overlap.

*****
Q: Length of video is too long.
A: I agree, I'll do something about that. I was worried that if I edited it people may feel light there was some slight of hand.

*****
Q: Could you add the baffle to an existing design like the Lee Valley that has the exit port in the center.
A: Results less than optimum because of the angles of these ports. Really need more typical cyclone alignment. Don't get me wrong, the baffle will help the LV, just won't be optimum.

*****
Q: If you suck up the pile as fast as you can [in the video], does the filter still stay clean?
A: EXCELLENT question. There are two things you don't want to do to a cyclone: (1) Stuff it (where you feed it too much debris and too little air) and, (2) Starve it (where you feed it little debris but little air, too).

If I do a full seven gallon stuffin of my cyclone (feeding the entire seven gallons in what I approximate as 1.5 minutes or less--I've never timed it), probably about a cup of debris will make it to the vac.

Of course, full cyclones are not immune to this either.

The answer is that the entrained particles have to be entrained (in air) to be separated. If there is no air, there is no separation.

*****
Q: Very impressive. How does the cyclone do with extreme fines, like you get using your random orbit sander, or a drum sander? I would be interested in making one as I'm tired of beating the squat out of my filter to knock the dust out of it. Please keep us appraised of the larger size tests, this might be the ticket for my drum sander with my HF dust collector.

A: It would certainly help but in reality I've read complaints even from people with full-blown cyclones that have not been satisfied w/ performance on their drum sanders. If I can get some interest I may try to find a volunteer in the Milwaukee area w/ a drum sander where I can do some testing.

I have done testing w/ orbital sanders and have been happy with the results as long as I make sure I don't starve the cyclone. The 1" hose I have for my Porter Cable sander is okay. But I always open another blast gate a little to make sure the cyclone has enough air to perform separation.

*****
Q: I assume there's something special about the angle of the inlet as well as its placement? What's that wing-nut looking thing in the middle of the baffle?

A: The placement of the inlet is somewhat of a compromise no matter what we do. That is because the ideal placement would be a tangent to side of the can, but given the size of the inlet and other obstacles, the best we can hope to achieve is to get the inlet close enough to the side that it somewhat gracefully eases the incoming air into a circular rotation in the can.

The wingnut is just there to cover the 1/4" hole in the center of the baffle that is a remnant of the circle cutting jig I used to make it.

*****
Q: I use the cyclone system but it is the small pail size. As an amateur who doesn't run the shop every day, at least not the sawdust making machinery, this seems to be all I need. So, how will this work with this size? It appears reasonably straightforward to build, but your lid will cut down on the available space inside. Should I go to a larger size can?

A: The baffle reduces the capacity of my 10-gallon steel garbage can by approx. 2.5 gallons. If you can live with approx. 7 gallons of capacity (quite a bit for the average part-time home woodworker), I'd say to just use the 10-gallon can. Otherwise it can be ramped up, but there is a limit (we need enough CFM to enduce full rotation and therefor adequate separation).

*****
Q: Phil, that is awesome. Do you think a similar one would work if you used smaller pipes and a five gallon bucket and ran it with a shopvac? And did you do anything to ensure that the cyclone lid was snug against the trash can? How did you avoid leaks?

A: This small model I demo actually is used w/ a shop vac. I used a ten gallon steel garbage can, which really isn't that much larger than a 5-gallon pail. You might be able to squeeze the design into a smaller 5-gallon pail, but the performance would probably suffer a little unless (like you mentioned) you reduced pipe diameter accordingly.

There is another reason to avoid plastic pails, though. And that is to avoid static problems. I'm not concerned with the chance of explosion, but rather have noticed that static charged plastic can interfere with the operation of the cyclone.

The tops I've made during have been of MDF and sometimes (when I've got scraps that are large enough) baltic birch. I cut a lip on the plywood and it "snaps" down into the steel can quite nicely. No leaks, and I don't use any weather stripping.

A special thanks to Frank Hagan over at sawmillcreek.org for suggesting I switch forum software from phpbb to SMF.  The phpbb software posed a number of challenges that the SMF software seems to alleviate.

I apologize to posters that will have to re-subscribe, but at least we'll have an easy way to upload pictures and graphs now, and new users will find signing-up a lot easier.