J. Phil Thien's Projects

QuoteHowever, I find that there is quite a bit of dust piling up around the circular intake pipe, above and below it. I don't have the skills (or didn't then) to fit the vertical cylinder of the drum to the horizontal cylinder of the intake pipe.

Peter-  It sounds like you have a slightly different intake, which we will call 5. I hope the updated diagram is accurate.

I can't say I am suprised by the strange voids you mention, and the inability for the filter to work on different size material.  I applaud your removable top, but it sounds to me like it does not work to specifications that I would consider acceptable.  If I spend a number of hours and some money or scrap building this thing, I want it to work considerably better than you describe.  Good work, I dont mean to sound like a dick, but... I expect more. 

QuoteMy understanding of the way this thing works is that we want airborne particles to collide with the sides of the drum, reducing their velocity which causes them to drop out of the air stream

As far as I can tell, and based on the literature I have been exposed to, The entire thing is about a cyclone, and while smashing particles into the vertical sides of the sphere (cyclone) decelerates them considerably, this is a by product of a more general and efficient idea; the heavy particles in the air are forced to the bottom and outside of the cyclone, where they wind up falling due to loss of velocity, which may or may not be caused by them contacting the side walls; in most cases I would imagine particles would not/never impact the side walls.  Again, this is not based on demonstration. 

QuoteRectangular duct is less efficient than round duct, so HVAC practice is to upsize rectangular duct by about 10%.  That's the number I used for the design of my round to rectangular transition piece.  However, with the transition piece bolted directly to the separator, the rectangular run is so short it probably has very little negative effect.  That said, I would still use the 10% guide and I would definitely not go smaller.   

While the last poster had some slightly useful information, I feel like this is much better.  I do imagine that if there is a difference, it is slight.  I am not sure what you mean by a 10% guide.  Say your intake is 6" circular, so you recommend 6.6" rectangular to accommodate the square?  You have a design similar to 1, and are providing the specs relative to the section? I could see this working in almost the same manner as #4, because the fit of circle to square is closer than most other projects (think 4" circular and 6.5" square... or even a larger ratio.  can site if necessary). 

What if we try to increase the rectangular run?  I am imagining a multiphase action of dust- it hits the tube and is sucked based on how it is received... it hits the short (long?) rectangular appendage, where it is scattered about the perimeter, most concentrated towards the edges of the perimeter of the square/rectangle, then it hits the cyclone where it is more (less) efficiently separated?  This makes sense- moving air quickly through a rectangle should push any heavy airborne material to the edge of the square/rectangle so that the air column can more more efficiently towards the center. 

I think we may be getting somewhere.  Peter, thanks for sharing, sorry if I was a little harsh.  Exactly what sizes are your shop vac (hp/gallon), your baffle (height, diameter, intake position and diameter, and outtake position and diameter) and perhaps your receptacle (20/30 gallon?)

I will be using a 5 (peak) HP 16 gallon rigid shopvac, with a 20 gallon can receptacle, planning on the baffle being 3.25" tall (interior, with approx

Quoteyou need to go back and see if people are using a rectangular. or round pipe coming in. i think the best is to use a rectangular pipe and keep it the same height as the inside of the separator. this really helps smooth out the air flow as it hits the wall and begins it way around the separator.

See you are assuming model 1 is best.  I assume that either model 1 or 4 is best, but I do not feel comfortable judging (without testing) that a transition from a circular to a square input is more efficient than maintaining a circular intake... If that is the case, and model 1 is more efficient, what length of square appendage is best?  Are they different for shop vacs and DCs?

QuoteI'm sure that theory and opinions would be plentiful.

I can only imagine that most everyone would consider their design superior, but would not be able to cite reasons other than "because I made it and it took considerable time/effort/money"  :)

QuoteI'm also sure in the end, you will wind up building the best system possible taking into account your particular needs.It sounds like you've really done your homework and are ready to start building your Thien Separator. I say, go for it!

I sure hope so.  That is the idea behind this post :)

QuoteIf you should decide to build various intake designs, run tests and share your findings; I can only imagine that many forum members would be grateful.

Ideally, I made this post to see if anyone has experimented with different inputs, and what they found.  If this has been done, I am surprised there is no post on it.  Perhaps I will be the one to do it.  Frankly, if/when I build one, I do not imagine that I will get it functional and then feel the need to mess with it more.  I am just a hobbyist, so I don't really care about squeezing a few extra percent efficiency. 

It would be very difficult to control all the variables in such an experiment.  Perhaps you guys could help a bit with that, or offer suggestions.   

Here are a few important controls and design considerations for the experiment:  I would imagine that any tests would need to require the inside of the shop-vac to be nearly spotlessly clean, and the dust piles would need to have consistent quantities/ratios of fines and chips, as well as consistent mass.   It would need to be sucked up at the same rate, to avoid unfairly overwhelming one design or another, although there could be differences in efficiency based on speed.  Perhaps 1 works better when there is lots of dust flying in fast, for example, but requires those conditions to outperform 4...  Next, and most importantly and most difficult, I would need a way to measure the dust collected on the inside of the shop vac and the filter.    If I had a very precise (laboratory grade) digital scale, that gave grams to 3 or 4 decimal places, perhaps I could weigh the whole thing before and after sucking dust, and report the weight change.  This assumes I have access to a very fancy scale.  Qualitatively looking/shaking the filter may provide a good answer to the question, but only if there is considerable difference, which there likely isn't.  It would also likely take large amounts of dust to create a noticeable difference, adding to the experimental challenge.  Do I test 1 cubic foot of dust, 1 lb of dust, or more?  . 

It seems easy enough to make model 1, and then modify it to be either 2, 3 or 4.  Using one tophat with two different intakes, I would need to design the tophat such that the top could be sealed well, yet still be removable, which seems like a huge challenge. This all seems like an awful lot of work If i want to experiment with the "classic" then input as originally described by Phil (iunput from top of cricle, not side), the mod would be more tricky to make without compromising my original "appendage and circle" design...  If trying the classic input, It would need to be in a tophat and not recessed into the can so that the edges of the baffle are airtight (not relying on the wall of the can to make a perfect fit" to further control variables.  The tophat seems to be far more efficient for this reason.  Again, this conclusion is not based on any empirical testing, but seems to be a consensus among the community. 

QuoteJust for reference, my system was designed and built using option '1' in your illustrations.

Gale, I actually know yours was style 1.  In fact, your design was the last one I looked at when I felt the need to address this problem.  Of all the models I looked at, yours seemed to be one of the most efficient, if not the most efficient from the sample I reviewed, but I am curious if the appendage helps or hurts, and under what circumstances.  An area where the intake diameter is considerably increased just before the circular cyclone....Hrmm....

Either 1 or 4 has to be the best designs in my opinion, but I can't tell which just by looking at it.  I would also guess (as I have nudged at) that one design may be better for a DC while the other would be better for a shop-vac.  4 would seem to increase the speed of the vortex, but would likely increase pressure as well, while 1 seems to decrease both speed and pressure somewhat...

After writing this gigantic essay, I think I likely will test differences between 1 and 4, but as I do not have a DC, will be using a shopvac.  I would not be surprised to find they are almost identical,  (small margin of error), but I'd be less even surprised if gains were circumstantial, and based on power source (shopvac/dc).   Any conclusions I make will be limited in application to shopvacs... Sorry fancy DC boys :)  It would be interesting if I found that the appendage model works best, but that it works even better if the appendage is unusually long. 

Any results or experimental preparation will be posted in this thread, but I will likely start a different thread for my build. 

Thanks to the great community, and the all wise Phil (he is kinda like a god on these forums, isnt he? :) ) for sharing such a great concept. 

You may have missed the point of the post.

Which is better? 

Or is the "classic thein" input the best, from the top, with nearly 180 degree bend, but a perfect circle, and likely more strange dead spots from the input... Something makes me think his way is best for shop vacs, and the arm model is best for dust collectors... Perhaps the near 180 degree bend in the pipes near the input reduces some static pressure yet maintains cfm?  Beats me :)

Hello! 

This is my first post, I have read the forums considerably, and have not found a specification on intake design.  I have seen intakes with four different styles, and one seems to stand out as superior.  I am curious if it might be the extra 9% I have heard people talking about - "mine gets 90% of particulate, it would be great to get 99%". 

I hope I can be clear in describing the diagrams I threw together in paint.

1 is where the entire baffle system is coated with something, and the hose attaches at the end of the square appendage (where the little x are). 

2 is where the hose or assembly sticks into the appendage somewhat,  but does not approach the circle, and is not cut at all, it is left square (from top view).  Typically only the circle is coated in these diagrams, not the appendage. 

3 is where the hose or adapter sticks into the appendage, approaches the circle, and is trimmed to match the circle.  In this case, the coating material wrapped around the circle is not the length of the circumference of the circle, it is shorter than the circumference by the width of the appendage.  I imagine this leaves strange voids where the x are found. 

4 seems to me to be superior.  It functions in much the same manner as style 1, but it elimanets the problem from switching from a square appendage to a circular hose.  It also functions in a similar manner as style 3, but with the voids surrounding the intake adapter filled in to match the circle.  In 4, A lexan piece would be cut to length of the circumference and would have a very long ellipse cut out to receive the intake hose/adapter

I am not sure if one might want the strange little dead spot (x) or not.  I have seen videos, pictures and descriptions of people using all 4 styles. 

Which is best and why? 

I hope I was clear enough with the picture/thought.  If you guys want, or if I am unclear, I can go through design videos/thread and link them to these 4 styles. 

Thanks,

Ged