J. Phil Thien's Projects

When I started to build my dust separator, I found that all the stores I shopped had the very same 20 gal. galvanized trashcan.  Nothing heavier.  I purchase the least expensive one I could find and went ahead with the project.  Everything worked as planned except I experienced the ?oil can? effect many have talked about.  In order to solve the problem, I first bought a second trashcan and, cutting off the top and bottom, I inserted it inside the original can and essentially made a double walled can.  A lot better, but still some oil-canning if the vacuum hose was blocked.  Then, I thought I might come up with a relief valve that might be cheap and easy to make.  The following photos show the results.  And it works!  All parts were purchased from my local True Value hardware store.  

VACUUM RELIEF VALVE
Parts List:

?-20 x 6? full thread hex head bolt (1)
?-20 nylon inset lock nuts (4)
?? x 1-1/2? steel fender washers (2)
?? x 2? rubber fender washer (1)
?? standard steel washers (2)
7/16? x 2-1/8? compression spring, Century Spring #C-676 (1)
3/8? lamp nipple with set of nuts, lock washers and washers
2? PVC pipe cut to 5-1/8? (1)
2? PVC pipe cap (flat ended) (1)
2? PVC pipe slip union (1)
8 x 3/8? automobile trim screws with points ground down (3)
6 x 5/8? machine screws (3)
?? MDF to make 2 disks
?? MDF to make one disk
   
Special Tools Needed:

Adjustable circle cutter - for precisely cut disks ? use on drill press
3/8? Forstner drill bit - seems to work better than standard drill bits

Theory:

Disk on right end of photo is secured in pipe with 3screws.  Middle and left disks move as a piston.  When vacuum tube is blocked, suction moves piston down pipe until middle disk clears the inlet holes in pipe.  Outside air bypasses the trashcan and normalizes pressure until blockage is cleared.  Once blockage is cleared, spring returns piston to normal position and full pressure comes through dust separator.

Construction and Assembly:

Disks:  These are the critical parts.  If they are not made to the correct size, the valve will not work smoothly.  MDF works better than plywood or hardwood.  Keep in mind that PVC pipe interior is not perfectly round or smooth.  Middle and left disks must move freely.  Carefully measure the inside of your pipe and, using the adjustable circle cutter on a drill press, cut disks to be as big as possible and also allow this movement.  Right disk can be a little snugger.  Don?t be afraid to remake disks to get the right size.  Drill 6 air holes in middle and right disks with 3/8? Forstner bit.  Try drilling small pilot hole for each air hole and enter from each side with Forstner bit to avoid break-out.  Drill 25/64 hole in center of right disk for 1-1/4? to 1-1/2? length lamp nipple.  Assemble with bolt, nuts, washers and spring as shown.  Left disk is solid.  I suppose it could be made with the ?? MDF, but there must be enough room between the middle and left disks to let in air.

PVC Union:  In order for the union to fit on the 2-1/2? ?T? I used, I had to stretch one end of it.  I did this using a heat gun.  I heated the end for several minutes and forced it over a 2-1/2? disk I had cut from ?? plywood and screwed to a flat board.  The same can be done using an oven, boiling water or boiling oil.  I find the heat gun the easiest.  Be sure to use gloves!  I used a short piece of PVC pipe inserted in one end for leverage, but be sure it is scrap pipe because it may deform also.  You can either wait for the union to cool off gradually, or dip it in water.

PVC Cap:  Be sure you purchased a cap with a flat top and not one of the rounded type.  Find the exact center and drill a 1-3/4? hole using the adjustable circle cutter on a drill press.  In order for the cutter to not damage the hole you are cutting, I found it best to cut a ?? plywood disk and insert it inside the cap.  This gives the center drill something to steady itself on when the cutter goes through the PVC.  

PVC Pipe:  6 holes should be drilled with your 3/8? Forstner bit.  They will be about 1-1/4? apart.  Position them so their centers are exactly under the middle of the middle disk.  This keeps air from flowing through the valve under normal air pressure.  Be sure the inside is smooth, and no burrs interfere with the piston.

Assembly:  I assembled the PVC with screws instead of glue so I could disassemble it and adjust it if needed.  Drill pilot holes and attach cap end with short screws.  Be sure points are ground down so they do not protrude onto the interior.  Position slip union on other end and drill pilot holes.  Insert piston.  It should protrude past the pipe end (not the union) about 3/16?.  Push it level with the pipe end and fasten it with three 5/8? screws.  Screws go through the union and the pipe and into the disk.

Use:  All my system uses 2-1/2? fittings.  I have used a 2-1/2? ?T? between the dust separator and the vacuum.  I put the relief valve on this T.  Under normal use, the relief valve is closed.  When the hose is blocked with trash or my hand, the relief valve opens, letting outside air into the vacuum, and the trashcan does not collapse.  If I do not want to use the relief valve, I put a modified 2-1/2? PVC end cap over the opening.  I stretched the cap the same way I modified the PVC union described above.  Can this valve be scaled to a 4? system?  Make the same valve and connect with an adapter.

It is possible there is another way to do this, but this is what I originally drew up and it seems to work as planned.  My only experimentation was with springs.  The first one was way to strong, but my second try worked.  I did try making the pipe a little shorter to get more pressure on the spring, but I went back to the original length.  So feel free to experiment all you want, but this plan does work if built as described and shown.  Good luck!

Pure genius!

Another nice description, and beautiful photography of a great idea!

Like you, I just put together a shop vac/Thien separator setup with a 20-gallon garbage can that collapses when the inlet hose gets blocked.  And, like you, I started thinking about a vacuum relief valve.  Unlike you, however, I haven't got as far as actually trying to build the valve.

I was considering something like what I've shown in the picture.  There would be a bolt extending through an oversized hole in the lid with the head of the bolt on the inside of the can.  Under the head of the bolt would be a big washer-like plate, perhaps with an O-ring or something similar to help it seal against the inside surface of the lid.

On the outside of the lid, a spring would live between the lid and a nut and washer threaded onto the bolt.  This spring would be in compression and would act to keep the plate under the head of the bolt sealed against the inside of the lid.

When a blockage occurred, the increase in differential pressure across the valve would compress the spring so that the inner seal would leak and thus allow air to flow around the bolt and into the can.  The compression in the spring--and therefore the pressure at which the valve opened--could be adjusted by moving the nut up and down on the bolt.

After seeing your valve, which would be quite a bit more complicated and difficult to build, I'm wondering if you might have started with something simpler like I'm talking about and discovered some reason that it wouldn't work.

Any insight?

Thanks,

-- Russ

I just thought that any valve should be away from the separator.  I didn?t want sawdust to prevent a gasket from sealing.

For me, this valve was a lot easier to build than it was to describe.  Once I drew up the plan and had all the parts assembled, it took less time to build than to write up.

John

Hi John,

Thanks for the reply.  I thought about the sawdust issue, too, and thought it wouldn't be a problem for a two reasons: 1) If the separator is working right, there shouldn't be any sawdust in the upper chamber where the valve would be, and 2) when the valve opened, the clean air rushing in should blow out any debris that might disturb the seal.  That's the theory, anyway.

I think I'll go ahead and try my idea and see what happens.  Then I can report on exactly why it didn't work!

-- Russ

PS: What kind of lighting did you use when you took the pictures of your valve?  Those pictures really are nice.

Great idea!!

Hi Russ,

As you plan your valve, I?d also have a way to plug your opening while you are trying various set-ups.  Otherwise your separator may be out of commission while you are experimenting.

I use a Nikon D40 with an ED 18-55mm zoom.  I have a mini studio set up in the basement.  The light tent is built out of PVC tubing and white nylon rip-stop fabric.  My 3 lights are 5000K compact fluorescents mounted in reflectors.  I always use a tripod for close up shots.  See photos.  The photos of the completed valve on the separator were taken outside.  I am not an expert in Photoshop Elements, but I use it to crop, adjust exposure and color, and adjust image size for export.

John

Got it.  Thanks!  The light tent explains everything.

-- Russ

that is a really great idea and application.

I didn?t realize the flat cap would be hard to find since my hardware store always has them.

The reason I made the open end like I did was so I could make adjustments in the position of the disks if needed.  When I designed this valve, I didn?t know how well it would work, so I made it for maximum flexibility.  The rubber washer is to provide an air seal when the piston is in the ?closed? position.  But there is no reason that the end can?t be capped off.  Just adjust the bolt so it stops at the end cap.  You could also eliminate the left disk.  In other words, weather the end cap is closed or open, it determines the position of the piston in the ?closed? mode.  You probably don?t need to cut the bolt - just adjust the nuts so it touches the cap.

The spring is only preloaded about 3/16? when the piston is assembled, and then another 3/16? when the right disk is attached to the pipe. 

Hope this helps.

John

Response to dabullseye-

I was wrong about the need for an open end cap.  When you close off the end cap and eliminate the solid end disk, there is nothing for the extra vacuum pressure to pull against.  So, when the vacuum intake tube is blocked, nothing happens to the relief valve.
I guess you could try to cut a hole in a dome cap, but I think a flat cap would be better.  I found mine at a True Value Hardware.  Also try Ace Hardware.  If all else fails, make one by heating a dome cap with a heat gun and flatten the end.

--johncarl

This an old post, but Russ did you ever make a release value?  If so what where the results?

thanks

erik

Last week I purchase a 50 cent spring and $5 1.5 inch PVC relief value.  Going to see what the spring does and if it does not work, enlarge the hole on the lid and put int PVC relief value.  While testing the the various size PVC values, the smaller ones seem to be better engineered and smoother operationally than the larger ones.

I was hoping someone would responded to this, but alas looks like I am flying blind.

So I started by drilling a 1/2" hole in the top of the lid about an inch behind the center vacuum intake pipe at 9 clock when the collection pipe is at 12 o clock.

I Really was not sure if the hole here would be an intake or outtake or reverse directions on a clog. The hole seemed to remain an intake the entire time.  Did some testing of the hole its self uncovered and with no value.  It did relieve some pressure, but not much.  My 4.25 HP shop vac had an adapater on it to 1.25 for my router and the garbage can was still was crushing.  Then increased the hole to 3/4" and it was a little better, but still crushing.

When with the simple washer value approach and could not get the value to release with a very weak spring.  Initially I had the spring on the wrong side, but even with that corrected, the pressure was not great enough.  Not sure the physics, but the baffle cyclone generated pressure is probably much greater than the vacuum pressure. 

Overall with a 3/4 hole I did not product a noticeable reduction in suction.  Not sure what to do now, leave the hole, fill the hole, move the hole to behind the collection intake, or try the pvc value on side on the vacuum intake.

Hi

I was just doing the math...

A one inch round hole has a cross sectional area of 0.785 sq inch
Six 3/8" holes have a total of 0.662 sq inch.
A 3/4" hole has 0.44 sq inch.
A half inch hole has 0.196 sq inch area.

A two inch tube has an area of 3.14 SQ inch.
A 2.5 inch vacuum tube has a cross sectional area of nearly 5 sq inches!!

So if you use a 3/4 inch hole in your can lid, you are asking 0.44 sq inches to absorb the volume being sucked through at least 3 to 5 sq inches of vacuum tube.

Secondly, I think that the force pushing/pulling the valve open would be related to the area over which the force is applied.  (Force = Pressure X Area)
Vacuum and pressure are really the same thing (the vacuum created just lowers the pressure inside the tube or can,
the outisde atmospheric pressure remains the same, and is what is crushing the can.

The idea is that a rather large disk would need a stronger spring to keep it closed.

So if you increase the size of your washer and hole in the can, your weak spring will succumb to the vacuum in the can (pressure from outside the can).
The tension on the spring could be adjusted to just keep the valve closed under normal operation, then a skosh more.
When a blockage occurs the valve would open allowing air in to buffer the collapsing force.

I am making a 20 gal tank, I could do some experiments with different sized holes to see when the can no longer bows in with hole wide open.
It seems to me you would want that size, and then add a little more, to get a relief valve which would be adjustable.
Maybe the bigger the better... you can always add more spring power. This would allow finer adjustment and control.

Jerry