Bell Siphon questions

[JRComito]

I recently designed and built my siphon, and I wonder how important it is to have a "gravel guard" around it. The other question is how much more effective is a siphon with a snorkel versus without one? I am testing with a tank of water without gravel, so I am aware that the flow rate is going to change. Is it enough to restrict the siphon's flow? Is having the gravel in the bed going to have a big impact on flow, and should I "over-build" the siphon to be ready for that?

Thanks for any help.

[Jim Fisk]

My opinion: Go with a Bernoulli funnel that in your case would be 1.5" and stop fooling around with little tubes that are known to clog up. After your gb is full of gravel and God willing full of plants is NO time to be "experimenting" Just say'n Also it occurs to me that it should work BETTER with the gravel in as the water raises 3 times faster given the same flow, after gravel is added. (the gravel displaces about 2/3 the volume) I was a little worried about mine until I added gravel and then they really took off and have been flawless for 2 months now.

JRComito said:

Well my siphon worked fine last night with 2/3 of the gravel in. This morning I filled it with the remaining gravel, and now the siphon won't start. I am running a 3/4" bulkhead with a 3/4" standpipe, a 2" bell housing fitted with a 1/4"mpt to a 3/8 barb for the snorkel. I need to get inside to my pc so I can post some photos.

[JRComito]

Ok so here are the photos:

I removed the trap before I came in, and it seemed to help, so hopefully when I go back out it will not be flooding over the sides.

[JRComito]

The reason the water inlets at the bottom of the bell housing are up so high is because the 2" pipe fits snugly around the top of the bulkhead fitting. I think this is where I am having troubles. I tried a different configuration by cutting 3 triangles in the bottom and it seemed to work alright. I may go back to that idea.

[Robert Rowe]

http://community.theaquaponicsource.com/photo/albums/mason-jar-bell...

[Jim Fisk]

I have to ask: why do you have a trap in the drain? I have done hours of research online before I built my bell siphons and I have never seen a trap before. Like I said above: traps are to prevent sewer gas from entering your house. I see absolutely no reason for them in AP. One 90 is all you need just to interrupt the flow enough to fill the tube and start the siphon process.

Also If you are going to feed the bed in one corner I suggest you drain at the opposite corner for the sake of flow thru the bed.

Also I suggest you drain from as low a point on the bell tube as possible, not that far up. You want to evacuate as much as possible. Get rid of the fancy bulk head fitting and get some cheap threaded adapters and an "o" ring for the seal. Nickel and dime components.

This is a simple process and you are making it way too complicated. KISS. Try building one like I show above just for shits and giggles. I think you will be blown away and you will get rid of all the extra junk. The only mods I perform are temporary ones like extending the funnel 1" to raise the level for seeds or baby transplants so the roots reach water. Once established I remove the collar so as to keep the water out of the sun, like 1-2" down. I use a 2" temp collar to temporarily raise the water an inch above the media in order to flush cut worms to the surface (works awesome by the way), etc.. I have never had any problems ever with the siphon doing a perfect job. When the roots get big enough to maybe interfere with the gravel shield you just rotate it and cut them off. I know this is a whole different approach but TRY IT. Your toilet is SIMPLE and works every time (yah, yah, I know. That's because of SOLIDS). Your siphon should be no different except you do not need the trap for the reasons stated above but like your toilet you do need the FUNNEL. In fact the lower the flow the bigger should be your funnel. In general a 2-1 ratio (funnel dia to down tube dia). I don't know where all the straight tube siphons came from but in my opinion they are the odd ball and not the proven Bernoulli design. Crapper knew better way back in his day.

Hope this helps.

[John Cubit]

I agree with Jim.  I have been involved in Aquaculture and sometimes a problem becomes more complicated when it is over thought, keep things simple and build off of that when needed and that is exactly what I did.  Using 3/4 cpvc to a special purpose cpvc adapter.  But instead of an "O" ring I purchased a sheet of 1/32 plumbers rubber.  Made my own gaskets and used one on either side of the fitting, works great.  Now that I cuttings in the bed as well as some seeds I replaced my 1" to 3/4" adapter to bring the water level up to just below the top  of the media.  So far so good. Pull it off and the level drops almost 2".

I have tested traps and that led to problems like siphon lock, constant run regardless of flow.  The only thing I changed was adding additional pipe to add aeration to the water.

Good luck though it is a hell of a lot of fun playing with PVC pipe regardless..

Jim Fisk said:

I have to ask: why do you have a trap in the drain? I have done hours of research online before I built my bell siphons and I have never seen a trap before. Like I said above: traps are to prevent sewer gas from entering your house. I see absolutely no reason for them in AP. One 90 is all you need just to interrupt the flow enough to fill the tube and start the siphon process.

Also If you are going to feed the bed in one corner I suggest you drain at the opposite corner for the sake of flow thru the bed.

Also I suggest you drain from as low a point on the bell tube as possible, not that far up. You want to evacuate as much as possible. Get rid of the fancy bulk head fitting and get some cheap threaded adapters and an "o" ring for the seal. Nickel and dime components.

This is a simple process and you are making it way too complicated. KISS. Try building one like I show above just for shits and giggles. I think you will be blown away and you will get rid of all the extra junk. The only mods I perform are temporary ones like extending the funnel 1" to raise the level for seeds or baby transplants so the roots reach water. Once established I remove the collar so as to keep the water out of the sun, like 1-2" down. I use a 2" temp collar to temporarily raise the water an inch above the media in order to flush cut worms to the surface (works awesome by the way), etc.. I have never had any problems ever with the siphon doing a perfect job. When the roots get big enough to maybe interfere with the gravel shield you just rotate it and cut them off. I know this is a whole different approach but TRY IT. Your toilet is SIMPLE and works every time (yah, yah, I know. That's because of SOLIDS). Your siphon should be no different except you do not need the trap for the reasons stated above but like your toilet you do need the FUNNEL. In fact the lower the flow the bigger should be your funnel. In general a 2-1 ratio (funnel dia to down tube dia). I don't know where all the straight tube siphons came from but in my opinion they are the odd ball and not the proven Bernoulli design. Crapper knew better way back in his day.

Hope this helps.

[Rob Torcellini]

The trap in the bell siphon has nothing to do a a toilet design...trapping sewer gas.  A real bell siphon (with a trap and bell) is designed to allow the siphon to start regardless of the in-flow rate.  This design dates back to the 1800's, maybe earlier where they used the bell siphons in sewer system to flush large quantities of waste through a sewer line at once to help prevent the lines from clogging due to slow flow rates.  (Today, it's not necessary since the lines are much smoother...)  The trap is used to create a burping effect to release the pressure that is created inside of the bell area and the trap.  When this pressure is released, the water level inside the bell quickly rises and overflows over the standpipe.  If you remove the trap and just have an elbow, it will require a larger water flow to create the some turbulence at this elbow to get the bell area to "burp" a bit.  (go back somewhere in this thread where I posted the video about how a bell siphon works.)

The Affnan siphon does work well, but is designed entirely on a different principle whereas you need enough flow through the system to create the Bernoulli effect down the standpipe.  If you have a very slow fill rate, this siphon will never start (unless you narrow the pipes)

There are several types of siphons...before bashing on the different types, please understand the physics behind how they work.  I've worked in the sewage disposal industry back in the 80's so I know a thing or two about some fluid dynamics.  ;-) 

[Jim Fisk]

Good point Rob. I did notice that JR's flow looks to be very small but it is a still pic so hard to be sure. That makes sense with regard to allowing the pipe to fill better thru the use of a trap. Did not mean to bash. I am just an impatient fellow:-) (just ask my wife) and I just want to see everyone's system run flawlessly. Perhaps he needs a bigger pump for a more foolproof system. The smaller the flow the more tweaking involved. I can see that now but I have never run such a small system so I had not considered the problems associated. Perhaps he should try 2 90's and no trap for experiments sake. The trap can make the siphon much harder to break. It is a balancing act that is very flow dependent and not all traps are created equal. Personally I would simply up the flow.

Hoping I did not step on any toes this time

[Rob Torcellini]

no harm done!  ;-)

[JRComito]

You guys keep at it! I am learning with every post, and I thank you for the invaluable advice. I have changed things a bit since these photos, and the siphon works, it just has trouble starting and stopping. I have been playing with the trap a bit. I have it going straight down from the bulkhead to a 90, then another 90 up, and another 90 over and out into the FT. I played with the level of the terminating 90 to see if I could achieve the "burp". The bed would fill but flood the gravel and be about 1"-11/2" above the standpipe level. The siphon will eventually start and empty the bed, but it doesn't stop! It keeps trickling out of the 90. I have also found that by the 3rd or 4th cycle, it will only fill about half way and the volume of water starts the siphon over again. I have set and watched the entire cycle enough times now to be aggravated. Occasionally, I will intervene with the flooding of the bed by grabbing hold of the bell housing and lifting it until the siphon starts. It sucks the housing down to the bottom of the bed and siphons as it should. It doesn't break the siphon every time, and that is what I'm after. I tried shortening my standpipe about 1/2", but it still floods the gravel. I leveled the top of my standpipe to be 2" below the top of the bed's framework, which is about 1" below the top of the gravel. I need some help to get this thing working correctly. I have planted the bed with some seedlings, and they can handle the flooding for now, but I want to sow some seeds, and I don't want them to sink out of germinating range every time it floods and drains.

[TCLynx]

If the trap is causing the siphon not to start it is probably too big a trap.  You might simply try coming down to an elbow and over to the tank, that might be enough restriction to let the siphon start without overflowing your gravel and then without the complete trap the siphon may be able to stop. 

If the siphon is able to start but is still having trouble stopping, then you add the breather tube.

[JRComito]

Ok, thanks TC. If I need the breather tube, at what level should it "suck" from? Should it be above or below the inflow point of my bell housing? I am experimenting with housings at this point because I am only informed enough to be dangerous to my project! :-) I will post some more photos soon.