Semi-automatic frozen food dispenser

ReeferAl · 01/31/2008, 06:15 PM

Well. I've been fooling around again with some ideas I've had for making a device to dispense (thawed) frozen food. The result is a bit impractical as it ended up costing about 3X what I intended, but it seems to work.

My goal was to make a feeder that could:[list=1][*]dispense small amounts of food throughout the day[*]feed anything frozen up to about the size of mysis [*]keep the food fresh for an extended period of time- at least several days[/list=1]

I decided that to do this I would need:[list=1][*]a chilled container[*]something to stir the food to keep it suspended[*]a method for dosing that would not be affected by the particle size[/list=1]

What I came up with is a bit more complicated than absolutely necessary, maybe even a bit "Rube Goldberg", but it does everything I wanted it to do.

This is unit disassembled:

The picture doesn't include it, but there is also a solenoid valve that can be opened for a set amount of time (currently about 5 sec) as often and whenever I want. It is controlled by the Aquacontroller and is set to open every 12 min throughout the day. This feeds about 1 gal of food suspension each day.

This is the inside of the cooler showing the Ice Probe inside:

And finally, here it is assembled:

It sits above the level of the tank and feeds by gravity.

ReefEnabler · 01/31/2008, 06:23 PM

wow, what a cool idea!

have you been using it much? I might have to try that someday

ReeferAl · 01/31/2008, 07:50 PM

I've been using it for a couple of weeks now. I figure I'll empty and clean it every weekend. Otherwise I just add about 1 gal of saltwater with a bunch of food each morning. The water added is just about offset by the skimmate produced, although I still check salinity periodically and adjust as needed. I've been keeping an eye on nitrate and phosphate as this is an increase in the total amount of food added to the tank and so far they've been stable.

prugs · 02/01/2008, 09:35 AM

How are you regulating the amount of food into the tank?
Is it a drip system or solenoid system?

ReeferAl · 02/01/2008, 01:31 PM

Quote:

Originally posted by Gem Tang Rider
How are you regulating the amount of food into the tank?
Is it a drip system or solenoid system?

I use a solenoid that I got from McMaster-Carr . I can't link directly to the valve, but its item no. is 7876K12

It is a 3/8" solenoid so it causes less restriction of flow which would prevent large particles such as mysis from passing. If you were to only feed small foods such as cyclopeeze then a smaller valve would be fine. It's also easier to keep smaller foods suspended so stirring isn't as crucial. I initially tried using a peristaltic pump, which worked well for smaller foods but not the mysis. The solenoid opens for 4 to 5 seconds every 12 minutes only during daylight hours. I arrived at this timing by some trial and error.

Allen

ReefEnabler · 02/01/2008, 01:48 PM

have you considered trying multiple solenoids run off some kind of mini-manifold? i wonder if you let out less food but spaced at different locations in the tank if you could feed more with less, so to speak. maybe like 1 outlet per powerhead or something to get the food everywhere quicker? or are you seeing that it diffuses pretty quickly throughout the tank as-is?

ReeferAl · 02/01/2008, 02:56 PM

The outlet actually goes into my surge tank. When it surges it distributes the food somewhat via the 2 surge outlets. It does tend to concentrate it initially more at one end of the tank and the fish are always keeping an eye on the surge outlets. I could run it into the wavebox as that distributes it quite well, but I've found it tends to "gum-up" the propeller a bit. That was with larger food though so that may still be an option. I could also aim it at the closed loop intake. It would then be distributed all through the tank. I'm afraid that would chop up the larger pieces though.

Those are options I may try in the future as either would be pretty easy to do. As far as varying the feed interval, the timer function could do that, but with the food going into the surge tank and then into the DT it is varied somewhat anyway. In fact, it results in there being some food (like cyclopeeze) in the water almost all the time. My intention is to eventually add a school of anthias.

Allen

"Umm, fish?" · 02/03/2008, 12:05 AM

Very nice! I've been trying to figure out how to do this for years. How cold does the water stay?

ReeferAl · 02/03/2008, 12:09 PM

The water stays around 35 to 40 deg.

If I hadn't already gotten the Ice Probe for another project a couple of years earlier I would have thought about leaving out the chiller and just adding ice each morning along with the food. That would add a little more time to the process every day but would also save about $100. The way I built this it cost me about $300.

It might also be possible to use an air pump to keep the food stirred, but to do that I would want to find a way to have the air pump recirculate air from the cooler rather than always using warm outside air. That would cost a bit less than using a motor for stirring.

The other major component was the solenoid and timer. I'm not sure how to trim any expense from that.

Allen

ribs · 02/03/2008, 06:28 PM

Is that a 1 qt cooler and why so large.....so you can dissolve the foods in a larger volume of water? What volume of water to frozen foods do you tend to use so far?

ReeferAl · 02/03/2008, 08:03 PM

The cooler is a 2 gal cooler. The larger the volume the easier it is to regulate the amount dosed. For example, I am dosing 5 times per hour (every 12 minutes). The dosing goes on about 10 hrs per day for 50 times daily. I aim to dose about 1 gal per day. That means that each dose is about 2.5 oz of mixture. If I were only dosing a quart it would mean each dose is only about 2/3 of an oz. It's difficult to dose such a small amount very accurately. Additionally, at the end of the day about half is still left, which keeps the flow more consistent than if the jug were almost empty. Since I'm using a chiller it's also good to keep some liquid in the jug because once it drops below the level of the chiller probe it isn't kept chilled any more. I'm not saying this is the only way or even the "right" way to do it, but those were my reasons for the way I did it.

Allen

ReeferAl · 02/03/2008, 08:09 PM

Oh, another thing- If you go back to my first post you'll see that I had 3 criteria that I wanted to meet with this project. If the only criterion had been:

#1 dispense small amounts of food throughout the day

and I were only feeding something very small like cyclopeeze then it would be alot simpler. The stirring would be easier and there would be more choices of how to dispense. I had tried using a peristaltic pump initially. It worked fine with just cyclopeeze, but couldn't handle mysis which is why I went with the solenoid. With just cyclopeeze the solenoid valve could also be smaller if you went that route. Since most of my fish are larger though I didn't feel that cyclopeeze alone would be worth the effort.

Allen

"Umm, fish?" · 02/03/2008, 10:31 PM

Actually, you've stirred my imagination. What about adding RO/DI ice and frozen food to a blender, chopping the ice, then adding the mess to (essentially) an insulated cone and allowing it to simply gravity feed to the tank as it melts? It'd be hard to standardize the dose, but is that so necessary?

ReeferAl · 02/04/2008, 08:39 AM

When you say "gravity feed", do you mean controlling the amount by using a pinch device to adjust the flow or a solenoid?
The reason I use a solenoid is that the pinch valve causes a dam that allows only the liquid and smallest particles to pass and holds back the larger particles until the tubing plugs.
Not stirring the mixture will mean that the sinking particles will feed first, followed by those that are suspended and finally those that are floating. Many of the floating particles may also end up coating the inside of the container as it slowly empties.

As far as using RO/DI ice and a blender, I think that would be fine. You could also add cold or frozen saltwater. I've found that my skimmer removes at least as much water everyday as I add via the feeder so the feeder offsets the skimmer losses somewhat.

Allen

JaredWaites · 02/04/2008, 08:56 AM

This is a great idea...but I found something you might be interested in looking at

I'll attach it in my next post damnit

I can't edit this post and post the attachment...that sucks.

JaredWaites · 02/04/2008, 09:05 AM

http://www.audet.biz/499finalreport.doc

There ya go.

It won't let me upload it as an attachment as its too big.

ReeferAl · 02/04/2008, 09:29 AM

Very cool concept. I considered trying to build a feeder using the same concept with a screw device and thermoelectric chiller but decided it was beyond my skill level. It looks like this was a college engineering project so it may never actually be developed, although it would be nice if it were. Still, it may be a few hears until it hits the market.
Right now I am dosing about 8 cubes of food per day so I would still need to refill it everyday.

Allen

"Umm, fish?" · 02/04/2008, 10:07 AM

Actually, I wasn't thinking about pinching the drain at all, just letting it flow at the rate of the ice melting. So, maybe controllable with the amount of insulation you use. Hmm.

Good point about the food coating the sides.

The reason I thought RO/DI ice was that you'd have to keep it so much colder to keep saltwater iced.

For what it's worth....

funman1 · 02/04/2008, 11:17 AM

The Aqualifter will pump air just fine, and it's got an input and an output hose connections.
And the best part is it's only like $14.00

ReeferAl · 02/04/2008, 01:08 PM

Quote:

Originally posted by funman1
it's got an input and an output hose connections.

That would be ideal for mixing as it could recirculate the air in the cooler and not add alot of heat.

ReeferAl · 10/05/2008, 01:38 PM

It's been a long time since I posted here. I've deleted all the old pictures I'm afraid. The Ice Probe chiller failed within about 2 weeks.
I came up with a new design based on a small refrigerator. I bought a "dorm fridge" from Best Buy that was marked way down due to a damaged case- nothing that would affect the function. It was also based on a thermoelectric chiller. It would only keep the fridge interior at 40 to 45 deg and it too failed within a couple of weeks. Interestingly, there was nothing about the way I was using it that should have caused a premature failure. It appears that thermoelectric chilling has a long way to go to be reliable.

I used the same principle again, but with a standard, compressor dorm fridge. That feeder has now been running reliably for about 6 months so I figured I'd post the concept here.

Allen

ReeferAl · 10/05/2008, 01:47 PM

Here is a schematic drawing I've done showing the components of the feeder,

I have a 1 gallon container that I fill with salt water each day and add about 8 small cubes of frozen food. About an hour after the lights go on in the morning the feeder begins dispensing food about every 15 or 20 minutes (I'm not sure of the current setting- it's adjustable). Each feeding occurs by opening a 3/8" ID solenoid valve (no 1) for about 15 sec. I found that the tubing or valve would periodically plug so I added a second solenoid (no 2) which is hooked directly to the RO line so it is under pressure. That valve opens for 1 sec at the start of each feed cycle. That amounts to just a small squirt of water but it has kept the line from plugging ever since. The feeding go on for 8 to 10 hrs each day before the jug is "empty". The timer then quits triggering about an hour before lights out. I'll post some photos later.

Allen

ReeferAl · 10/05/2008, 02:09 PM

Some more information: The quick disconnect has an integrated stop so it closes off when disconnected. That allows me to just remove the whole jug to fill or clean. The check flow valve prevents the RO water from refluxing into the jug. The control box contains the timers and the magnet end of the smaller solenoid.
The timers are from Home Tech. I used 2 of the model 6030, but it might be possible to use just 1 of model 6062, with the RO flush occurring as a 1 sec blast at the end.
I just refill the jug each morning for a total of 5 days then I clean the jug and fill with RO water to run through the 6th day to flush the system.

ReeferAl · 10/05/2008, 02:33 PM

This is a picture of the control box and solenoids outside the feeder:

ReeferAl · 10/05/2008, 02:43 PM

This is what the inside looks like with the door open:

I also had to remove the shelf from the interior of the door and replace it with a flat piece of aluminum to allow room for the gallon jug and tubing. Due to the angle you can't see where the airline enters and where the air intake is positioned. Both enter the fridge through the top and the air pump sits on top of the fridge. The whole thing is sitting on a shelf suspended from the ceiling to allow it to flow by gravity into the surge tank. I could have it sit a bit lower if it were flowing directly into the display tank instead.

01/31/2008, 06:23 PM	#2
ReefEnabler Premium Member Join Date: Oct 2007 Location: Cary, NC Posts: 3,760	wow, what a cool idea! have you been using it much? I might have to try that someday __________________ - Ryan B "that is enough skimmate to ruin lives." - GSMguy Current Tank Info: 220g Display, 70g sump, 35g frag, 50g fuge, 2x250w MH, 1x400w MH, 2x80w T5, 2x140w VHO Actinic

02/01/2008, 09:35 AM	#4
prugs Moving Out Join Date: May 2003 Location: San Tan Valley, AZ 85140 Posts: 4,898	How are you regulating the amount of food into the tank? Is it a drip system or solenoid system? __________________ A Picture Is Worth A Thousand Words. Current Tank Info: Getting rid of the last equipment :{

02/01/2008, 01:48 PM	#6
ReefEnabler Premium Member Join Date: Oct 2007 Location: Cary, NC Posts: 3,760	have you considered trying multiple solenoids run off some kind of mini-manifold? i wonder if you let out less food but spaced at different locations in the tank if you could feed more with less, so to speak. maybe like 1 outlet per powerhead or something to get the food everywhere quicker? or are you seeing that it diffuses pretty quickly throughout the tank as-is? __________________ - Ryan B "that is enough skimmate to ruin lives." - GSMguy Current Tank Info: 220g Display, 70g sump, 35g frag, 50g fuge, 2x250w MH, 1x400w MH, 2x80w T5, 2x140w VHO Actinic

02/03/2008, 12:05 AM	#8
"Umm, fish?" Premium Member Join Date: Jun 2005 Location: Boulder, CO Posts: 4,112	Very nice! I've been trying to figure out how to do this for years. How cold does the water stay? __________________ --Andy "And chase the frothy bubbles, / While the world is full of troubles. . . ." --W. B. Yeats Current Tank Info: ~480 gal. display, ~1,400 gal. system

02/03/2008, 06:28 PM	#10
ribs Registered Member Join Date: Jan 2002 Location: Marlton, NJ Posts: 1,599	Is that a 1 qt cooler and why so large.....so you can dissolve the foods in a larger volume of water? What volume of water to frozen foods do you tend to use so far? __________________ Delaware Valley Reef Club - PA NJ DE area.

01/31/2008, 06:15 PM	#1
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	Semi-automatic frozen food dispenser Well. I've been fooling around again with some ideas I've had for making a device to dispense (thawed) frozen food. The result is a bit impractical as it ended up costing about 3X what I intended, but it seems to work. My goal was to make a feeder that could:[list=1][]dispense small amounts of food throughout the day[]feed anything frozen up to about the size of mysis []keep the food fresh for an extended period of time- at least several days[/list=1] I decided that to do this I would need:[list=1][]a chilled container[]something to stir the food to keep it suspended[]a method for dosing that would not be affected by the particle size[/list=1] What I came up with is a bit more complicated than absolutely necessary, maybe even a bit "Rube Goldberg", but it does everything I wanted it to do. This is unit disassembled: The picture doesn't include it, but there is also a solenoid valve that can be opened for a set amount of time (currently about 5 sec) as often and whenever I want. It is controlled by the Aquacontroller and is set to open every 12 min throughout the day. This feeds about 1 gal of food suspension each day. This is the inside of the cooler showing the Ice Probe inside: And finally, here it is assembled: It sits above the level of the tank and feeds by gravity.

01/31/2008, 07:50 PM	#3
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	I've been using it for a couple of weeks now. I figure I'll empty and clean it every weekend. Otherwise I just add about 1 gal of saltwater with a bunch of food each morning. The water added is just about offset by the skimmate produced, although I still check salinity periodically and adjust as needed. I've been keeping an eye on nitrate and phosphate as this is an increase in the total amount of food added to the tank and so far they've been stable.

02/01/2008, 02:56 PM	#7
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	The outlet actually goes into my surge tank. When it surges it distributes the food somewhat via the 2 surge outlets. It does tend to concentrate it initially more at one end of the tank and the fish are always keeping an eye on the surge outlets. I could run it into the wavebox as that distributes it quite well, but I've found it tends to "gum-up" the propeller a bit. That was with larger food though so that may still be an option. I could also aim it at the closed loop intake. It would then be distributed all through the tank. I'm afraid that would chop up the larger pieces though. Those are options I may try in the future as either would be pretty easy to do. As far as varying the feed interval, the timer function could do that, but with the food going into the surge tank and then into the DT it is varied somewhat anyway. In fact, it results in there being some food (like cyclopeeze) in the water almost all the time. My intention is to eventually add a school of anthias. Allen

02/03/2008, 12:09 PM	#9
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	The water stays around 35 to 40 deg. If I hadn't already gotten the Ice Probe for another project a couple of years earlier I would have thought about leaving out the chiller and just adding ice each morning along with the food. That would add a little more time to the process every day but would also save about $100. The way I built this it cost me about $300. It might also be possible to use an air pump to keep the food stirred, but to do that I would want to find a way to have the air pump recirculate air from the cooler rather than always using warm outside air. That would cost a bit less than using a motor for stirring. The other major component was the solenoid and timer. I'm not sure how to trim any expense from that. Allen

02/03/2008, 08:03 PM	#11
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	The cooler is a 2 gal cooler. The larger the volume the easier it is to regulate the amount dosed. For example, I am dosing 5 times per hour (every 12 minutes). The dosing goes on about 10 hrs per day for 50 times daily. I aim to dose about 1 gal per day. That means that each dose is about 2.5 oz of mixture. If I were only dosing a quart it would mean each dose is only about 2/3 of an oz. It's difficult to dose such a small amount very accurately. Additionally, at the end of the day about half is still left, which keeps the flow more consistent than if the jug were almost empty. Since I'm using a chiller it's also good to keep some liquid in the jug because once it drops below the level of the chiller probe it isn't kept chilled any more. I'm not saying this is the only way or even the "right" way to do it, but those were my reasons for the way I did it. Allen

02/03/2008, 08:09 PM	#12
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	Oh, another thing- If you go back to my first post you'll see that I had 3 criteria that I wanted to meet with this project. If the only criterion had been: #1 dispense small amounts of food throughout the day and I were only feeding something very small like cyclopeeze then it would be alot simpler. The stirring would be easier and there would be more choices of how to dispense. I had tried using a peristaltic pump initially. It worked fine with just cyclopeeze, but couldn't handle mysis which is why I went with the solenoid. With just cyclopeeze the solenoid valve could also be smaller if you went that route. Since most of my fish are larger though I didn't feel that cyclopeeze alone would be worth the effort. Allen

02/03/2008, 10:31 PM	#13
"Umm, fish?" Premium Member Join Date: Jun 2005 Location: Boulder, CO Posts: 4,112	Actually, you've stirred my imagination. What about adding RO/DI ice and frozen food to a blender, chopping the ice, then adding the mess to (essentially) an insulated cone and allowing it to simply gravity feed to the tank as it melts? It'd be hard to standardize the dose, but is that so necessary? __________________ --Andy "And chase the frothy bubbles, / While the world is full of troubles. . . ." --W. B. Yeats Current Tank Info: ~480 gal. display, ~1,400 gal. system

02/04/2008, 08:39 AM	#14
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	When you say "gravity feed", do you mean controlling the amount by using a pinch device to adjust the flow or a solenoid? The reason I use a solenoid is that the pinch valve causes a dam that allows only the liquid and smallest particles to pass and holds back the larger particles until the tubing plugs. Not stirring the mixture will mean that the sinking particles will feed first, followed by those that are suspended and finally those that are floating. Many of the floating particles may also end up coating the inside of the container as it slowly empties. As far as using RO/DI ice and a blender, I think that would be fine. You could also add cold or frozen saltwater. I've found that my skimmer removes at least as much water everyday as I add via the feeder so the feeder offsets the skimmer losses somewhat. Allen

02/04/2008, 08:56 AM	#15
JaredWaites Premium Member Join Date: Sep 2007 Location: Destin, FL Posts: 1,482	This is a great idea...but I found something you might be interested in looking at I'll attach it in my next post damnit I can't edit this post and post the attachment...that sucks.

02/04/2008, 09:05 AM	#16
JaredWaites Premium Member Join Date: Sep 2007 Location: Destin, FL Posts: 1,482	http://www.audet.biz/499finalreport.doc There ya go. It won't let me upload it as an attachment as its too big.

02/04/2008, 09:29 AM	#17
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	Very cool concept. I considered trying to build a feeder using the same concept with a screw device and thermoelectric chiller but decided it was beyond my skill level. It looks like this was a college engineering project so it may never actually be developed, although it would be nice if it were. Still, it may be a few hears until it hits the market. Right now I am dosing about 8 cubes of food per day so I would still need to refill it everyday. Allen

02/04/2008, 10:07 AM	#18
"Umm, fish?" Premium Member Join Date: Jun 2005 Location: Boulder, CO Posts: 4,112	Actually, I wasn't thinking about pinching the drain at all, just letting it flow at the rate of the ice melting. So, maybe controllable with the amount of insulation you use. Hmm. Good point about the food coating the sides. The reason I thought RO/DI ice was that you'd have to keep it so much colder to keep saltwater iced. For what it's worth.... __________________ --Andy "And chase the frothy bubbles, / While the world is full of troubles. . . ." --W. B. Yeats Current Tank Info: ~480 gal. display, ~1,400 gal. system

02/04/2008, 11:17 AM	#19
funman1 Registered Member Join Date: Jan 2006 Location: Roseville, CA Posts: 4,898	The Aqualifter will pump air just fine, and it's got an input and an output hose connections. And the best part is it's only like $14.00 __________________ Professional Pyro: NO REALLY; I get PAID to blow stuff up!! ~Steve~ Current Tank Info: 12G NANO

10/05/2008, 01:38 PM	#21
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	It's been a long time since I posted here. I've deleted all the old pictures I'm afraid. The Ice Probe chiller failed within about 2 weeks. I came up with a new design based on a small refrigerator. I bought a "dorm fridge" from Best Buy that was marked way down due to a damaged case- nothing that would affect the function. It was also based on a thermoelectric chiller. It would only keep the fridge interior at 40 to 45 deg and it too failed within a couple of weeks. Interestingly, there was nothing about the way I was using it that should have caused a premature failure. It appears that thermoelectric chilling has a long way to go to be reliable. I used the same principle again, but with a standard, compressor dorm fridge. That feeder has now been running reliably for about 6 months so I figured I'd post the concept here. Allen

10/05/2008, 01:47 PM	#22
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	Here is a schematic drawing I've done showing the components of the feeder, I have a 1 gallon container that I fill with salt water each day and add about 8 small cubes of frozen food. About an hour after the lights go on in the morning the feeder begins dispensing food about every 15 or 20 minutes (I'm not sure of the current setting- it's adjustable). Each feeding occurs by opening a 3/8" ID solenoid valve (no 1) for about 15 sec. I found that the tubing or valve would periodically plug so I added a second solenoid (no 2) which is hooked directly to the RO line so it is under pressure. That valve opens for 1 sec at the start of each feed cycle. That amounts to just a small squirt of water but it has kept the line from plugging ever since. The feeding go on for 8 to 10 hrs each day before the jug is "empty". The timer then quits triggering about an hour before lights out. I'll post some photos later. Allen

10/05/2008, 02:09 PM	#23
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	Some more information: The quick disconnect has an integrated stop so it closes off when disconnected. That allows me to just remove the whole jug to fill or clean. The check flow valve prevents the RO water from refluxing into the jug. The control box contains the timers and the magnet end of the smaller solenoid. The timers are from Home Tech. I used 2 of the model 6030, but it might be possible to use just 1 of model 6062, with the RO flush occurring as a 1 sec blast at the end. I just refill the jug each morning for a total of 5 days then I clean the jug and fill with RO water to run through the 6th day to flush the system.

10/05/2008, 02:33 PM	#24
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	This is a picture of the control box and solenoids outside the feeder:

10/05/2008, 02:43 PM	#25
ReeferAl Premium Member Join Date: Dec 2001 Location: Fond du Lac, WI Posts: 2,040	This is what the inside looks like with the door open: I also had to remove the shelf from the interior of the door and replace it with a flat piece of aluminum to allow room for the gallon jug and tubing. Due to the angle you can't see where the airline enters and where the air intake is positioned. Both enter the fridge through the top and the air pump sits on top of the fridge. The whole thing is sitting on a shelf suspended from the ceiling to allow it to flow by gravity into the surge tank. I could have it sit a bit lower if it were flowing directly into the display tank instead.