I need to time an AC outlet by seconds, or a break in a DC line

[Madsalts]

PROBLEM:


I am about to lose my mind. I am trying to accomplish what should be a really simple project.


I am trying to automate feeding in an aquarium. For various reasons, commercially available feeders won't work. From a mechanical perspective, I can build one. Timing one is quite difficult, though.


Essentially, I need filters to turn off while a feeder turns on. The feeder will then turn off (so, for a period of time, the feeder and filters will be off so that uneaten food won’t be sucked into the filters). At this point, the filter will turn back on. This then repeats over and over.


The filter is commercially made, and the feeder will be either AC powered with a wall wart, or DC powered to 3V.


So far, I've thought that the easiest way to do this would be to use the “Woods 50015” timer for the filter. It can turn on and off up to 20 times per day. Also, I can set these on/off periods to correspond with the actual time of day. The big problem is that I cannot set this in increments of less than 1 minute. This is not an issue for the filter, but it is an issue for the feeder.


For the feeder, I am thinking of using a timer designed for a hydroponic growing system (Titan Controlls Apollo 15). The timer will go on for a period of time, and then off. The periods of time for which it can be off are in 5 sec. increments, then in minute increments, then in hour increments. I.e., it can be set to be on for 15 seconds. It cannot, however, be set to be on for 1 min and 15 sec. Times are EXCLUSIVELY in seconds, minutes, or hours. A cycle of being on combined with a cycle of being off constitutes one “period”.


So, if period a feeder period is comprised of 15 sec. on and 5 min off, the period is 5 min + 15 sec. As time goes on, this can’t correlate to any sort of chronological time on the other feeder unless I have a period which is a factor of 24 hrs. It is highly unlikely that it will cork out this way.


POSSIBLE SOLUTIONS:


1. It seems that some sort of control company might offer a commercial solution to this problem. I wouldn’t want to pay more than $150, and preferably half that. Despite looking, I haven’t found a commercial solution. Suggestions would be welcome.

2. I can use a wall wart and AC power for the feeder, and use a more appropriate timer, but I know of none.

3. I can use a DC timer on the power once it leaves the wall wart. This seems like it might be most promising in terms of either having a timer built or finding something commercially available. Again, suggestions are welcome.


Thanks.

 

[KrisBlueNZ]

PROBLEM: I am about to lose my mind.

Don't be afraid. It only happens once. Then you'll never go back. Bwahahahahaaaaaaa!

Essentially, I need filters to turn off while a feeder turns on. The feeder will then turn off (so, for a period of time, the feeder and filters will be off so that uneaten food won’t be sucked into the filters). At this point, the filter will turn back on. This then repeats over and over.

OK. What kind of time periods are we talking about?

Do you have (or can you get) a timer with a changeover output? In other words, it plugs into the mains supply and has two sockets on it, one for the filter and one for the feeder. Then all you need is a circuit in the low voltage path to the feeder motor that turns it OFF after it has been ON for a certain time.

The filter is commercially made, and the feeder will be either AC powered with a wall wart, or DC powered to 3V.

Right, that's an advantage. It's easy enough to make a delay circuit that's powered from low voltage DC (e.g. 5V, 6V or 12V; even as low as 3V) that can directly switch the current to a low-voltage DC motor.

3. I can use a DC timer on the power once it leaves the wall wart. This seems like it might be most promising in terms of either having a timer built or finding something commercially available. Again, suggestions are welcome.

Yes, I think that's the simplest option.

So your feeder has a 3V motor? How much current does it draw, running and stalled? And you could get a 6V DC wall wart? And you want the feeder to run for how long? Sounds like a job for a 555 and an emitter follower.

 

[Bluejets]

Easy with an Arduino pro mini and a couple of relays, especially if the timing turns out to be hours.

 

[KrisBlueNZ]

Easy with an Arduino pro mini and a couple of relays, especially if the timing turns out to be hours.

Yes, and it's also easy to crack a walnut using a pneumatic drill... If you have one, and know how to operate it!

 

[Arouse1973]

Yes, and it's also easy to crack a walnut using a pneumatic drill... If you have one, and know how to operate it!

Yes and people new to electronics don't suddenly become programmers over night, if at all.

 

[Bluejets]

I see your point......

 

[KrisBlueNZ]

"If all you have is a hammer, everything looks like a nail" (Not exactly relevant, but I like it.)

 

[(*steve*)]

Yes, and it's also easy to crack a walnut using a pneumatic drill... If you have one, and know how to operate it!

But consider the skill required to get the walnut kernel out intact!

 

[hevans1944]

What is the "ON" time duration of the feeder? How often does the feeder operate? Do you want to feed at particular times each day, or only a few times per week? How about a cheap web-cam like camera to inspect the water for food not consumed, and then with pneumatic-drill software detect and act on this? Do you have any electrical or electronic construction experience? Seventy-five bucks for a solution seems reasonable if you build (and possibly program) it yourself, as @Bluejets suggested. That's probably how I would do it.

 

[Madsalts]

The "on" duration would be anywhere from a few seconds to perhaps three minutes (multiple times/day). I don't know how a webcam would solve the problem, unless a program analyzed light transmittance through the water (telling me when the food has been cleared) is used. This doesn't address the problem of how much to initially dispense, though.

This is beside the point, though. I want to keep this as simple as possible. I'm skilled in building things like this, but lack knowledge regarding electronics.

By the way, I changed the design to use 6V, not 3V.

The simplest solution I can think of is to have the feeder (which will incorporate two motors, one for dispensation and one for "shaking" the hopper to help food fall toward the screw) turn "on" when the transformer is energized via a wall timer. Since the wall timer will only be able to be set for minute increments, the feeder will have to turn "off" prior to being de-energized. A digital dial of some kind (would an optical encoder to optimal?) could be turned. The position of the dial would correspond to the amount of time that the feeder would remain active once energized.

One way this might work, and I don't know if this is practical, is to have the system turn "off" after "x" number of pulses received from the dial. The frequency of pulses would be determined by dial position. I.e., if the system shuts off after 10 pulses, and each pulse is 20 sec. apart, the system shuts off after 200 sec. If the pulses are 1/10 sec apart, the system shuts off after 1 sec.

Any thoughts on this?

 

[Gryd3]

The "on" duration would be anywhere from a few seconds to perhaps three minutes (multiple times/day). I don't know how a webcam would solve the problem, unless a program analyzed light transmittance through the water (telling me when the food has been cleared) is used. This doesn't address the problem of how much to initially dispense, though.

This is beside the point, though. I want to keep this as simple as possible. I'm skilled in building things like this, but lack knowledge regarding electronics.

By the way, I changed the design to use 6V, not 3V.

The simplest solution I can think of is to have the feeder (which will incorporate two motors, one for dispensation and one for "shaking" the hopper to help food fall toward the screw) turn "on" when the transformer is energized via a wall timer. Since the wall timer will only be able to be set for minute increments, the feeder will have to turn "off" prior to being de-energized. A digital dial of some kind (would an optical encoder to optimal?) could be turned. The position of the dial would correspond to the amount of time that the feeder would remain active once energized.

One way this might work, and I don't know if this is practical, is to have the system turn "off" after "x" number of pulses received from the dial. The frequency of pulses would be determined by dial position. I.e., if the system shuts off after 10 pulses, and each pulse is 20 sec. apart, the system shuts off after 200 sec. If the pulses are 1/10 sec apart, the system shuts off after 1 sec.

Any thoughts on this?

You could build a gravity fed system and actuate some kind of gate to cutoff the supply.
Once the gate opens, the required amount of food will dispense and no more.
The gate would be optional if you think the supply needs to be limited.

 

[hevans1944]

The "on" duration would be anywhere from a few seconds to perhaps three minutes (multiple times/day). I don't know how a webcam would solve the problem, unless a program analyzed light transmittance through the water (telling me when the food has been cleared) is used. This doesn't address the problem of how much to initially dispense, though. ... <snip> ...

This is beside the point, though. I want to keep this as simple as possible. I'm skilled in building things like this, but lack knowledge regarding electronics.


Any thoughts on this?

It has been awhile since I have done anything with video camera signal processing, or fish aquariums for that matter, but I was trying to think "outside the box" on your problem. As I understand it, you want to dispense fish food periodically but not over-dispense. You also do not want the aquarium water re-circulation/aeration filter system to operate if there is unconsumed food in the water or, more likely, unconsumed food floating on the surface of the water. IIRC unconsumed food eventually gets water-logged and sinks to the bottom, which along with fish feces has to be cleaned out regularly... a completely different problem.

I tried to think how a human might do this. For example, I instruct my young son or daughter:

"Take care of the aquarium every day and I will reward you with room and board and perhaps other favors, such as an allowance. Several times each day, inspect the tank to see if there is any food still floating at the top. If not, turn off the filter system and add some quantity of fish food. Observe that starving fish rush to the surface to eat the food. After a few minutes look to see if there is any fish food remaining on the surface. If not, turn on the filter system and schedule another inspection some time later. If fish-food is still present, wait an interval to see if the fish are still consuming the remaining food. If fish are not eating the food, use a skim net to remove the un-eaten food and make a note to self to add less food the next time. Check for dead fish and remove same. Turn the filter system back on. Let me know if you find any dead fish."

Surely, I thought, since my sons and daughters have all grown up and left home, and my wife would not look favorably on my suggesting she take care of the aquarium (which may be why we no longer have an aquarium), there must be some way to automate the fish-feeding process. Perhaps if I point a video camera, focused on the surface of the water in the tank, I will be able to see if the fish need feeding. Sure, this will require some image processing software to detect food-on-the-water versus a clean surface, but how hard could that be? Maybe I will purchase a small aquarium and a few gold fish to observe with a web cam to find out. It's unfortunate I have some competence with electronics, not so much with software.