Backyard Aquaponics
A place to discuss aquaponics

I'm having another crack at a demand feeder. The fish press a lever and get fed. The device regulates the maximum amount of feed the fish are allowed.

I'm having another go at my demand feeder, and might add a few other ideas.

I just couldn't figure out what some of the components did on the first one. (I blame prescription opiates during the construction - other might just blame my head)

It's hard to fix something when you don't know what it does

So I've started again, and gone for a 20 pin chip instead of the previous 14 pin version.

I thought I might drop the design in here as it progresses, in the hope that someone who knows stuff might keep an eye on me and see if what I'm designing will work. I check and double check everything, but I know so little about electronics that sometimes I get confused by, for example having two power sources that share the same "-" terminal. Clearly electricity works nothing like plumbing, but rather a lot more like magic.

So, if anyone or everyone is willing, can you let me know if the design works as I post it.

The general idea, is there will be a main board with the chip on it, and a second board (above all the XXXXXXXXXs) that will hang down under the main board at 90 degrees to hold switches and adjustable things. They will be accessed by removing the PVC cap.

All things like led's and motor connections will be on plugs so I can slide out the board from its 90mm storm drain PVC housing to reprogram it as needed.

The main board will connect to the front board by either pluggable pins or those L shaped wire bridge things that are soldered in place.

The back of the main board will probably have pins that connect to the motor that drives the feed auger. Those pins will also serve to secure the main board.

The main board will slide in and out like a drawer.

Before I start blowing up chips, I'll design it on Pebble, and hopefully get some feedback.

Here is a screen grab of the first stage.

The "X"s are cuts in the strip board, and the top line of them represent the separation between the main board, and the front board. The Black wire on the right indecates the right hand limit of the board size that will fit into the PVC.

Power connection, and serial connection (baby steps ) The serial data connection info can be found at the PICAXE website here in this pdf on page 8, but I dont think that will be an issue.

My main concern is the birds-nest of stuff hanging off the "-" terminal. I have no idea if that all works, because I dont believe in electricity.




All feedback/derisive laughter welcome, and if you know about this stuff please hang around and help me as I go.

I'll drop the code in as a text file at the top of this thread once its complete.

Go hard bw. Hope it works out.

Depending on how many amps your motor draws, you may need to use a darlington array or relay to turn it off and on.

Huzzah . Ur having another go at it. il be watching bw . But I doubt ill be much help :p

Sent from my HTC Explorer A310b using Tapatalk 2

Good to see some proper electrickery spells being crafted :-)

I'd put the 470 Ohm resistor for the power LED on the other terminal of the switch, that way it goes red when you turn the switch on, rather than when there's power on the 5V plug. Also I think the 100k pull-down resistor for the switch should be a pull-up instead -- the switch grounds the signal wire, but so does the resistor.

Depending on your program, you may need to debounce the switch with a capacitor as well.

Regarding power supply, I like to use a 5V regulator on the board, that way I can give it anything over 6V and it will work (like a 9V battery or 12V plugpack). Also it's usually a very good idea to put a capacitor across the power lines.

That's if I'm reading the diagram correctly, I might be way off :-)

Feel free to laugh at my solar woes... I just spent a week trying to buy a USB->Serial converter because mine was "dead", I can't program my chips without it. Turns out to be a dud driver, I feel a bit silly now!

Re other terminal of the switch - yeah, it should go on when the switch is on, But I dont understand the other bit. I dont think there was a resistor for the switch. There are two resistors in the serial circuit...



but the resistor under the switch is for the power LED

I think



I think I'll do the debounce* with the code. In the last version I had an "on switch count to 1000" line that made it be only seen once, and it also only triggered when the switch was released so if something was stuck up against it (say a dead fish) it wouldn't keep getting false hits. I didnt know you could do it with a capacitor.

Psudo code was something like...



on pin1 high

count to 1000

add some stuff to variables to note the event, and reset time to next feed etc

loop this line until pin1 is low (button released)



re:5V regulator - yeah wise, thanks.


And there is a capacitor just before the power pins on the chip. Perhaps I left it out of the diagram, but it's meant to be there. It was on the first version, so perhaps I dreamt it.

That first version is still a bit of a blur.


How much battery would be wasted by the 5 v regulator?





*For anyone reading that's wondering what debounce is all about... a switch doesn't open and close in one neat click. Up close it might be a series of minor contacts or even a bounce leading to the chip seeing it as many, rather than just one event. In the code you create a pause after each time the switch is activated so it ignores any extra hits.

The original version had a transistor looking thing so I could use 12 volts, so thanks, I'll try to remember it. I'm trying to avoid looking at the original because it keeps confusing me

And I think I've just noticed a missing cut in the stripboard under the second to top resistor. I think that resistor does nothing as it is.

Also, would anyone who can draw circuits be willing to contribute a circuit diagram as this takes shape?

It will all be open source, software and the PVC and motor bits, so it's for the good of the people

I cant figure out how to use any of the design tools, and they drive me nuts when they dont have the components I need. One program I tried has all the parts listed by their catalog number. I just want to pick a drawing of a resistor from a list and stick it on the the circuit. AARRgghhh. It listed parts by manufacturer, then serial number. Must be for mass producing stuff, but it drove me crazy.

I think I might get rid of the power led from where it's connected now, and run if from a pin. The picaxe chip has a known voltage stored in a variable so you can compare voltages to it. I think that means I can make the power led flash when the batter is low (If it's run on battteries - I want the batter option in case someone else makes one where mains power is inconvenient. One of system plans involved having a pump and sump next to my power source, and the growbeds and fishtank miles away)

So I think I poll the internal referance voltage, and compare that )ADC?) with the voltage at the + pin. This will take an extra pin, but will make sure there is some warning of a battery going flat.

Also I think I need to drive not just the chip with a 5volt regulator, but also the motor sith something similar. I'll use pulse width modulated method of controlling the motor, and also time. ie the motor will run a (user set) time each feed to deliver the correct dose. But as the Motor driver power supply gets low, the motor will turn less often in a given number of seconds, so the fish wont get quite as much food.

The 12 v motor runs well enough off the 5v batter supply for the chip.

Is there any problem running a 12 volt motor at 5 v?

Now I've decided to leave the power led where it is, set to not very bright, and then add another very bright one that flashes if the batteries get too low.

Ah I thought the serial port was the switch for the fish. My mistake, carry on, nothing to see here.

Cool.

Please keep the feedback combing though.

I really wont be able to do this on my own