Rotifer Tower

[herring_fish]

My robotic feeder is not a technical big deal. It’s low hanging fruit. If, after reading a few tutorials, people could tell me what motors to buy, how to hook them up and give me some example started code, …then in the word of the famous philosopher, Jay Leno, ”This isn’t rocket surgery.”

Technically, a tank glass cleaner would not be that hard to do. It is basically turning a standard motor on and off, along with a slide somewhat like what I have, that would move the brush assembly across that tank and back. It might be pretty ugly though.

While you do need to know or learn to program, an Arduino controller with various add-on boards can do stuff like that with ease. It can turn things on and off at particular times. It can run things at desired speeds, even if the speeds need to vary. It can monitor multiple sensors, like anything from optical limits switches to pH and temperature probes. They can respond to pushed buttons, time events or sensor input by triggering pre-set actions or groups of actions.

Building the hardware while keeping a full time job, having a family and other hobbies, is the hard part.

 

[dave willmore]

I know what you mean, families take all the time we've got. That's why the combination of automated tank ceaning with automated food production appeals to me. I love arduinos but perhaps an automated router could do the same job without the programming.

If an arduino is used it seems like your track arrangement would still be necessary, right? One track for the brush spinner to travel, another track to move the motor to the next cleaning area.

 

[herring_fish]

Yes, you have it.

 

[herring_fish]

These are links to my progress so far. I am about to put it on the wall.

http://www.youtube.com/watch?v=zBPnccwZjO4

http://www.youtube.com/watch?v=VMpPz_xaEC0

http://www.youtube.com/watch?v=dK4hXVRhH2k

 

[dave willmore]

Your videos look good and it looks like it will do what you want it to. When will you set it into operation?

 

[dave willmore]

How is it working?

 

[herring_fish]

I really don't know. I took the week before Christmas, Christmas week and the following week off. I thought that I would spend most of that time working on it but the Arduino board died the first day. I understand that is very rare but it happened to me. I ordered another one but it turned out to be an Arduino knock-off. It was dead on arrival. What bad luck.

I am very close but things move very slowly for me. I have lots of other stuff that I have to do.

I put two 4' by 8' plastic sheets up in the garage to keep splashes, spills and spray off of the walls. I drilled the main rail so I can put it up this weekend. Then I have to add the races for the two wire bundles. Finally, there will be lots of details, some expected some not.

I have also been following your thread. When are we going to see some pictures?

 

[herring_fish]

FINALLY!
I finally have the robot up on the wall. It doesn't look like much does it. It took me 4 days to put in 4 screws. Everything that could go wrong did. Now it's on to the raceways for the wires.

Maybe, if I'm lucky, I should be done with this project in, somewhere between 2 weeks and 5 months.

 

[dave willmore]

I am very close but things move very slowly for me. I have lots of other stuff that I have to do. QUOTE]

Sorry to hear about the arduinos. I know what you mean about delays, my tank build has been going on a year and it keeps dragging out. And that's just the tank, the custom equipment hasn't started yet.

Sometimes I think I work full time picking up after the kids and have no time left over for work.

 

[dave willmore]

Your setup looks great. Don't worry about the time, original work always happens that way.

 

[Northside Reef]

What a fun thread. Thanks for posting it.

I know it's way to late for input but I wish I could have read this when it was new. I think you could have saved yourself a ton of programming if you wold have used proximity switches for each station. Want to skip a station? pull back a prox and let it find the next kind of thing.

These are the kinds of projects that look fun as heck once you've establshed your setup. I really look forward to learning more about how your system works. Again thank you.

 

[herring_fish]

I went on several web sites to get advice and that path was one that was suggested. I looked at using proximity switches and servo motors instead of a stepper motor. I am just more familiar with CNC machines and using either absolute or relative location assignment. I wanted my robot to be more like a 2 axis CNC machine. I just think that it's cool and a challenge.

I do have three optical switches. One at the far right for a home location, one on the far left to shut it all down incase thinks go haywire and one on the boom of the second axis so that I am sure that is all the way up before the carriage takes of to the next station.

For me absolute is the best way to do things if you have the accuracy. You just tell the system that you want to go to this point in space and it does it. Relative is the next best thing. That is, you tell the system to go so many steps to the left or right, up or down. I think that I could have gone absolute but I wasn't confident that I had all the accuracy that I needed. I went with relative positioning. That was the easiest for me and the way that I think.

With relative positioning I was able to simply write some code (that I got help with) that reads a text file at compile time. That file has d_val(s) for taking distance values. In other words, how far should the carriage move to get to the next station. I also have f_val(s) for food values and w_val(s) for water values. In other words the value is the number of second that the peristaltic pump gets turned on for and how long the water peristaltic pump gets turned on for.

#define d_val 11, 21, 31, 41, 51, 61
#define f_val 12, 22, 32, 42, 52, 62
#define w_val 13, 23, 33, 43, 53, 63

On power up, the program does a move to the left of about 3 inches and then does a go home and a calibration function before it moves to station one, by looking at the first d_val which in this case is 11 steps. Then it stops and lowers the boom. (pun intended) The pump for the food is turned on by looking at f_val which is 12 seconds and finally it looks at w_val and turns on the water pump for 12 seconds. Then it looks at the second d_val to go 21 steps to station two and so on. After station six, it goes home and turns off.

This way I just edit the text file. I can go to locations that might be non-standard. Perhaps I want to bunch a couple close together for some reason and move others further apart. I can fine tune locations to the step position that I want. There was a little more programming but less switches and wiring, which I am bad at. It also gives me more freedom and ease of use in the long term"¦..I think.

 

[Northside Reef]

Hello Herring yes I'm pretty familiar with CNC controls I design and build 3 axis robots used for polyurethane dispensing. So I get the love of the machines.

I was just saying a proximity switch and a simple dc motor could have done that job since the target area for the drop was so large and prox switches are pretty darn accurate considering they're used in the homing sequence. so home position needs to be pretty close to exact each time you home the machine or your programs won't be of much value to you.

But trust me I totally understand the desire to incorporate a little of what we know with our hobbies I plan to do my DYI lights using all PLC code lol. So I get it, it's just cool to add a little of ourselves to these projects.

PS: here is a pic of the one of my babies before shipment.

 

[herring_fish]

I figured that you knew what you were doing since you were offering programming help. You can see by my explanation that I had to go in a weird direction to get the effect that I wanted. I didn't want to get into a CNC module even if it was free because I wanted to understand everything in the code. I over explained so that others could see what was I doing.

So what is that wheel? It looks kind of like the one that I used.

 

[Northside Reef]

it's a fan sitting behind the machine. but it does look like it is attached from the photo. Trick photography I tell ya

 

[herring_fish]

Again, since I came up with a liquid switching unit that would be a simpler fit for this project, this is now more for fun, the challenge and for the cool factor.

I had a simple way of putting the wiring for controlling the 2nd axis carriage by putting the two cables in separate inexpensive channels so that they didn't fall over or get tangled. Although I have been around E-chain for years on big machines, once I saw Northside Reef's pictures of his latest smaller CNC project, I had to look again at the prices of that stuff because it looks so cool.

I changed my strategy, back to using a standard cable cage track once I found it on Ebay under drag chain instead of E-Chain. I found it much cheaper than I was looking at before so I ordered it today. I'll have to wait another 5 to 12 business days for that.

I will try to do some of the check out work in the mean time, now that I have a REAL Arduino microcontroller that works.

 

[dave willmore]

Asa,

I'm not trying to sidetrack your great thread but I wonder if your track and roller idea could be adapted to a different type of automated plankton harvest.

I have in mind a trough full of rotifers, into which phyto is dripped. A 53 micron screen is lowered into the middle of the trough, then a pulley moves it to one end, concentrating half the rotifers. At the end of the trough is a slotted pipe on a solenoid that opens for a few seconds and lets a few gallons of concentrated rotifers into the display tank. Then the screen is lifted out of the water and moved back to the middle of the trough to wait 12 hours or so for the rotifer population to recover.

The screen would need to be moved a foot or two vertically and maybe 3 or 4 feet horizontally. Do you think this would be easy or hard? I don't know enough about programming to know if there is any original work involved here, or if something else could be easily adapted.

Please give an update on your system, it is fascinating.

 

[herring_fish]

Speaking from the programming side only, you are basically talking about a two axis machine. One axis moves the screen to the end and back. A second axis would lift and/or rotate the net out of the way.

If someone had done multi-axis work before, it would not be very hard. What I have is the same except that you probably need one or two less sensors. My program could be pruned down from six stations to one and the motion feed rates would be simply lowered.

As for my project, I burned out the controller for the second axis by hooking it up wrong. I got another one in and I am back to final assembly now. I re-built the cable for the third time to take advantage of the new e-chain so that I will get a smaller bend radius. Now I'm testing both axes at the same time.

Sorry for being so slow.

 

[dave willmore]

Original work is never fast or easy. You are doing it in half the time it would have taken others.

 

[ReefKeeper64]

How are things coming along with the project?