Automatic Litterbox Central

I am interested in assisting with this. Here is a sample of my current project (http://forums.adafruit.com/viewtopic.php?f=8&t=14763), ironically, dealing with my cats, and soon (this weekend) the catgenie (as detailed in the post). Had posted a general question for help on the schematic, and to a lesser extent, the code.

Interesting reading on your project. I do look forward to seeing/helping in any stage. I like to tinker with code..

GenieRefiller wrote:Currently, I have completed the schematics and tested functions for all of the motors, sensors, and LEDs on the processing unit. I am now working on detecting/debouncing the keyboard and the main structure of the program. I am working on this in my spare time. So, it is going slowly at best.

It's OK. Good luck.

Currently, I have completed the schematics and tested functions for all of the motors, sensors, and LEDs on the processing unit. I am now working on detecting/debouncing the keyboard and the main structure of the program. I am working on this in my spare time. So, it is going slowly at best.

Excellent,

I thought I was doing this alone...

I am going to attempt writing this in Mikrobasic.

Do you have the pinouts mapped out? Can you post them. I think using the stock controller is the best idea...

JT

Here's one for the "logic list"

Cat should be in the box for more then X seconds before triggering a wash cycle... My goodness! The (original) Cat mode is triggered by the slightest (200ms) motion... Mitchcat has triggered 4 or more wash cycles unnecessarily already in the few days it has been running. He was just snooping around - granted it's a new thing to check out, But surely a cat can't drop a deuce bomb in under 3 seconds...?

Jason

JasonT wrote:Cat should be in the box for more then X seconds before triggering a wash cycle... My goodness! The (original) Cat mode is triggered by the slightest (200ms) motion... Mitchcat has triggered 4 or more wash cycles unnecessarily already in the few days it has been running. He was just snooping around - granted it's a new thing to check out, But surely a cat can't drop a deuce bomb in under 3 seconds...?

That would not improve box operation. The sensor doesn't oversee the whole box. My cats often sit in a corner of the box, triggering the sensor only briefly while entering and leaving the box, buy still doing some dirty business in between.

What you mention is called 'debouncing' in software. It ensures that the sensor state is steady for a specified period before it is considered to be valid, 3 seconds in your case. However, it is not possible to distinguish the differ a cat jumping in briefly for fun and a cat moving around in the box. Trust me, it did some measurements on this and I have never seen a consecutive detection time of over 1 second.

The only solution to your problem is a scoop-only function, 5 minutes after each use:
http://www.youtube.com/watch?v=Yx5zOR2LaMY

I have noticed me smaller cat is able to use the Cat Genie and make it out without setting of the sensor.... I am not sure how she does it, but it happens every now and again... I picture her using it like Tom Cruise in Mission Impossible with the lasers!

CatGenius wrote:a scoop-only function, 5 minutes after each use

Well on your way to new firmware I see...

Rather then using debounce, (problems I know all to well) Set a variable that increases with each blip of the IR and when that var hits a selectable threshold then trigger a "Cat Activation"

I was going to wait to tear into my 7 day old 120, But more and more of these little things are starting to bug me.

If I have time this weekend I will probably replace the PIC with a zif socket and start mapping out pins. With the RFID reader and serial port no longer needed there should be enough IO to get a LCD screen, and sonar going.

Jason

psu1994ca wrote:I have noticed me smaller cat is able to use the Cat Genie and make it out without setting of the sensor....

Funny you say this, I got home today and there was the #2 with no cat light flashing...

Let the hacking begin!

Jason

CatGenius,

That video link is very very interesting!!!

JasonT,

There's a hefty "secret market" just lurking about waiting for such a firmware or add-on device that will "Turbo Charge" their CG's!!!

I can't wait to see what you devise!

JasonT wrote:Rather then using debounce, (problems I know all to well) Set a variable that increases with each blip of the IR and when that var hits a selectable threshold then trigger a "Cat Activation"

Hm, perhaps moving average will improve things here indeed. I will consider it, but only if I have memory left. Currently the box is fully operational and I'm using about 67% of code memory.
Now I'm beta testing and implementing copy protection: I will need to distribute this firmware as a binary, and I don't need this project to make me rich, but I wouldn't like to see other people exploiting my labor. The copy protection it notching up memory usage to 87% and I still need some headroom for improvements and bug fixes.

JasonT wrote:If I have time this weekend I will probably replace the PIC with a zif socket and start mapping out pins. With the RFID reader and serial port no longer needed there should be enough IO to get a LCD screen, and sonar going.

There's no need for a ZIF socket. You can reprogram the PIC on the original board. Even an ISP connector is mounted, but you will loose the original firmware with no way to back it up. I would be careful with these modification on a box that young because some boxes have bad scooper motors and you're not likely to get it replaced if you've tinkered with the box.
There's no use for sonar: The existing GenieEyes are very clever and hard to beat. There's more to them than meets the eye (pun intended). An LCD screen would be useful, but hard to install both electrically and mechanically. The serial port is available, but not used on the original box. If I have memory left, my firmware will offer an small interface there so advanced users can attach a serial-to-Ethernet converter to Telnet their way in.
Don't think this is something you can do in one or two weekends. I've been working on it in my spare time for 6 weeks now and I think I've still got a couple of weeks to go. But it's fun and I think it will make other people happy too. If you're in it for the fun and you're persistent, I'm sure you can do the same. But if you just want a solution for you problem you can get my firmware, at a very affordable price, and probably sooner than you can make something stable yourself

flyguyjake wrote:That video link is very very interesting!!!

Well thank you FlyGuy! This weekend I will post videos of the other menu features, a full washing cycle and a scoop-only cycle.
(BTW, I'm not a genius: The project and firmware is called CatGenius)

CatGenius wrote:There's no need for a ZIF socket. You can reprogram the PIC on the original board. Even an ISP connector is mounted, but you will loose the original firmware with no way to back it up.

Right, the Zif would alow me to take my time without torching the box

CatGenius wrote:you can get my firmware, at a very affordable price, and probably sooner than you can make something stable yourself

yes, if the price is right.

Jason

phil1001 wrote:Is it possible to put the Arduino board with a LCD backlit display into a wall mounted project box and attach this to the main CG board with a 1 meter shielded harness such a old serial data cable ?
This would mean disabling or removing the CG PIC

I would like a simple level indicator for the sani solution
A scoop mode only
realtime timer to set full cleaning cycle
display for real error messages
linking to home automation

The CatGenie had several features I really liked, namely the scooping action, grinding and flushing.
What I didn't like was the long (almost 30 min) cleaning and drying time. Plus the large amount of odor and moisture it dumped into my home.
If I was to modify the CG I would make the bowl solid and use flushable clumping litter.And only use the the water for grinding and flushing. This would hardly add to the expense of use and would cut the cleaning time down to just a few minutes.

Please comment.

Does anyone have any schematics they have created available to send me or post online?

I am still into replacing the current PIC for an Arduino as I figure if you have to strip the unit apart you might as well create a complete open source version and install a new daughter board instead of just reprogramming the current PIC especially on the CG60

This will potentially also enable WiFi, LCD display and other features that just reprogramming wont overcome

Thanks

KE5AFU, is the writeboard open? Still waiting....