Automatic Litterbox Central

Has anyone tried this?

I have the original LR that worked fine for almost 2 years. It appears as if the hall effect sensor (there are actually two back-to-back if you pull them out) became less sensitive recently, as the globe will rotate past the dump position (about 90° too far), stop, think, rotate back, stop approximately in the right place but not quite. It is still usable in this condition but I am reluctant to rely on it for more than a couple of days - it could stop in an unusable position.

Here is what I was able to figure out:

The flat area right next to the exposed gear (on the base) has a dual hall effect (magnetic field detector) underneath it. The magnets are of course in the globe's track (easy to locate with a magnet). The sensors look just smaller than a TO-92 3-pin package (typical small transistor), and are not soldered to wires at all - their pins are instead inserted directly into a female harness connector, then the whole thing is shrink-wrapped (in clear shrink-wrap, helpfully). I am assuming that there are two sensors for redundancy (or false-alarm detection?)

My sensors are not dead. They are weak. If I take a small magnet (Sonicaire brush head magnet), the sensor works at a distance of about 1/3 inch. If I slide that magnet on the top of the base above the sensor, it works. The same magnet definitely finds the magnet in the track, but clearly the toothbrush magnet is stronger since the globe does not trip the sensors. My assumption is also that it must be the sensor and not the magnet since there are two in the track, and neither one trips the sensor. (Why would both permanent magnets suddenly get weaker?)

The sensors seem to have limited writing on them, from what I can tell they are labeled with a "R"-like logo and the letters "01E" (or "o1E"?). I failed to find a manufacturer with that logo so far, but I figured that hall effect sensors should be reasonably standard, right?

Any EE's out there (or manufacturer employees willing to help) willing to comment on my selections of possible alternatives?

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=US5781EUA-ND

Based on my marginal understanding of magnetic fields, a "15mT Trip, 3.5mT Release" seems about right - a small bar magnet is supposed to be about 0.01 mT, and the magnets in the track seem like they are stronger than that. The voltage tolerance is within my expectations of operating voltage too. I realize that there are more parameters to the operation of an IC, but the application is hardly high-precision or high-speed.

REASONS for doing this myself: I have major reservations about ordering a new base, for two reasons: I don't want to spend hundreds of dollars every 2 years, and I don't want another almost-fully-functional piece of hardly biodegradable equipment to go into the landfill if I can help it.

I screwed up the numbers a little (a small magnet, according to my reading, is 0.01T or 10mT). For some reason the most sensitive hall effect sensors I am finding start at 15mT. Maybe I don't understand the mechanics or measurement distances.

Also, I recognize that my other (and simpler) option is to install additional stronger magnets in the globe track. I can either replace the existing magnets or place a new one in the next spot (I'd have to cut out a little cave to insert one). My concern with this approach is that the sensors' condition could be degenerative.

One possibility I should probably consider is that the sensors are fine, the magnets are fine, but the control circuit is not (either not supplying adequate IC power or not registering a sufficient voltage change, assuming these are linear sensors). In which case, I'm out of luck with a DIY solution, I guess - so I'm ignoring this possibility

In short, if anyone has any thoughts, please contribute.

If I fail to fix this thing, I'm kind of considering the Cat Genie - I passed on it previously because I didn't have water facilities accessible where I want the box, but now I do, and it's not much more money.

Sounds like you have the mechanical/tech knowledge and explorer willingness for the Cat Genie. Wish I could help you with the magnet questions but I haven't a clue. Anyone??

If you buy a replacement globe or base, you can send the old one back to APCP where it will be disposed of properly.

My litterbox is now working more or less acceptably.

Just in case anyone else out there ever faces the same situation, here is what I learned:

• There are two magnets, one is used to stop when dumping waste, the other is used to go back to the home position.
• That second magnet is used twice! Once on the return trip towards home position to know to go another 30 degrees to level litter. Then on the final approach to stop.
• Magnet polarity is significant. The sensors only detect a magnetic field in one direction. My experiments with magnets proved this - only one orientation works
• In my case, the problem is either sensors or the circuit. Interestingly, the sensors operate properly when first powered on, and then start to get flaky in a few operations or a few hours.

What I actually did:

• Upgraded magnets. I cut rectangular holes directly over the existing magnets and epoxied in my little Sonicaire brush-head magnets. They are probably not any stronger, but they are closer to the sensors. Note that they must be positioned in the "repelled" orientation towards the existing magnet underneath (you have to tape them down until epoxy sets). Can't say that makes a lot of sense to me, but it doesn't work the other way
• I used regular plastic-welder epoxy, sets in 7 minutes. Nothing should be protruding or the globe will lift up and the magnets may be too far from the sensors at a crucial moment
• Doing the above made it work reliably when first plugged in. I figured I'm ok with that, I don't really want the thing to run as frequently as it does, it just wears itself out and makes a lot of noise.
• Put the litterbox on a digital programmable timer (about $12), on twice a day for 5 minutes, that should give it enough time.
• I assume that the litterbox won't run when plugged in while the pressure sensor is tripped, but I set the times for inactive cat periods.

On an unrelated note, in a properly functioning Litter-Robot I would prefer to have control over the cat activation delay. I'd probably set it to 4 hours or so. Maybe more. Of course this delay would have a different meaning then - it would be in addition to the 7 minutes, eg: "Wait 4 hours after first activation, then wait for a 7 minute period of inactivity as before". (I guess I'm just talking to myself here - sorry, I'm a programmer)

In my experience with my two cats, the litter is in nearly acceptable condition when operated just once in 24 hours (I mix Companion Scoop with a dash of tidy cats crystals). A delay of 4 hours is a compromise between caring about the cats and saving energy/litter robot/my ears.

On another unrelated note, there is a vulnerability in the litter robot sensor logic: Should the power go out while the box is leveling the litter (final stage), it will fail in the worst possible way. Once power is resumed, it will detect the magnet and think it has rotated fully to the dump position instead, reverse, and dump all of your litter, leaving the cats nothing and itself upside down.

akom wrote:

• Put the litterbox on a digital programmable timer (about $12), on twice a day for 5 minutes, that should give it enough time.
• I assume that the litterbox won't run when plugged in while the pressure sensor is tripped, but I set the times for inactive cat periods.

Unfortunately it will ignore the pressure sensor when the power is first turned on so it's a small safety hazard. The LR is only intended to be turned on while attended. APCP products doesn't recommend putting the LR on a timer.

Unfortunately it will ignore the pressure sensor when the power is first turned on so it's a small safety hazard. The LR is only intended to be turned on while attended. APCP products doesn't recommend putting the LR on a timer.

Good to know. Though as a programmer, I object to such a design assumption - some areas have frequent power outages.

In a continuing (however useful) effort to gather like data in one place, here is some additional info from probing and taking things apart:

• The base has two sensors because they are magnet polarity-specific. They are mounted together at 180 degrees to each other, hence each one detects only one of the two magnets on the globe, therefore the LR "knows" what position the globe is in, even if it winds up in a completely unexpected orientation (This negates my vulnerability comment, in theory). This is of course assuming that the sensors are working right.
• The senors are fairly standard hall effect (non ratiometric) pull-down type in a mini SO-92 package, 3 pin. They are powered with regulated +5V, and the output pin is pulled down to 0V when it detects the magnet. As far as I can tell, they are exactly like this one: http://www.adafruit.com/index.php?main_page=product_info&products_id=158. I might replace mine if it acts up again.
• There are two other sensors in the unit. The well-known pressure sensor for cativation, and the lesser-known cable-tension sensor (the metal cable on top of the base). I assume that the latter is to avoid injury to cat or LR should an object or cat wind up in the open waste door during the "dump" phase of operation. Interestingly, this one only stops the LR for a second.

I will try to leave it alone now

Hi there,

You might also try re-seating the black connector on the circuit board as the connection may have become oxidized over time. Some WD40 sprayed into the terminals helps too. If you contact customer service 877-250-7729 or customer-service@litter-robot.com they can order a new harness for you which includes the hall sensor. We don't offer this as a standard item on the web site only because it requires opening up the base to replace it and it's a bit more complicated than swapping out a circuit board. But if you're up to it, you can get one and try. I strongly recommend re-seating the black connector first to see if that works.

Regarding the pressure sensor, current code will stop the globe if the pressure sensor is activated - even during start-up. This has been the case since January of 2005.

Thanks,
Brad

APCP_Inc wrote:Regarding the pressure sensor, current code will stop the globe if the pressure sensor is activated - even during start-up. This has been the case since January of 2005.

Oh, you folks changed this just a few months after I started telling everybody that I noticed it was ignoring it during start up! I didn't notice that you fixed it.