Possible drain hose leak and unrelated observations
Sonagi
increased stench, and I noticed that the drain hose is not very
securely tied to the bottom of the cat genie. I read a couple of posts
on litter-box about drain hose failure with earlier model (due to
corroded metal clamp, leaky hose, etc), and the new model (12) has
opted for plastic clamp with different kind of glue. But looking the
the assembly itself closely, it doesn't look very secure as the drain
hose can be rotated from the plug. I thought about applying some
caulking but decided against it as I don't know how my particular (non-
silicone based adhesive kind that was not meant for constant water
contact) would react to the softer caulking (looks like silicone to
me) found on the cat genie.
The easiest preventive fix I can think of is put some duck tape to
immobilize the drain hose or use a zip tie through a small hole on the
side, and put more silicone caulking. Or I can put a big plastic basic
under the Cat genie and not worry about possible future leak. Does
anyone have this concern?
While also cleaning, I noticed that the genie impeller didn't actually
pulverize half of the poop. I have two cats, and they both used the
cat genie before I ran full cycle. Half of the poop seems to have been
pulverized and expelled, but the other half was stuck at the impeller
and hardly pulverized. I also am concerned about the durability of the
pump/impeller blades as they are made of thin plastic.
On a separate note, I am using the latest B3 (the last one attached to
the e-mail with reduced ticking sound). The tick is still there but
beat is slower. A few days ago, I found that cat genie bowl was keep
rotating even after the genie scooper finished scooping, and the red
led was blinking constantly. (two quick bursts of red led, then pause,
then two quick bursts of red led, etc). I stopped the cycle manually,
and attempted to run the scoop only cycle again, but same error. I
fixed the problem by unplugging the cord, and no more bowl keep
rotating without stop. Unfortunately, it is not attached to the PC to
log errors. Has anyone experienced this behavior?
Fraggboy
didn't notice any leaks underneath. What I perform is remove the
plastic clamp, and clean both "O" rings and lubricate them with
vaseline.
Robert Deliën
> On a separate note, I am using the latest B3 (the last one attached to
> the e-mail with reduced ticking sound). The tick is still there but
> beat is slower. A few days ago, I found that cat genie bowl was keep
> rotating even after the genie scooper finished scooping, and the red
> led was blinking constantly. (two quick bursts of red led, then pause,
> then two quick bursts of red led, etc). I stopped the cycle manually,
> and attempted to run the scoop only cycle again, but same error. I
> fixed the problem by unplugging the cord, and no more bowl keep
> rotating without stop. Unfortunately, it is not attached to the PC to
> log errors. Has anyone experienced this behavior?
Version B03b still doesn't have error handling implemented. It does detect errors, but it doesn't act upon them. Instead it just waits for error to go away automatically. Older versions stopped all motors at that time, rendering the box useless without user intervention. But in my specific case, the error seemed to resolve itself after some time. At that time we though a residual drop of water still hanging onto the light guide was causing false-positive water detections. After up to 20 minutes, the drop fell off and the program continued. Logging on the serial port would reveal the occurrence and the box would be in business again. That's why it is the way it is now. At that time, we were discussing the need to 'season' the light guide with WD40 of candle wax, to make the last drop fall off easily.
Later I too had situations where the bowl has been spinning for a very long time. My record is 2 days... In my case, it turned out that my water sensor LED was weakening, resulting into false-positive water detections. The error you're reporting (Error 2) is a draining error, indicating that water is still being detected after pumping. This either means you have a weakened water sensor LED too, or you have a water drop hanging onto the light guide (or both). It fixed my problem with the bowl spinning for a long time for sure. Though I'm no longer logging the serial output lately, so I'm not sure if a drop hanging onto the light guide is still a problem. I do know that I haven't cleaned or seasoned it for a long time, so I consider both problems cured by installing a new LED.
If you haven't done so already, I strongly recommend you to change your water sensor LED. If you already did, we would like to hear that too because we may learn something from that. In any case, you could check your LED's intensity with a digital photo camera. You will need to reflash you box with a GenieDiag version that still uses the old water sensor algo. I'm not sure what algo the most recently posted version does. But you gave me the idea that implementing a water sensor LED diagnosis mode in GenieDiag will add some value here.
But we have more good news on this: Two developers have access to the code, a third one is still hesitating, I think. Supporting these guys has given me new inspiration and I have just implemented Pause functionality of the washing program. This function is essential for proper error handling. When an error occurs, the running program will be paused and the error will be displayed. The user can either resume the program, have the washing program retrying to solve to problem, or he can stop the program and resolve the problem himself.
Until that works, using B03b firmware, the best thing you can do in such a situation is to unplug the box, clean and dry the light guide and plug in the box again. This will trigger the cleanup program, which will drain and dry your box.
Cheers,
Robert.
scot kight
The easiest preventive fix I can think of is put some duck tape to
immobilize the drain hose or use a zip tie through a small hole on the
side, and put more silicone caulking. Or I can put a big plastic basic
under the Cat genie and not worry about possible future leak. Does
anyone have this concern?
Sonagi
without use of any caulking/glue but just re-lubricating the O-rings?
Thanks.
Sonagi
> Version B03b still doesn't have error handling implemented. It does detect errors, but it doesn't act upon them. Instead it just waits for error to go away automatically. Older versions stopped all motors at that time, rendering the box useless without user intervention. But in my specific case, the error seemed to resolve itself after some time. At that time we though a residual drop of water still hanging onto the light guide was causing false-positive water detections. After up to 20 minutes, the drop fell off and the program continued. Logging on the serial port would reveal the occurrence and the box would be in business again. That's why it is the way it is now. At that time, we were discussing the need to 'season' the light guide with WD40 of candle wax, to make the last drop fall off easily.
guide. Maybe I should redo this.
> Later I too had situations where the bowl has been spinning for a very long time. My record is 2 days... In my case, it turned out that my water sensor LED was weakening, resulting into false-positive water detections. The error you're reporting (Error 2) is a draining error, indicating that water is still being detected after pumping. This either means you have a weakened water sensor LED too, or you have a water drop hanging onto the light guide (or both). It fixed my problem with the bowl spinning for a long time for sure. Though I'm no longer logging the serial output lately, so I'm not sure if a drop hanging onto the light guide is still a problem. I do know that I haven't cleaned or seasoned it for a long time, so I consider both problems cured by installing a new LED.
>
> If you haven't done so already, I strongly recommend you to change your water sensor LED. If you already did, we would like to hear that too because we may learn something from that. In any case, you could check your LED's intensity with a digital photo camera. You will need to reflash you box with a GenieDiag version that still uses the old water sensor algo. I'm not sure what algo the most recently posted version does. But you gave me the idea that implementing a water sensor LED diagnosis mode in GenieDiag will add some value here.
by the unofficial recall by cat genie manufacturer (some of the
earlier batches seemed to have bad LEDs that failed prematurely).
> But we have more good news on this: Two developers have access to the code, a third one is still hesitating, I think. Supporting these guys has given me new inspiration and I have just implemented Pause functionality of the washing program. This function is essential for proper error handling. When an error occurs, the running program will be paused and the error will be displayed. The user can either resume the program, have the washing program retrying to solve to problem, or he can stop the program and resolve the problem himself.
>
functionality? I pour hot water or extra detergent (oxyclean) when the
cat genie is pumping water. It will be nice to pause cat genie at this
state (bowled filled with water) to allow for this. Or, pause the
program to allow for manually picking up the cat poops too small.
Right now, I have to be quick or cancel the program (which will result
in re-running the entire cycle, and running the scooping function when
the granules are wet seems to waste a lot of granules because granules
clump up when wet and get thrown down the drain.
> Until that works, using B03b firmware, the best thing you can do in such a situation is to unplug the box, clean and dry the light guide and plug in the box again. This will trigger the cleanup program, which will drain and dry your box.
>
twice thus far (in many months of using your firmware) but otherwise
the firmware has been rock solid.
Sonagi
Robert Deliën
> by the unofficial recall by cat genie manufacturer (some of the
> earlier batches seemed to have bad LEDs that failed prematurely).
Ok, but I still recommend you to check it because it's fairly easy to do and it will either learn us that you'd better replace your LED or that light guide seasoning is still necessary with a healthy LED.
The new water sensor algo pulses the LED, instead of keeping it on continuously. If you have a modern camera, it will be able to see this. For the best result it's better to check using either an older version or GenieDiag with the older algo
> Will it be possible to allow users to trigger manually the pause
> functionality?
Yes; The pause function is intended to work like it does in the original firmware, but better: If the original firmware is pause while the arm is moving, timing is lost upon unpausing and the arm will make a too long movement. CatGenius will store the context of all timers, so the program should resume exactly as if it were never paused.
Besides that, the pause function is also activated in case an error occurs.
> I pour hot water or extra detergent (oxyclean) when the
> cat genie is pumping water. It will be nice to pause cat genie at this
> state (bowled filled with water) to allow for this. Or, pause the
> program to allow for manually picking up the cat poops too small.
> Right now, I have to be quick or cancel the program (which will result
> in re-running the entire cycle,
Yes, you will be able to do that. We are suffering from the same problem, because we do the same.
> and running the scooping function when
> the granules are wet seems to waste a lot of granules because granules
> clump up when wet and get thrown down the drain.
You could start a cleanup cycle to deal with that by just unplugging and replugging the box. If your granules are wet, the box will find so in it's non-volatile memory. If not, you can still force a cleanup cycle by pressing and holding the start button (if I'm correct) while replugging the box in.
If you don't have time for that, you can just attend the scooping and manually shake the scoop while it's going up. That's what I do.
> Got it. Thanks for the suggestions. The stuck bowl has occurred only
> twice thus far (in many months of using your firmware) but otherwise
> the firmware has been rock solid.
Twice is not that much. But did both cases occur recently?
'Rock solid', I like the sound of that!
Fraggboy
> without use of any caulking/glue but just re-lubricating the O-rings?
> Thanks.
nipple into the base. I cleaned the surfaces really well, and
lubricated the two O-Rings with vaseline. I then placed some vaseline
inside the drain hose as well.
I just checked the bottom, and there is no seepage. I clean the CG
thoroughly every 2 months.
Fraggboy
> without use of any caulking/glue but just re-lubricating the O-rings?
> Thanks.
Sonagi
error). I should probably change the LED with a new one from
Radioshack but am procrastinating for now (especially since my last
time servicing the cat genie resulted in one of my cats taking a big
dump on my bed :( ). I could take out the very reliable Omega Cat Paw
litterbox, but I opted to just clean the water sensor for now (no
priming with candle this time). I noticed that the water sensor was
dirty like the last time and I am making following guesses.
The candle wax priming seems to make the plastic water sensor more
prone to become dirty. It might allow the water to drop off more
easily from the water sensor, but it might make the water sensor more
dirty by making the water sensor's surface less smooth. Maybe the
trick is to use really really little bit of candle wax. I will try
WD40 next time as this will be easier to apply a thinner layer.
Last time, I thought if I used the original cat genius 120 cartridge
it might keep my unit more clean and also have cleaner water sensor.
But it seems this is not the case. I know that the original cleaning
solution is stickier than my custom (water and liquid dish washing
soap) detergent as the granules seem to stick more to cat genie wall
(almost none with my custom mix or with no detergent).
Right now, I am running on auto cat activated mode (full wash with
water every 3 uses) but I have two cats. So I might be running full
wash once every six use. I think that's like four urination plus 2
defecation before full cycle with water. So this might make my water
sensor especially dirty quickly as it is dealing with more dirty
water. So I might set the full wash to run more often. So full wash
once 2~4 uses instead of 4~6 uses.
You mentioned that you changed the water pump relay to reduced the
tick with beta 3-3, and I indeed notice slower tick. I rarely had
false water sensor error with the original Beta 3 (with faster ticks).
I might try going back to the original Beta 3.
Related observations:
The Beta 3-3 seems to use up the detergent more quickly than either
previous versions or the stock firmware. In just 20 days, the new
original cartridge was almost empty (this with about 2~3 full washes
daily, so about 40~60 full washes before empty). Anyways, I don't have
any original cart anymore (other than the maintenance cart which I
have yet to use).
On the future version of the firmware, would it be possible for the
scoop-only/full-wash cycle to reset after running the full cycle
manually? So if I run a full cycle manually, the scoop only cycle
count will reset before running the full cycle.
Thanks. :D
Robert Deliën
> prone to become dirty. It might allow the water to drop off more
> easily from the water sensor, but it might make the water sensor more
> dirty by making the water sensor's surface less smooth. Maybe the
> trick is to use really really little bit of candle wax. I will try
> WD40 next time as this will be easier to apply a thinner layer.
You could be right here. Even though the wax may repel water, it may be sticky to other dirt. I'm no longer priming my light guide anymore since I have replaced the LED.
> Last time, I thought if I used the original cat genius 120 cartridge
> it might keep my unit more clean and also have cleaner water sensor.
> But it seems this is not the case. I know that the original cleaning
> solution is stickier than my custom (water and liquid dish washing
> soap) detergent as the granules seem to stick more to cat genie wall
> (almost none with my custom mix or with no detergent).
That's good to know: I thought the original fluid would keep the water sensor working better too, but I suppose any other soap - breaking surface tension - will do just was well. BTW: I'm running without soap for a year now and I don't have any problems.
> Right now, I am running on auto cat activated mode (full wash with
> water every 3 uses) but I have two cats. So I might be running full
> wash once every six use. I think that's like four urination plus 2
> defecation before full cycle with water. So this might make my water
> sensor especially dirty quickly as it is dealing with more dirty
> water. So I might set the full wash to run more often. So full wash
> once 2~4 uses instead of 4~6 uses.
I'm running a full wash after every use, one on one. The scooping smears the urine around the box. It dries pretty quickly, so it's not a problem for a long time, but one of my cats doesn't like to use the box shortly after the smearing.
> You mentioned that you changed the water pump relay to reduced the
> tick with beta 3-3, and I indeed notice slower tick. I rarely had
> false water sensor error with the original Beta 3 (with faster ticks).
> I might try going back to the original Beta 3.
You have to check the change list on the release page on the group to see what more has changed. I don't think you need to replace the relay. Some may tick, others may not, but it doesn't harm the relay because it's not a full move of the lever. It's just a very brief rattle.
> Related observations:
> The Beta 3-3 seems to use up the detergent more quickly than either
> previous versions or the stock firmware. In just 20 days, the new
> original cartridge was almost empty (this with about 2~3 full washes
> daily, so about 40~60 full washes before empty). Anyways, I don't have
> any original cart anymore (other than the maintenance cart which I
> have yet to use).
The dosage is base on a manually triggered washing program, so you should get 60 washes out of a cartridge. We could make this configurable.
If there's a difference with other releases, it could be caused by the auto-dose command. Again you can check the release notes to see when this was introduced.
> On the future version of the firmware, would it be possible for the
> scoop-only/full-wash cycle to reset after running the full cycle
> manually? So if I run a full cycle manually, the scoop only cycle
> count will reset before running the full cycle.
It should. If it doesn't, that's a bug.
Earl
base's discharge nipple. It was siliconed on there just like it was
on the CG60. The clip on there does not secure the hose...it seems to
be only as a stop. I was able to twist and pull on it to disconnect
it from the discharge nipple. I'm using some white Gorilla glue to
secure the new clear hose back on. The new clear/black striped hose
can be purchased at a marine store as bilge hose. This will allow you
to see if the hose is clogged or not.
http://www.tridentmarine.com/stage/bilge.htm
Sonagi
First error: It seems like water sensor error in that the bowl kept
spinning, but there was no blinking LED and the cat genie didn't even
begin to scoop. So basically cat genie started spinning the bowl, but
didn't scoop. So the bowl kept spinning (for more than 10 minutes)
until I canceled the running program. After reset, the catgenie
successfully ran the full wet mode after a use by a cat.
Second error: Freezing box. A few days ago, there was another error
where the Catgenie just froze after a cat has used it. No spinning
bowl, no led error, no scooping, and all the buttons didn't seem to do
anything (no beep). I reset the machine by removing the power. There
was a few times before where the catgenie would just freeze. It has
been only 13 days, but I have yet to get the classic water sensor
error (blinking red led plus spinning bowl).
Interestingly the Beta 3 posted on the file page (rather than the
slower click version that was e-mailed) didn't have this strange
behavior.
I am now logging the Catgenie Beta 3 (e-mail version) using serial
port. It seems I am getting strange error once a week so hopefully the
log would catch it.
Anyway having to unplug the catgenie to fix a similar error? Thanks.
Richard W.
Richard
Sent from my iPhone
Robert Deliën
> First error: It seems like water sensor error in that the bowl kept
> spinning, but there was no blinking LED and the cat genie didn't even
> begin to scoop. So basically cat genie started spinning the bowl, but
> didn't scoop. So the bowl kept spinning (for more than 10 minutes)
> until I canceled the running program. After reset, the catgenie
> successfully ran the full wet mode after a use by a cat.
I had that once, I think. In my case the box didn't respond to the keyboard either, so I had to power-cycle it. In my case, probably the whole software stack was hanging. In the next release we will re-enable the watchdog again, which will reset the whole board in these situations. I had to disable it, because I noticed a number of un-explained resets.
In your situation, it was probably just the washing program that was hanging. We run a program-within-a-program to do the washing cycle. This could be due to a bug in the timer implementation. We could guard the program for that, but I doubt if that's of any use.
> Second error: Freezing box. A few days ago, there was another error
> where the Catgenie just froze after a cat has used it. No spinning
> bowl, no led error, no scooping, and all the buttons didn't seem to do
> anything (no beep). I reset the machine by removing the power. There
> was a few times before where the catgenie would just freeze. It has
> been only 13 days, but I have yet to get the classic water sensor
> error (blinking red led plus spinning bowl).
That will be worked-around by the watchdog.
> Interestingly the Beta 3 posted on the file page (rather than the
>slower click version that was e-mailed) didn't have this strange
> behavior.
I can check if beta 3 has it enabled.
> I am now logging the Catgenie Beta 3 (e-mail version) using serial
> port. It seems I am getting strange error once a week so hopefully the
> log would catch it.
That would be great! Without it, it's just guessing. With it, we may have some insight.
Robert Deliën
> posted it. I believe it was before B03. I went on a vacation for several
> days and when I came back the CG had never cleaned. It was a horrible
> smell. It never happened again.
I these situations, it's very important to check if the box still responds to the keyboard and the cat sensor. If it still responds to the keyboard, we know that the software didn't hang up. If it even responds to the cat sensor, it must have missed a couple of detections.
Sonagi
experience a lockup or spinning bowl thus far, and it will probably
take a week or so before I catch it. Meanwhile, I was looking at the
log so far, and I find a few errors. I had two Error 3 and one Error
12 and I have no idea what they are. Meanwhile, I am using Monserial
under Ubuntu to log the serial output. If you know a more robust
program with gui under linux please let me know. On the other hand,
the USB serial adapter that I got for four dollars from Ebay seems to
reset every 10 hours. It might be cause it's defective or my USB port
is not supplying enough power to the serial adapter since I am
plugging in USB keyboard, USB mouse, a USB flash memory stick, and the
USB serial adapter using a four-port unpowered USB hub to an
underpowered Clarion Mind (a atom-based GPS naviator hacked to run
regular ubuntu).
Anyways, I plugged in a separate power adaptor for the USB hub, so
hopefully the usb serial adapter won't reset. Thanks for your help.
_______________________
0 17:23.40 IP 0x12C4: CMD_START, wet
0 17:28.58 Cat in
0 17:29.02 Cat out
0 17:29.36 Cat in
0 17:29.37 Cat out
0 17:29.56 Cat in
0 17:29.56 Cat out
0 17:29.59 Cat in
0 17:30.01 Cat out
0 17:30.02 Cat in
0 17:30.03 Cat out
0 17:30.04 Cat in
0 17:30.07 Cat out
0 17:30.08 Cat in
0 17:30.13 Cat out
0 17:30.18 Cat in
0 17:30.20 Cat out
0 17:30.22 Cat in
0 17:30.22 Cat out
0 17:30.22 Cat in
0 17:30.23 Cat out
0 17:30.26 Cat in
0 17:30.26 Cat out
0 17:30.27 Cat in
Err 3 #first error#
0 17:30.33 Cat out
0 17:30.34 Cat in
0 17:30.34 Cat out
0 17:30.35 Cat in
0 17:30.36 Cat out
0 17:30.38 Cat in
0 17:30.40 Cat out
0 17:30.41 Cat in
0 17:30.50 Cat out
0 17:30.51 Cat in
0 17:30.52 Cat out
0 17:30.52 Cat in
0 17:30.54 Cat out
0 17:30.54 Cat in
0 17:30.57 Cat out
0 17:30.57 Cat in
0 17:31.01 Cat out
0 17:31.01 Cat in
0 17:31.15 Cat out
0 17:31.16 Cat in
0 17:31.17 Cat out
0 17:31.18 Cat in
0 17:31.18 Cat out
0 17:31.19 Cat in
0 17:31.22 Cat out
0 17:31.29 Cat in
0 17:31.31 Cat out
0 17:31.32 Cat in
0 17:31.41 Cat out
0 17:31.42 Cat in
0 17:31.55 Cat out
0 17:31.57 Cat in
0 17:32.09 Cat out
0 17:32.09 Cat in
0 17:32.12 Cat out
0 17:32.12 Cat in
0 17:32.13 Cat out
0 17:36.13 Cattimer expired
0 17:36.13 IP 0x12C4: CMD_START, dry
0 17:41.49 CMD_END
Err 12 #Second error#
0 18:39.14 Cat in
0 18:39.40 Cat out
0 18:39.53 Cat in
0 18:39.57 Cat out
0 18:40.11 Cat in
0 18:40.22 Cat out
0 18:40.26 Cat in
0 18:40.35 Cat out
0 18:40.40 Cat in
0 18:40.42 Cat out
0 18:40.43 Cat in
0 18:40.45 Cat out
0 18:40.45 Cat in
0 18:40.48 Cat out
0 18:40.48 Cat in
0 18:41.14 Cat out
0 18:45.14 Cattimer expired
0 18:45.14 IP 0x12C4: CMD_START, wet
0 18:50.07 Filling
0 18:51.42 Water high
0 18:51.42 Filled
0 18:52.46 Draining
0 18:52.48 Water low
0 18:52.48 Drained
0 18:53.11 Draining
0 18:54.27 Draining
0 18:54.51 Draining
0 18:54.59 Draining
0 18:55.24 Draining
0 18:55.32 Draining
0 18:56.48 Draining
0 18:56.48 Filling
0 18:58.14 Water high
0 18:58.14 Filled
0 18:58.58 Draining
0 18:59.02 Water low
0 18:59.02 Drained
0 18:59.23 Draining
0 19:00.39 Draining
0 19:01.03 Draining
0 19:01.12 Draining
0 19:01.36 Draining
0 19:01.44 Draining
0 19:03.00 Draining
0 19:03.00 Filling
0 19:04.26 Water high
0 19:04.26 Filled
0 19:05.38 Draining
0 19:05.40 Water low
0 19:05.40 Drained
0 19:09.35 Draining
0 19:28.21 CMD_END
Err 3 #Third Error#
1 4:06.00 Cat in
1 4:06.08 Cat out
1 4:06.09 Cat in
1 4:06.17 Cat out
1 4:06.21 Cat in
1 4:06.49 Cat out
1 4:06.51 Cat in
1 4:06.52 Cat out
1 4:06.53 Cat in
1 4:06.56 Cat out
1 4:06.58 Cat in
1 4:07.05 Cat out
1 4:07.05 Cat in
1 4:07.07 Cat out
1 4:07.07 Cat in
1 4:07.31 Cat out
1 4:07.32 Cat in
1 4:07.33 Cat out
1 4:11.33 Cattimer expired
1 4:11.33 IP 0x12C4: CMD_START, dry
1 4:17.09 CMD_END
1 5:38.15 Cat in
1 5:38.16 Cat out
1 5:38.28 Cat in
1 5:38.29 Cat out
1 5:38.30 Cat in
1 5:38.36 Cat out
1 5:38.49 Cat in
1 5:38.51 Cat out
1 5:42.51 Cattimer expired
1 5:42.51 IP 0x12C4: CMD_START, wet
1 5:47.43 Filling
1 5:49.16 Water high
1 5:49.16 Filled
1 5:50.19 Draining
1 5:50.22 Water low
1 5:50.22 Drained
1 5:50.45 Draining
1 5:52.00 Draining
1 5:52.25 Draining
1 5:52.33 Draining
1 5:52.57 Draining
1 5:53.05 Draining
1 5:54.21 Draining
1 5:54.21 Filling
1 5:55.49 Water high
1 5:55.49 Filled
1 5:56.34 Draining
1 5:56.38 Water low
1 5:56.38 Drained
1 5:56.59 Draining
1 5:58.15 Draining
1 5:58.39 Draining
1 5:58.47 Draining
1 5:59.11 Draining
1 5:59.19 Draining
1 6:00.35 Draining
1 6:00.35 Filling
1 6:02.02 Water high
1 6:02.02 Filled
1 6:03.13 Draining
1 6:03.15 Water low
1 6:03.15 Drained
1 6:07.10 Draining #this is when the usb serial cable reset#.
Sonagi
to see how the USB serial adapter reset, and I noticed something about
the usb port itself being reset and dmesg suggesting EMI
(electromagnetic interference) as the cause. The logging computer and
usb hug is right above the Catgenie, so I wonder if catgenie is
creating too much EMI to reset the device. I don't know whether it
will work or not, but I wrapped the USB hub with a single layer of
aluminum foil to shield. :D Hopefully this and supplying separate
power to the USB hub will fix the usb serial problem.
On the other hand, after checking the new log, I noticed error 9 which
occurred without the catgenie being used. This was after changing the
Catgenius mode a few times to check the logging function and it seems
error 9 occured a few seconds or minutes after mode change.
Robert Deliën
Errors logged as "Err xx" can be ignored: They are debug errors for the cartridge tag reader. I've put them in to debug the copying protection tag I was still planning at that time. They are unrelated to any other functional parts of the box.
Sonagi
catgenie crashing or erroring out. On the other hand, the USB serial
cable that I bought seems to reset every few hours (almost always
after the start of the wet washing program or the start of the drying
phase of the wash). I first thought it was due to inadequate power to
the USB hub but plugging in the hub adapter made the serial adapter
crash even faster (less than one hour). Shielding the hub with
aluminum foil didn't seem to improve stability. Now I have wrapped the
USB serial adapter with aluminum foil, but I don't think it will work.
I think there is a surge in voltage during the whole wash cycle
(especially during the blower drying phase), and this is somehow
resetting the adapter. Perhaps my cheap RS232 Serial level shifter
(bought for like $2 from Ebay) is defective and can't handle the
voltage fluctuation.
Any updates on the next beta of the Catgenius? I know Rob is working
on the pause functionality, but re-enabling the watchdog would be
nice. On the other hand, I think I can proactively reduce the chance
of catgenie locking up by resetting power to the catgenie every week
or day (to clear the buffer and such). Rob, was the watch dog function
enabled in the Beta 3 posted on the files page (as opposed to the e-
mail version)?
On an unrelated note, I am using Ajax dish washing detergent (about 1
part in 10 part water), and this seems to be extremely effective. One
thing I have definitely noticed is that, unlike other detergents I
have tried, including the original quad formula, there is no more
brown ring of excrement in my toilet after a full wash. Before, I
would have to flush the toilet immediately after the finish of the wet
wash OR I will get a rather hard to clean brown ring of excrement on
my toilet bowl. Now, with Ajax dish washing detergent, I don't have
this ring even after now flushing the toilet after the finish of the
wet cleaning cycle. I guess this detergent is really good at cutting
grease and making the excrement non-sticky after being liquified, and
I think this (in addition to weekly hot water treatment) will lead to
less frequent Water sensor error in the future. Rob has mentioned that
he is running his machine with just plain water, and has yet to
encounter any water sensor error after replacing his LED. As for me, I
noticed with my last water sensor error, my original cartridge has
almost run out and I have been running 2 dry to 1 wet (with two cats
this can mean that I would have had only one wet wash after 3 ~ 6 uses
by 2 cats).
Thanks.
Robert Deliën
> (especially during the blower drying phase), and this is somehow
> resetting the adapter. Perhaps my cheap RS232 Serial level shifter
> (bought for like $2 from Ebay) is defective and can't handle the
> voltage fluctuation.
That has been my suspicion too. The next version will have brown-out detection enabled, so at least the whole processing unit will reset along. I'm intending to buy a logging multi-meter, a Fluke 289, but even that one will only take 10 measurements per second so a brief surge will still go unnoticed.
> Any updates on the next beta of the Catgenius? I know Rob is working
> on the pause functionality, but re-enabling the watchdog would be
> nice. On the other hand, I think I can proactively reduce the chance
> of catgenie locking up by resetting power to the catgenie every week
> or day (to clear the buffer and such).
The pause function is in. So is a more decent error handling: The box will go into pause in case of an error. The big bug Christian found in the LitterLanguage interpreter is fixed too. For the rest it's just cosmetics and small bug fixes. For example: If you unplug your box while it is running a wet cleanup program, it ran a dry cleanup program when plugged in again. I still need to test all these features to some degree.
I'm also thinking about making an improvement in the washing program we discussed in the group earlier, called 'flushing'.
After scooping we stop the bowl before we open the tap. When the tap is open, we wait until the level hits just below the bottom of the bowl. At that moment we start the pump, while leaving the tap open. The pump drains faster than the tap fills, so the water level will decrease. When the water is almost drained, and the waste is flushed away, the pump will stop, the bowl will start spinning again and hence the bowl will fill. The purpose of this flushing is that we will flush out the waste before we give it a chance to spread over the whole box when filling it.
The copy protection is mostly implemented too, but we won't activate that.
> Rob, was the watch dog function
> enabled in the Beta 3 posted on the files page (as opposed to the e-
> mail version)?
I have checked, but I cannot see: The code was in at that time, but it is not active when a debug version is built. Unfortunately I cannot see if the binary I have posted is a debug build or a release build.
> On an unrelated note, I am using Ajax dish washing detergent (about 1
> part in 10 part water), and this seems to be extremely effective. One
> thing I have definitely noticed is that, unlike other detergents I
> have tried, including the original quad formula, there is no more
> brown ring of excrement in my toilet after a full wash. Before, I
> would have to flush the toilet immediately after the finish of the wet
> wash OR I will get a rather hard to clean brown ring of excrement on
> my toilet bowl. Now, with Ajax dish washing detergent, I don't have
> this ring even after now flushing the toilet after the finish of the
> wet cleaning cycle. I guess this detergent is really good at cutting
> grease and making the excrement non-sticky after being liquified, and
> I think this (in addition to weekly hot water treatment) will lead to
> less frequent Water sensor error in the future.
Warm water helps, but don't make it hot: The water valve is not made to withstand hot water. Jaime had his water valve partially molten, leading to a weak stream filling the box. It took him quite some time to diagnose the filling timeouts down to the hot water damaging the valve.
Dish washing detergent seems like a good idea. Are you sure that it won't damage the internal silicon tubing? Or dull the water sensor light guide? It is excellent in degreasing and I'm pretty sure it's killing germs as well. But how do you get it into the cartridge? I have tried to solve dish washer detergent powder in water before and It was very difficult. It seems to take a long time, probably to spread the time it is active within the machine.
> Rob has mentioned that
> he is running his machine with just plain water, and has yet to
> encounter any water sensor error after replacing his LED. As for me, I
> noticed with my last water sensor error, my original cartridge has
> almost run out and I have been running 2 dry to 1 wet (with two cats
> this can mean that I would have had only one wet wash after 3 ~ 6 uses
> by 2 cats).
We are no longer using scoop-only, but I know other people who still use it. One of our 3 cats refuses to use the box if it's not completely clean, so we're back at one-on-one. This, and the lack of the 12 hour sleeping of the box, has caused our water usage to increase to 150%. But fortunately water is still cheap.
Sonagi
> The pause function is in. So is a more decent error handling: The box will go into pause in case of an error. The big bug Christian found in the LitterLanguage interpreter is fixed too. For the rest it's just cosmetics and small bug fixes. For example: If you unplug your box while it is running a wet cleanup program, it ran a dry cleanup program when plugged in again. I still need to test all these features to some degree.
error? At least this will put the catgenie in a useable state.
> I'm also thinking about making an improvement in the washing program we discussed in the group earlier, called 'flushing'.
> After scooping we stop the bowl before we open the tap. When the tap is open, we wait until the level hits just below the bottom of the bowl. At that moment we start the pump, while leaving the tap open. The pump drains faster than the tap fills, so the water level will decrease. When the water is almost drained, and the waste is flushed away, the pump will stop, the bowl will start spinning again and hence the bowl will fill. The purpose of this flushing is that we will flush out the waste before we give it a chance to spread over the whole box when filling it.
scum buildup at the bottom of the catgenie and thus reduce overall
smell.
>
> Warm water helps, but don't make it hot: The water valve is not made to withstand hot water. Jaime had his water valve partially molten, leading to a weak stream filling the box. It took him quite some time to diagnose the filling timeouts down to the hot water damaging the valve.
catgenie is filling with cold water using a saucepan. So the water
valve should be safe. However, I wonder if the hot water will melt the
silicon glue that connects the drain hose to the catgenie base (as I
found the hose to get pretty warm to touch).
>
> Dish washing detergent seems like a good idea. Are you sure that it won't damage the internal silicon tubing? Or dull the water sensor light guide? It is excellent in degreasing and I'm pretty sure it's killing germs as well. But how do you get it into the cartridge? I have tried to solve dish washer detergent powder in water before and It was very difficult. It seems to take a long time, probably to spread the time it is active within the machine.
>
the silicon tubing. A couple of litter-box forum users have used the
same detergent. It is a just regular dish washing detergent that
should be also gentle on people's hands. Considering that the original
quad solution is pretty harsh stuff, this should be relatively safe. I
am using aquarium clear plastic air tubing to connect to the cartridge
nipple directly. A clear plastic one-gallon juice bottle holds the
diluted liquid detergent. I have experimented with power detergent
(Oxyclean) but encountered a few problems later. First, when I didn't
use the catgenie for a weak and went back to the old manual litterbox,
the oxyclean + water solution clogged up the catgenie silicone tubing
pretty badly. This can be unclogged by running very warm water only
solution and the diagnose mode (detergent dispenser test) for a few
minutes. It clogged because the powder detergent didn't fully melt in
the water, or it produces a lot of foamy bubbles (which dried up
inside the tubing when I opened it). Another issue I had was this mix
produced a lot of bubbles inside the toilet bowl. The bubble/foam
tended to lift up the liquified cat stool, and make a large ring
inside the toilet bowl (less than original solution though).
You can find the ingredients inside the liquid detergent I am using.
(Water, Ammonium C12-C15 Pareth Sulfate, Lauramidopropylamine Oxide).
I think it's the Ammonium C12-C15 that is doing the degreasing.
http://www.colgate.com/app/Colgate/US/HC/Products/Dishwashing/Ajax/Product-Ingredients/Antibacterial-Orange.cvsp
Again, I am using very little of this (1 part of this liquid detergent
plus 9 part water) and it works just as the original solution with no
tracking, less smell, and cleaner toilet.
> We are no longer using scoop-only, but I know other people who still use it. One of our 3 cats refuses to use the box if it's not completely clean, so we're back at one-on-one. This, and the lack of the 12 hour sleeping of the box, has caused our water usage to increase to 150%. But fortunately water is still cheap.
to have the catgenie not run for a few hours at night to reduce noise,
or at least have a mode where at night, it only scoops. Or I guess I
can just unplug it at night Do you think the daily power-cycling will
damage the catgenie.
A new version to flash would be really great though you are probably
more occupied with CartridgeGenius. Thanks. :D