B4 - Catgenie stops after filling with water

[Cassie Humphreys-Kearnes]

I have two Cat genies, (16F877A) was successful, and (16F1939) was not. The 16F1939 cat genie female connector was melted as previously mentioned in other posts. I filed it down and connected leads. The female connector is a little fragile and has to have the right touch to be recognized by the computer. Upon connecting a notification pops up notifying about low volt device and target. I click ok, then target is found. Sometimes it pops up Device ID (00000000) does not match target device ID. Which I figured this is MPLAB telling me poor connection. But upon touching the connector and cable right, target was identified and program verified/complete. Once running the cat genie, it only makes it to filling the bowl with water, at which point it rotates forever and then stops. If I lift the cat genie motor head off, and back on, it then beeps and continues it's cycle by draining and drying. Why does it do this?

So far I have done a geniediag, and all features run smoothly. I thought it could be the water senor that everyone seems to have trouble with, but IR light looks fine (Attached below), I put a new plastic tong and cleaned sensor lights with WD-40. I have tried re-downloading/flashing a few times now and the program still acts funny. What should I do? Do you think it is the female connector? Or the water sensor? Should I replace the connector or the IR light? Or is it something else? 

Thank you for your help, 

Cassie

[Michael Conner]

You are very correct in stating that the removal of the melted program header "Takes a Gentle Touch". Myself, I used a dental pick made of stainless steel.

I did have a very similar situation in programming one one my CG-120's Just as you did I had initial  CG operation hangs. And, Robert's diag program operated perfectly. So, I suspect that you may  have the same problem that I had.

After numerous re-programming and testing that failed, and . Since B4 failed and Diag always operated perfectly, I decided that I must have a bad download file, as everything else was the same...

So, I down loaded B4 from Robert's site, connected up the CG-120 , again, and programmed the Cg-120 once more. Everything operated perfectly - and still does after several months.

Don't forget to erase the MPU and power up the CG-120 whitest programming

Good Luck,

Michael...

. 

Cassie

--
You received this message because you are subscribed to the Google Groups "CatGenius" group.
To unsubscribe from this group and stop receiving emails from it, send an email to catgenius+...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Cassiehk]

Thank you for your fast reply Mike, very much appreciated.

I have reprogrammed several times now, and it still stops after it fills with water. I got excited with one download, it completed a whole cycle. But come the morning, water was in the bowl again. :(

The only way I can get it to empty is to lift the unit head up for a few seconds. Once back on, I hear the sensor click/check, then it drains and starts drying cycle. 

Every time I have programmed, I have download a new file from Roberts site, checked hex file id, erase unit, ensured power was on, and uploaded new file to unit. 

I cleaned the female connector, by spraying a little WD-40 on the pins connected to the jumper cable and Pic3 recognizes target immediately now. 

I have also installed a quick connect water hose, making it so easy to disconnect water and reprogram. 

I will keep re-programming in the meantime and let you know of my progress. 

But if that doesn't work, what can I do? Would installing a serial port help to diagnose the problem? 

Please help me, my cats keep peeing outside of the box cause mummy can't seem to fix it. :( 

[Michael Conner]

Cassie,

If it were me. I'd

Clean out the pump assembly

Clean the Light Tube, Top & Bottom,, install it DRY

Clean the IR LED's under the Control Box, try a Q-tip - check for Dust

Try again, without programming.

The Click ..Click is most likely one of the Relays. 

So I understand, the bowl is full and when you lift the controller box, the relay clicks and the water starts to pump out ??

Does it cycle properly though all cleaning ( water in.out) cycles ??

Try looking for the post that instructs how to empty the bowl with an Electric Drill and a socket. 

Good Luck,,,

Michael.

[Robert Deliën]

> Sometimes it pops up Device ID (00000000) does not match target device ID. Which I

> figured this is MPLAB telling me poor connection.

 

This is _always_ bad and needs to be attended first; MPLab, or any other programming utility, will try to determine the type of controller attached first. If one of the two signals is bad, an ID of 0 will be read. In my opion, this illegal ID should prevent the programmer to go any further, but somehow it does. And if it does, all kinds of things can to wrong.

 

> But upon touching the connector and cable right, target was identified and program verified/complete.

 

I'm not in favor of that because this unreliability introduces an extra variable in the diagnosis.

 

> Once running the cat genie, it only makes it to filling the bowl with water, at which point it rotates forever and then

> stops.

 

Having said the above, this probem is probably _not_ related to unreliable programming.

 

> If I lift the cat genie motor head off, and back on, it then beeps and continues it's cycle by draining and drying. Why

> does it do this?

 

That's the clean-up program. Before opening the tap, I put a marker in non-volatile memory. After drying the box, this marker is removed. When powering up, this marker is checked for, and if present the box is assumed to be wet, hence a clean-up program is triggered so you don't have to clean up the box yourself. The clean-up program just drains, dries and surfaces the box, to make it ready for use again.

 

> So far I have done a geniediag, and all features run smoothly. I thought it could be the water senor that everyone

> seems to have trouble with, but IR light looks fine (Attached below),

 

It looks fine to me too. GenieDiag is the way to go for diagnosis.

 

> I put a new plastic tong and cleaned sensor lights with WD-40. I have tried re-downloading/flashing a few times now

> and the program still acts funny. What should I do? Do you think it is the female connector? Or the water sensor?

> Should I replace the connector or the IR light? Or is it something else?

 

The water sensor is the most likely cause. It's just weird that it works with GenieDiag, but not with CatGenius; It's the same code...

 

Cheers,

 

        Robert.

[Robert Deliën]

> I have reprogrammed several times now, and it still stops after it fills with water. I got excited with one

> download, it completed a whole cycle. But come the morning, water was in the bowl again. :(

 

Is there a lot of light on your box, possibly shining into the hopper? The water sensor is sensitive to that.

 

There is a fairly good way to diagnose your sensor, but you have to flash GenieDiag for that, and you need ot have the PU on a bench top, plugged in, and with it's power switch triggered. Water should not be connected.

1. Using GenieDiag, go to the water sensor test.

2. Activate the test,.

3. Now stick the light guide into the two emitter and receiver holes in the bottom. This should activate the water valve: You should hear a sharp click, followed by a conituous mild hum of the water valve.

4. Now gently pull out the light guide, to determine when the hum stops.

 

You should be able to pull out the light guide almost entirely before the valve closes. You can play a little with it, pulling it out, tilting it a bit, to determine how stable it is. Valve closed == water detected.

 

The reason why your box freezes in washing program execution, is that is has filled up the box for the maximum amount of time, but still no water was was detected.

 

> The only way I can get it to empty is to lift the unit head up for a few seconds. Once back on, I hear the sensor

> click/check, then it drains and starts drying cycle.

 

That's what it's for.

 

> I cleaned the female connector, by spraying a little WD-40 on the pins connected to the jumper cable and Pic3

> recognizes target immediately now.

 

That's not a good idea, neither is contact spray. But I understand you're trying everything you can think of.

> But if that doesn't work, what can I do?

 

Try my suggestion above and report your findings.

 

> Would installing a serial port help to diagnose the problem?

 

Nope; We're fairly sure it's a water sensor problem.

 

> Please help me, my cats keep peeing outside of the box cause mummy can't seem to fix it. :(

 

Doing our best.

[Robert Deliën]

Any updates?

-- 

[Cassie Humphreys-Kearnes]

Sorry I haven't replied sooner, you were so fast responding to me and I feel horrible to didn't return the favor. Please forgive me. I deeply appreciated your responses :)

I have done GenieDiag test, and the water value turns on and off quite well when putting light guide in and out of emitter. The sensor is fast at shutting and opening value. My catgenie has the dome, and is situated in my hallway cupboard, no light is directed at the cat genie. 

When catgenie is going through normal cycle, it fills to the correct water height (so no flooding), and then bowl spins for a long while at which point it stops.

I don't know what else I could do? It worked fine before reprogramming. 

I'm currently operating activated scoop only, and then doing lift unit trick once a day to clean. 

Even though I have had trouble with this unit, my other one is working great! Thank you for your sharing your brilliance with us all.

  

[Robert Deliën]

Sorry I haven't replied sooner, you were so fast responding to me and I feel horrible to didn't return the favor. Please forgive me. I deeply appreciated your responses :)

Don't worry about that.

I have done GenieDiag test, and the water value turns on and off quite well when putting light guide in and out of emitter. The sensor is fast at shutting and opening value.

That proves that your water sensor is working well.

My catgenie has the dome, and is situated in my hallway cupboard, no light is directed at the cat genie.

Ok, so we can rule out that corner case.

When catgenie is going through normal cycle, it fills to the correct water height (so no flooding), and then bowl spins for a long while at which point it stops.

This can mean either two things:

1. Water wasn't detected before the safety timeout fired, so the valve was closed an no further error handling was done.

2. The water valve detects water and closes, but software did not.

In the first case, your water sensor is broken, which we have established is not the case.

In the second case, the analogue triggering value may be different for your box. Can you build the software yourself? If so, you can try using the digital algo by commenting '#  define WATERSENSOR_ANALOG' on line 23 of software/common/catgenie120.h.

I don't know what else I could do? It worked fine before reprogramming.

Hmyeah, that's the exact reason why this upgrade never went commercial; There's no way back.

I'm currently operating activated scoop only, and then doing lift unit trick once a day to clean.

That's the clean-up program; It doesn't get your box as clean as it should, but at least it gets it operational again instead of manually draining the box.

Even though I have had trouble with this unit, my other one is working great! Thank you for your sharing your brilliance with us all.

We'll get it to work, no problem, but I'm currently flooded in work.

Cheers,

        Robert.

[Michael Conner]

Robert,

Okay, so I'm confused again. 

If Her light tube operated with the Diag program, why would it not with the full normal program ??

--

[Robert Deliën]

> Okay, so I'm confused again.
>
> If Her light tube operated with the Diag program, why would it not with the full normal program ??

The water sensor directly operates he valves, by using a schmitt-trigger. Software evaluates the reflection intensity using an analog input since version B04.

[Michael Conner]

Robert,

Yes I recall that you informed me re this in the past. But does this mean the Diag code reads and operates the H2O relay (valve) differently ??

Michael

[Robert Deliën]

> Yes I recall that you informed me re this in the past. But does this mean the Diag code reads and operates the H2O relay (valve) differently ??

No, both applications are using the same driver code.

[Cassie Humphreys-Kearnes]

It is weird that it works in Diag mode and not normal mode. I have done some research into programming, and can't seem to work it out sorry. I'm clearly not as talented as some of you in here. 

I found your link referencing line #23 https://code.google.com/p/catgenius/source/browse/trunk/software/common/catgenie120.h?r=308&spec=svn375

But how to re-program or build software I still need to learn. It is a skill I hope to discover one day. 

I downloaded Pika software builder, and a little scared to continue without instruction. Sorry for being so new, you must get annoyed at us Newbies. 

No worries about being busy, your so kind to create this product and share. I'm sure we will fix this issue one day. 

On Fri, Aug 16, 2013 at 1:10 AM, Robert Deliën <rob...@delien.nl> wrote:

> Yes I recall that you informed me re this in the past. But does this mean the Diag code reads and operates the H2O relay (valve) differently ??

No, both applications are using the same driver code.

[Robert Deliën]

Just a quick reply: Learning to program, or even compile isn't the way to solve this problem. When I suggested it, I assumed you already were able to do so.

At this moment I cannot build a custom version for you, because I'm completely swamped. But there are a couple of versions posted on this group that should do more or less the same. Search for binaries build for the PIC16F1939 and check what the post says. The first versions didn't have analog water sensor readout, that came much later. Try one of those versions and post which ones work and/or fail.

If it's really the analog water sensor, we may need you to install a serial port to determine your water sensor value. With that value we can tune the current value.

(Sent from a mobile device; Please excuse grammar, typos and brevity)

[Michael Conner]

Cassie,

I read,,,  that you ...

""I cleaned the female connector, by spraying a little WD-40 on the pins connected to the jumper cable and Pic3 recognizes target immediately now. ""

This is not a good idea. Sooner or later you should clean the connector and board with a

circuit board cleaner.

This circuit board is coated with "conformal" coating" - but the ones I've seen are only coated on the top side.

Some other ideas would be to make sure your checking the box "" /MCLR " under the 

Microchip Logo on the right.

But, Robert is most certainly correct in that you should try an older rev.

Let us know your results...

As Roberts says, You will get the CG box operational, so do stick with it.

Michael

"

[Cassiehk]

IT WORKS!!!! YAY!!! 

After months of lifting the head unit to go into a clean cycle and several attempts of new downloads. I finally found the right download. I mis-read what you meant before Robert, I thought you were telling me to try older beta verisons, which I tried B3, B2 & B1 and all failed. I then re-read and discovered you meant to look up this post: https://groups.google.com/forum/#!searchin/CatGenius/binary$20PIC16F1939/catgenius/ZnW-6w-Inu0/BWiWN4eSKwQJ and downloaded your Binaries for PIC16F1939. It worked and has been operational for the last 3 days. YAY!

I don't understand why, but am very happy that my catgenie is working. Thank you again for all your help :)

[Robert Deliën]

IT WORKS!!!! YAY!!! 

After months of lifting the head unit to go into a clean cycle and several attempts of new downloads. I finally found the right download. I mis-read what you meant before Robert, I thought you were telling me to try older beta verisons, which I tried B3, B2 & B1 and all failed. I then re-read and discovered you meant to look up this post: https://groups.google.com/forum/#!searchin/CatGenius/binary$20PIC16F1939/catgenius/ZnW-6w-Inu0/BWiWN4eSKwQJ and downloaded your Binaries for PIC16F1939. It worked and has been operational for the last 3 days. YAY!

I don't understand why, but am very happy that my catgenie is working. Thank you again for all your help :)

Neither do I, and that worries me a bit… A lot has changed in the time in between an my box shows no problem, so I must be missing something. But I'm glad to hear you've got it working!