CLI-8 Cartridges, How does the chip really monitor ink?

[mikling]

Well here's an idea. What is the real reason you are trying to hack the chip? Just to get the prism system to work? Because other than that, you can print with refilled cartridges. How about going around the chip by just using the onboard system. If the system is rendered useless, then you will be able to either clip on another harness to simulate the existing leds and then add another external ink monitor system with LED indicators on it. I don't have a current printer to do this on but if someone has one they should take a look.
1. See if the LED thinktank sensors can be disconnected and how accessible it is.
2. Make up a circuit that simulates the prism system.
3. Install it onto the sensors.
4. interface the installed circuit to a USB connection and some software or an external INK low warning indicator.

Hey. My patent.

 

[pebe]

Well here's an idea. What is the real reason you are trying to hack the chip? Just to get the prism system to work? Because other than that, you can print with refilled cartridges. How about going around the chip by just using the onboard system. If the system is rendered useless, then you will be able to either clip on another harness to simulate the existing leds and then add another external ink monitor system with LED indicators on it. I don't have a current printer to do this on but if someone has one they should take a look.
1. See if the LED thinktank sensors can be disconnected and how accessible it is.
2. Make up a circuit that simulates the prism system.
3. Install it onto the sensors.
4. interface the installed circuit to a USB connection and some software or an external INK low warning indicator.

Hey. My patent.

If the sensors are not being used, would they even be fitted?

 

[canonfodder]

Pebe, the sensor is fitted. I can see it inside the printer. It probably does function when OEM cartridges are in place.

After going to refilling, the sensors are not really used, but what if the software objects to them not being present and functional? It could still be looking at the sensor signals, but not giving us the benefit of the information.

I have thought of checking to see if the light source (LED) is still powered. If so, I would try to tap the signal. That is like phone tapping. Don't interrupt the line, just listen in.

Remember Mikling, that the normal operation of the sensor is done with a slow passing of the head and the processer has to know which cartridge is over the sensor as it passes, so some synchronization is involved. I will check to see if the LED is on. Maybe we don't care about synchronization, we just want to know if ANY cartridge is going empty.

 

[canonfodder]

Well, if the LED is still powered, it must be just when the cartridges are over the sensor. I tried to detect the infrared from the LED, (assuming it is infrared because such are better at this sort of thing so even cheaper to work with) and could get nothing. When the cartridges are over the sensor, the path is so blocked that I could not expect my detector to receive any beam. I'm going to have to do partial disassembly in order to get at the wires and tap in.

An interesting note. With two cartridges definitely declared empty and refilled, by the printer, it does not light their red lights when the head is parked out front. But if you temporarily pop one of the cartridges, and reset it, the light starts blinking. Closing the lid, or faking that, the head goes to home and the next time out the light on that cartridge will be dark again. So if you remove a cartridge, and replace it, the printer has to assume, that it may be a fresh cartridge, but then can identify it as empty or should be empty, so it blinks the light to say so.

 

[pebe]

Canonfodder,
It is probably an IR device similar to an opto-isolator. If so, then the receiver would be a transistor switch that it would be possible to take a high impedance feed from without disturbing Canon's existing circuitry.

It could be used to trigger a 555 timer to make an LED flash or a buzzer sound every pass of the printhead. It wouldn't tell you which cart was low, but it would warn you.

 

[websnail]

So, am I right in assuming this discussion is now looking at forgetting about hacking the chip so much as just acknowledging that we can't change it but we can figure out a way to check for ink levels? Something of a rhetorical question but just wanted to be clear.

 

[pebe]

So, am I right in assuming this discussion is now looking at forgetting about hacking the chip so much as just acknowledging that we can't change it but we can figure out a way to check for ink levels? Something of a rhetorical question but just wanted to be clear.

Yes, I think you are right. Canon's code could be cracked but it would need a lot of data readings taken from different machines with as assortment of cartridges to do so.

 

[Grandad35]

So, am I right in assuming this discussion is now looking at forgetting about hacking the chip so much as just acknowledging that we can't change it but we can figure out a way to check for ink levels? Something of a rhetorical question but just wanted to be clear.

Yes, I think you are right. Canon's code could be cracked but it would need a lot of data readings taken from different machines with as assortment of cartridges to do so.

Unfortunately, that seems to be the consensus. We had fun on the trip, even if we ended up back where we started.

I thought that I would post photos of the hole in the bottom of the print head where the IR snesors can see the prism, as well of the sensors themselves.

Something else to think about - maybe Canon didn't protect the printer's code like they protected the data transmission between the printer and the carts. How hard would it be for the right person to hack the printer's code so that it ignored the data from the carts? Of course, a separate hack would be required for each model. Even if a complete hack proved to be too difficult, maybe it would be easier to just keep the prism active with refilled carts.

 

[tkalfaoglu]

I think hacking the printer code is the way to go.. But how

 

[Grandad35]

I think hacking the printer code is the way to go.. But how

Since the printer's internal drivers are all available for download, getting the object code is easy, as is installing the hacked version in the printer. This obviously assumes that the critical sections of code are included in the download.

I don't know how the actual hack is done, but there is a hack for a Canon 300D camera (http://satinfo.narod.ru/en/download.html), so this type of thing as been done. The camera hack did not add new features, but only enabled features which were already in the code and disabled (the 10D and 300D were basically the same camera and software, but with some features disabled in the 300D). Note that this hack comes from Russia - this may have something to do with being out of the reach of legal action.