Hacking cartridge HP 84/85

[MAsic12345]

and what protocol does hp 45 have for data transmission
I can not find the protocol with a seven-bit clock
there is at whom that oscillogram hp 45
I have not yet started to connect to arduino, waiting for materials with aliexpress

[dragonator]

I think logic is 2.5V. I have not seen more on the scope signals. Maybe the head also works on 3.3V, but you might damage it.

and what protocol does hp 45 have for data transmission
I can not find the protocol with a seven-bit clock

Do you mean HP84/85? I do not think it is a standard protocol, but a 2 data line (9 and 13) shift register. 5 is the clock and 19 is the latch/trigger.

The HP45 is a multiplexed array, with address on one side and primitive on the other. You are wired directly to all the nozzles and have to do everything yourself.

[math]

@dragonator

What do you think about 15?

[dragonator]

It baffles me a bit. With a high every 11 pulses it HAS to be related to the address, but it is very weird.

I assume that 19 is a trigger signal, and while it is high, the selected nozzles are powered (latched). However 15 turns high while 19 is already high. It might toggle the printhead side while it resets the address, but I do not see any reset for this, and I do not expect there to be only a toggle, that seems fragile.

I also do not expect there to be any clock in the printhead that might delay some of the functions. I have seen chips where the high time was enacted after the pin became low again. I also do not see how anything might reset if any pin is low for an amount of time. That would also require a clock.

Assuming there is no clock in the printhead, my biggest mystery is how does it know what address it is printing on. Triggers on 19 can advance it, but it needs to be reset at some point or you can easily lose synchronization.

Another possibility that 15 is a feedback signal, not a driver signal. Nothing in the scope files tells us about direction. We only assumed that everything is coming from the controller. If it is feedback, every 11 addresses is plenty to know what address you are on and not lose synchronization. Once for address 0, and once for 11.

[Wonko]

Aw, great - I did order a used HP Designjet 90 just because I want to figure out this puzzle. I also got some Teensy controllers. It's not that I don;t have enough to do in my life anyway!

[Wonko]

OK, here is my approach. Next to the digital signals, Reset, Clock, and Data (x2), we need a Fire signal which seems to be digital as well. There may also be a Store signal that moves data from the shift register into flipflops, so that the fire signal can be independent of the shift clock. Also, since the head is probably firing in some row/col scheme, the first firing signal will likely trigger a burst of fire signals on the row or column, so there needs to be a clock for that as well.

To keep drop size consistent, we need to be able to measure the head temperature and then regulate the voltage - or pulse length - when firing the nozzles.

Since reading the temperature is probably not digital, we will need to find the pin that either changes resistance or voltage when the head temperature changes. @MAsic12345 maybe you can take a cold head, measure analog voltages, then slowly and carefully heat the head with a hot air gun, and check if voltages change. If only one pin changes, this is likely the one that returns temperature to the head. That would be one less pin to figure out.

Pin 8 and 11 have 12V. Why are they divided? To run the left and right row? If we can put some masking tape on 8, but not 11, print a pattern, and see the changes, we will know. The same for pin 17 and 18. But why does it need 9V *and* 12V?

Is pin 10 the logic power supply voltage? Why is it 5V if the signals are at 2.5V?

Also, we assume that 9 and 13 are data 1 and data 2. If we mask one off, we can again see if only half of the printhead works, and which half.

The last thing that I find interesting: which of the supply voltages are present when the printer is *not* printing? Or even, when the head is not inserted? That could really help when trying the printhead outside of the printer.

Fun, fun, fun

[math]

II guess that this patent is the answer to protocol and other details related to the printheads like hp 84, hp 88, hp 11, that seem to use same circuit:

https://patents.google.com/patent/US6705694

[Wonko]

Very very good find!

[math]

@Wonko

Thanks!

[MAsic12345]

I tried to block the contact with tape and the printer gave an error tried any contact
I think it is connected with the fact that the rest of the contacts do not stick out of the thickness of the scotch Scotch was very thin, used Kapton
or the printer simply recognized that the contact is not connected
I measured the voltage at the input to the head without a head and only contact 12 gave 3.3v all other 0.01v
I did not find the analog signal
I checked the signal voltage more accurately, and they really 3.3v only pin 5 has for some reason 2.5v
ordered the pogo pin to assemble a testing carriage for the test

0-02-04-1f21272dd08cb305e393e8a75bd5798e518c274624235146fa36b5b87acd0880_a6db50f4.jpg (124.73 KiB) Viewed 20842 times

And now I solve puzzles, how to repeat the signal. I try to learn C ++ because I have no idea how to do it
Here at the moment I have

sfgsdfg.png (398.16 KiB) Viewed 20842 times

sfdg.png (430.56 KiB) Viewed 20842 times

sgffg.png (393.32 KiB) Viewed 20842 times

0-02-04-b8f5b050a5274abab5eade7ebf0a309aae57c2d02678ff4c2f0c3b9e485b7710_b6fd5ae7.jpg (157.33 KiB) Viewed 20842 times

Thank you for your interest in this topic and help.