HP printer cartridge control module

Here is some info about the current status of the project.

I've made a raw signal dump from a real HP DeskJet 2130 printer.
It does not contain F3 and F5 signals and has some noise, but it's pretty readable.
Printed image: test-page-fill.png
Data: test-page-fill-data.srzip - can be opened with PulseView

PCB debugging is in progress with the following results:

First: the DC-DC step-up converter changed from FP6292IR-G1 to HX3608 since the first one did not provide enough current to initiate ink firing.

Second: some issues were found while testing the module with ESP32 MCU using this firmware

- Protection circuit on F3 and F5 signals did not function properly, its width is more than 10us

As a result, I think that the magenta color of my cartridge has been burned out since the cartridge is no more printing with magenta after my experiments on the real printer.

- Some software issues exist around every third DCLK group.

Next steps to do:
1. Temporary remove protection circuit that limits F3 and F5 width and drives all signals directly with the software.

2. Debug firmware with an in-circuit debugger.

I have not been written for a while but debugging of the assembled solution is in progress. Slowly but it's moving forward.

After tests of each component independently, it seems that all parts work correctly:

• The software produces correct signal sequences.
• The PCB converts levels correctly.

But when I put it all together and tested it with a real cartridge, nothing happened. The cartridge does not print... unfortunately...

I have switched the environment from Arduino to ESP32, have been implementing some debug code for it, have been written a monkey generator tool that adds some adjustments to the base signal sequences... Minor sequence adjustments have not yielded any results... unfortunately...

I suggested that this might be so. So I decided to buy a whole printer, disassemble it and reverse engineer target signals to compare what is the difference between my implementation of sequences and the original one.

The contacts which are going to the color and black print cartridge would be brought out as two IDC ports.

Then all data will be logged with the oscilloscope and will be analyzed. All dumps will be presented here.

The journey continues.

Here is it, the software part of the project for Arduino board is ready and can be downloaded from github.

I've implement simple Arduino application to control cartridge ink jetting through module. It was not so hard as I expected, since all cumbersome bit logic is implemented and well documented by Sprite_tm in his repository of simple printcart.

This code has not been tested with real hardware yet, but output pins diagram looks promising.

The second prototype of the fully assembled mount .

As it can be seen mount fits pretty well with cartridge and PCB connected through pogo pins.

I've started to implement code to control cartridge with ESP32-WROVER here on github, but unfortunately it can not be tested with real cartridge since my first PCB prototype has an stupid issues. So I should fix it... again.

Stay tuned...

The first almost successful prototype of the mount:

It's the first time i deal with 3D printing and the second design I have tried to print. First design was really not so good as I expected:

It's printed with FDM printer and precision is not enough for this project, so the second time I have printed on SLA printer with ABS-like resin.

Second design fits pretty well with cartridge:

The pins are positioned fairly accurately, it can be seen by the scratches on on contact pads of the cartridge:

First time I have seen printed mount, I thought that many of the contacts will touch each other, but only CSYNC and D3 touches.

Next try I'll fix it and will try to connect to PCB which will be ready soon.