Finding the pinout
The cartridge that was chosen is an HP 51640M cartridge, also known as an
'HP type-40, magenta'. This cartridge is available in magenta (red-purple)
as well as cyan (light blue), yellow and black. The circuit has only been
tested on the magenta and cyan cartridges, but it can be assumed that all
type-40 cartridges have the same pin-out, therefore also the yellow and the
black. Other HP cartridges do not appear to have the same pinout, although
the operating principle is probably the same.
This is the type of cartridge we're playing with, although the one actually used
is magenta instead of yellow.
In the first generation of inkjets the cartridges had only about 24 nozzles.
Driving this system was simple: one side of all the nozzle resistors were tied
together and connected to one pin of the connector. The other side of the
little resistors each had their own connector pin. The pin-layout for such
a cartridge can be figured out with a simple multimeter. A newer cardridge isn't
as easy, tho:
The cartridge that we use here has over 100
nozzles. The number of connecting pads is a lot smaller, however. It appears
that HP has built some electronics into the print head that takes care of
driving all the nozzles. And now, as electronics hobbyist you are facing a
problem. What's inside that piece of electronics? You see, there are several
methods that could be thought of to enable to nozzles to be driven
individually. From shift registers and multiplexers to a simple matrix to
reach the nozzles one by one. The latter method can still be figured out with
a multimeter, but the other methods require at least a logic analyser to
discover what is going on, or a stroke of genius. Opening the printer where
this cartridge belongs leads to nothing in this case. The only thing that can
be seen is that a number of connections are tied to ground and the others run
into an unidentified IC.
In the end, the author, after thinking long and measuring a lot, arrived at a method copied from gene technology. Make a gene defective, observe what changes in the organism and then you know what that gene does. This is also possible with inkjet cartridges: cover some connector pads with sticky tape and look at what goes wrong. After much messing about, it appeared that with exactly three pins (one of which is ground) just one thin line was drawn by the printer. These pins are therefore enough to drive one nozzle. With this information and some tinkering with a microcontroller and a lab power supply set to 20 V, it was possible to discover the exact workings of the cartridge. In the end it appeared that the cartridge is divided into eight sectors with 17 nozzles each. The print head itself consists of two rows of holes. Each row of holes has four sectors. And each sector has its own power supply pin. When 20 V is applied to this the nozzles in that sector can be activated. Whether they actually do this depends on the nozzle inputs. When a positive voltage is applied an actual drop of ink will come out of the nozzle. If no voltage is applied to the nozzle input then nothing happens. The next table shows the positions of the various inputs, seen from the back of the cartridge with the printing end pointed downwards.
| Gnd | Gnd | Gnd | Gnd | |
| C1 | R | R | C | |
| R4 | Gnd | Gnd | R1 | |
| C2 | R | R | C | |
| R0 | Gnd | Gnd | R5 | |
| C3 | R | R | C | |
| R3 | Gnd | Gnd | R2 | |
| C4 | R6 | R7 | C |