Hacking the HP45

[DigitalShadow]

The two patents supplied at the beginning of this thread are great, but have precious little information on the thermal control circuitry. By spending some time referencing patents made around the same time period I came up with a few that may or may not be relevant to this particular print head, but they seem similar enough to get me thinking in the right paradigms. Here's a few I thought may be useful

-Determining the operating energy of a thermal ink jet printhead using an onboard thermal sense resistor

-Control circuit for regulating temperature in an ink-jet print head

-Thermal ink jet print head and temperature control apparatus and method

-Thermal turn on energy test for an inkjet printer


That last one has what looks like a good description of the function of the 10x resistor:

The integrated circuit printhead of the thermal ink jet printer of FIG. 1 further includes a sample resistor 21 having a precisely defined resistance ratio relative to each of the heater resistors, which is readily achieved with conventional integrated circuit thin film techniques. By way of illustrative example, the resistance sample resistor and its interconnect circuit are configured to have a pad to pad resistance that is the sum of (a) 10 times the resistance of each of the heater resistors and (b) the resistance of an interconnect circuit for a heater resistor. One terminal of the sample resistor is connected to ground while its other terminal is connected to one terminal of a precision reference resistor Rp that is external to the printhead and has its other terminal connected to a voltage reference Vc. The junction between the sample resistor 21 and the precision resistor Rp is connected to an analog-to-digital converter 24. The digital output of the A/D converter 24 comprises quantized samples of the voltage at the junction between the sample resistor 21 and the precision resistor Rp. Since the value of the precision resistor R.sub. p is known, the voltage at the junction between the sample resistor 21 and the precision resistor Rp is indicative of the pad to pad resistance of the sample resistor 21 which in turn is indicative of the resistance of the heater resistors.

As discussed more fully herein, the sample resistor 21 can be utilized to determine the pad to pad resistance associated with the heater resistors in order to determine the energy provided to the heater resistors as a function of the voltage VP and pulse width of the voltage pulses provided by the driver circuit.

[DigitalShadow]

While spending much time reading to determine weather or not a patent would be relevant, I did come across some interesting points. They seem to use 12 volts for the address line and a variable power supply on the primitive lines. The voltage is up to 12 volts but usually around 9. Current temperature is a factor in setting drive voltage. Separate voltages may be used for warm up or firing. In our current approach, the voltage regulation seems to be done by modulating the pulse width at a constant voltage.

They also specified using separately regulated power supplies to decouple interference between address and primitive lines, as that is apparently known to be an issue. Perhaps try a separate 9-12v supply (Instead of an inline capacitor) for your primitive drivers and see if that helps?

[dragonator]

That is a lot of information to take in. Initially (at digitalshadow's first post today) I was a bit skeptical about the influence of temperature, but I recon it makes sense now. I do not know if it is the reason why my current setup does not work yet, but it is most definitely a good thing to keep in mind when designing a better system.

I have not yet continued with the experiments. I want a few hours of undistracted time, which I do not have on weekdays. The first thing I wanted to test is continuity by adding a capacitor and a pulldown on the ground. If the capacitor charges (visible on the scope), I at least know power is going through the heaters. judging from the scope, I cannot see that much interference. I am only using one address and one primitive. If I do get to the full setup, I will split primitive and address power supplies. For now I will install capacitors.

I do have a regulated power supply I can use to get anything between 9V and 12V if that is what is being suggested. I will keep the 12V on the address line. I still have some trouble with the legal English and what the 10x does exactly. The 10x is used to determine the resistance of the pad, which in turn is used to determine the resistance of the printhead? They want to exactly determine the power going to a nozzle? Do you think they do these measurements on the fly as well, to modify the inkjet parameters real time? This is getting more and more advanced as I type. The C6602 was peanuts compared to this.

I have this event next week, I am meeting the man that has given me most information so far:
https://ultimaker.com/en/community/1957 ... ruary-22nd

If there are any question specifically that pop up, post them and I will try to ask them.

I am more and more beginning to think about buying an HP45 printer. They are a decade old, but they are still available second hand. By then using scopes and logic analyzers I might be able to completely figure it out. First I want to try a bit more though.

[DigitalShadow]

I am still trying to wrap my head around it all as well heh heh. While both the thermal sensor and 10x are both heat sensitive, it appears they may serve different purposes. I think the 10x resistor is used to measure the resistance of the heating resistors, while the temperature sensor, well, senses temperature. Apparently the "turn on energy" required by the print head can vary between cartridges and can change over a cartridges lifetime. I believe the 10x resistor is used in a calibration/warmup cycle prior to printing, while the temp sensor is used while printing. By determining the lowest possible useful power level, they are greatly increasing the lifespan of the print head

[dragonator]

Game plan for this weekend.

I want to take 2 steps back and get some initial data on the printhead. Last weekend I was basically banging rocks together. I connected everything in the hope that it would work. This weekend I am trying to get some data on the printhead.

There are 2 things I want to know before I continue with getting ink to eject. First I want to get capacitive data on all the contacts. I want to see the difference between the primitives (little capacitance) and addresses (more capacitance). My LC meter is out of power, so I might first try it with the scope and some resistors. The pin with addresses should have higher capacitance than pins with primitives.

The second thing is nozzle continuity. I want to attach a capacitor to the ground with a resistor, and them measure with the scope if the capacitor gets charged when I fire a nozzle.

If I have time after that, I might try a few other things, like measuring the temperature resistor and trying (again) to get nozzles to fire, but one step at a time.

[ezrec]

Sounds like a reasonable plan.

I look forward to your results.

[dae43]

Hello!

We are working on the HP45 like you are, effectively trying to control it to serve the purpose of controlling individual nozzles. For a couple weeks, we have been keeping up with your progress, and we saw that you were considering buying an HP45 compatible printer.

In our efforts, we have already purchased and dismantled an HP deskjet 990sce professional series printer, and we have an oscilloscope that has a very high frequency, 500MHz. We plan to connect wires to the printhead's circuit board, and read the signals generated by the printer through the oscilloscope. Below are images of our printer and the accompanying printhead and our oscilloscope.

We would like to offer you help and possibly work with you on the endeavor to control the HP45. We do not have very expansive knowledge in computing/circuitry/electronics, but we think that we can still help each other on this project. We have past experience using Nicholas C Lewis's InkShield setup with the HP C6602. We are looking to the HP45 to be able to sustain a higher resolution and thus, a smaller individual ink droplet size.

We hope to do the signal test from the printer next week, and we can post our results here afterwards.


Image of the front of the printhead from the printer.


Image of the circuit board behind the printhead from the printer.


This is the dismantled printer.


This is the oscilloscope we are using - Aigilent infiniium, 500MHz.

[dragonator]

Hello and welcome.

I (We?) indeed seem to be working on the same thing as you are. If you want to do the measurements, I would be extremely grateful. I am not yet at a point where I have give up with my experiments here, but getting a proper readout from an original HP45 printer would save me a huge amount of time. At this point I am fairly certain that the pinouts on the first page of this forum are correct, after this weekend I should be sure. That may help you narrow down what to connect to.

We are after the same thing, so I would be happy to collaborate. I am more of a mechanical man myself too, but I do have enough knowledge to make electronics if I completely know how the HP45 works. Even better is that most other people on this forum know much more than me about electronics and can probably do a lot better than me if they want to. And if all else fails I have the combined knowledge of the Tkkrlab (hackerspace Enschede) available. Electronics should be possible after the printhead is hacked.

How many channels do you plan to measure at once?

[ezrec]

I would suggest that for your application, you focus on the ribbon cable to the printhead controller, personally, instead of the printhead itself.


I did some prior work for the HP F4480, which seems to use a similar controller, but my scope was not up to the job:


http://www.evillabs.net/index.php/Proje ... Controller

[DigitalShadow]

Welcome to the project! looks like you will be able to get some great data with that

Something of note, if my understanding of the patent is correct, the address lines are connected to both the primitives, and an "identification circuit". This identification circuit gets its inputs from the address lines, and is also shared with the temperature sense lines in an effort to reduce pin count. It seems that low voltage is used to poll temperature queries, and high voltage is used to poll information regarding the current cartridge, such as manufacturing data and such (Confirmation of this is still required. Getting this infromation from this patent from the beginning of this thread, last few paragraphs)

Also, my PCB's arrived today. They look great Still waiting on the pogo pins and driver chips to arrive. I will likely focus on temperature control and monitering. Likely submerging the tip of the print head into an appropriate liquid solution and controlling the bath's temperature to a known value.