Hacking the HP45

Powder and inkjet printing
acastel
Posts: 9
Joined: Tue Aug 25, 2020 7:11 pm
Location: Brazil

Re: Hacking the HP45

Post by acastel »

The pulse duration of the primitive line is also much longer than that of the patent, which is 1.875us.
I guess the primitive can be between 1.9μs - 2.3μs, depending of the head temperature.

Dragonator, do you have a waveform on oscilloscope of the output primitive with the TLC59213 driving it?
User avatar
dragonator
Site Admin
Posts: 484
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Hacking the HP45

Post by dragonator »

I do think that the primitive time can be assumed to be constant. I have no reason to believe otherwise with the C8855 manual. Given the high power nature of the signals, giving us rise and fall times that are significant, and the measuring hardware, some variance is to be expected/. This seems to be confirmed by the fact that the warmup pulses are also varying by 1-2 dots or 200-400ns and the addresses overlap sometimes. The printhead varies the power of the head by varying the voltage, not the pulse time. This is at least the case according to the manual.

There are a few things that do stand out. The first is some chatter over the address lines before printing and in between the calm areas. I have absolutely no idea what this is. The addresses are high for around 14us when pulsing, but only 3.8us when there is warming pulses. There seems to be jetting, then warming, and then more jetting. I do not know the context of then this was taken to know if this is some sort of calibration (which I do know occurs).

What was it that you were printing here. Is is possible to get a waveform of the printhead when it is printing something like text. You mentioned the idea that the printhead might actually give warming pulses when it was not printing, and that is an idea that seems intriguing. The waveform here seems to suggest all black.

I can get a scope waveform of the TLC59213. However this is of my setup, and not something HP original. I do still have a C8855 laying around here, which I might actually try at some point, but I have never tried to control this. I am getting a bit curious what is going on here. I have one unresolved issue that is a major headache. I cannot pulse all 14 primitives at once. I need to pulse 5,5 and 4. Else I am getting misfires. HP seems to manage just fine, driving the head as fast as humanly possible, all primitives on one address all at once. I have no clue whatsoever why they manage this and I do not. The idea of warming pulses when not firing seems reasonable, but this will be a bit hard for me to implement right away. Fixing this issue will effectively triple my printing speed at once.

I cannot promise a waveform tomorrow, but I will try to get a scope to my controller in the next week.
math
Posts: 27
Joined: Tue May 24, 2016 1:08 am

Re: Hacking the HP45

Post by math »

dragonator wrote: Mon Sep 14, 2020 6:32 pm I have one unresolved issue that is a major headache. I cannot pulse all 14 primitives at once. I need to pulse 5,5 and 4. Else I am getting misfires. HP seems to manage just fine, driving the head as fast as humanly possible, all primitives on one address all at once. I have no clue whatsoever why they manage this and I do not.
I think this problem is due to the simultaneous supply capacity of the tlc59213, which is a maximum of 200mA when all 8 outputs are activated at the same time. To generate enough energy for the ejection of the ink, you need 400mA for 1.875us, according to the patent and c8855m.
farukbay
Posts: 1
Joined: Sun Oct 25, 2020 7:12 pm

C8855 documentation

Post by farukbay »

Hello Dragonator,

I have a project for custom printer, and i want to start with HP45,
i want to start with C8855 manuel, is it possible to share the C8855 manuel.

Thanks
Last edited by farukbay on Sun Oct 25, 2020 7:49 pm, edited 1 time in total.
User avatar
dragonator
Site Admin
Posts: 484
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Hacking the HP45

Post by dragonator »

I have mailed the manual to you. Good luck on your project.
acastel
Posts: 9
Joined: Tue Aug 25, 2020 7:11 pm
Location: Brazil

Re: Hacking the HP45

Post by acastel »

Hello everyone, sorry the delay.
dragonator wrote: Mon Sep 14, 2020 6:32 pm I do think that the primitive time can be assumed to be constant. I have no reason to believe otherwise with the C8855 manual. Given the high power nature of the signals, giving us rise and fall times that are significant, and the measuring hardware, some variance is to be expected/. This seems to be confirmed by the fact that the warmup pulses are also varying by 1-2 dots or 200-400ns and the addresses overlap sometimes. The printhead varies the power of the head by varying the voltage, not the pulse time. This is at least the case according to the manual.
I think the primitive time is constant, but it can change in different printer model. In manual on page 18 we have - CAUTION: DO NOT exceed 22.5Vμs input over the specified fire pulse width range of 1.9μs ≤ tpw ≤ 2.3μs when driving a 51654A-type print cartridge. So you can change the primitive time and voltage since you do not exceed 22.5Vμs. In my printer, with an 100MHz osciloscope, I see a primitive pulse about 2,2us and 10,0V, and Address about 11,7V during printing. On digital analyser the timing is the same.
dragonator wrote: Mon Sep 14, 2020 6:32 pm There are a few things that do stand out. The first is some chatter over the address lines before printing and in between the calm areas. I have absolutely no idea what this is. The addresses are high for around 14us when pulsing, but only 3.8us when there is warming pulses. There seems to be jetting, then warming, and then more jetting. I do not know the context of then this was taken to know if this is some sort of calibration (which I do know occurs).

What was it that you were printing here. Is is possible to get a waveform of the printhead when it is printing something like text. You mentioned the idea that the printhead might actually give warming pulses when it was not printing, and that is an idea that seems intriguing. The waveform here seems to suggest all black.
The temperature control and warming pulses work when we are printinting. But when the printer is in stand-by and we open the printer options on
Windows, the printer start to control the temperature. I saw warming pulses every 2 minutes, when the printer options window is open. I am attaching the waveform I got during printing some patterns from CorelDraw. Now you can see very clear the warming pulses, print pulses, etc. The pattern is seven black blocks with 12x3mm.
dragonator wrote: Mon Sep 14, 2020 6:32 pm I can get a scope waveform of the TLC59213. However this is of my setup, and not something HP original. I do still have a C8855 laying around here, which I might actually try at some point, but I have never tried to control this. I am getting a bit curious what is going on here. I have one unresolved issue that is a major headache. I cannot pulse all 14 primitives at once. I need to pulse 5,5 and 4. Else I am getting misfires. HP seems to manage just fine, driving the head as fast as humanly possible, all primitives on one address all at once. I have no clue whatsoever why they manage this and I do not. The idea of warming pulses when not firing seems reasonable, but this will be a bit hard for me to implement right away. Fixing this issue will effectively triple my printing speed at once.
About pulse all 14 primitives at once, I guess the math post is right.

During this time, I made some tests and printing here using the L293 and now using a mosfet transistor to do the primitives pulse. I have success on that, but for some reason, I burned out my cartridge. Probrably I exceeded 22.5Vμs, and I guess the voltage and the timing have a very tight requirements. I am attaching some pictures.

Now I´m making an PCB to do more tests and try to control the temperature during printing, eliminating parasit indutance from long wires.
Attachments
Teste para captura_12mm.jpg
Teste para captura_12mm.jpg (71.13 KiB) Viewed 57 times
Setup2.jpeg
Setup2.jpeg (70.03 KiB) Viewed 57 times
Setup.jpeg
Setup.jpeg (135.91 KiB) Viewed 57 times
Impressão Teste.jpeg
Impressão Teste.jpeg (122.23 KiB) Viewed 57 times
17_09_2020_1_Teste Impressao Barras 3mm x 12.zip
(217.22 KiB) Downloaded 3 times
acastel
Posts: 9
Joined: Tue Aug 25, 2020 7:11 pm
Location: Brazil

Re: Hacking the HP45

Post by acastel »

This is my primitive pulse using P channel mosfet.
Attachments
Primitive Pulse mosfet.jpg
Primitive Pulse mosfet.jpg (461.76 KiB) Viewed 57 times
math
Posts: 27
Joined: Tue May 24, 2016 1:08 am

Re: Hacking the HP45

Post by math »

acastel wrote: Sun Nov 22, 2020 12:11 am

During this time, I made some tests and printing here using the L293 and now using a mosfet transistor to do the primitives pulse. I have success on that, but for some reason, I burned out my cartridge. Probrably I exceeded 22.5Vμs, and I guess the voltage and the timing have a very tight requirements. I am attaching some pictures.

Now I´m making an PCB to do more tests and try to control the temperature during printing, eliminating parasit indutance from long wires.
I think that a good way to evitate cartridge burning is to limitate via circuit timing of primitives. In this page are a circuit explained that can do this, maybe:

http://spritesmods.com/?art=magicbrush&page=4
User avatar
dragonator
Site Admin
Posts: 484
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Hacking the HP45

Post by dragonator »

I can't really complain about delays given the number of posts I have made the past month.

The 22.5Vμs is a number I am seeing the first time, but it does mean that I run within spec (roughly). In my experience exceeding it by a bit is not an instant disaster, though I can imagine that it will shorten the life of the printhead.

I am not entirely sure yet that Math is right completely, but I do think more experiments to verify his point are warranted. I myself have tried using more TLC59213's without success. If it was purely from the TLC's, 2 channels per primitive should have helped. From what I understand, at least part of the issue with the current limit is to do with heat dissipation. There is probably a current limit somewhere in place too, though I have not been able to see a voltage drop on the TLC59213. Another thing I am curious about, which is much easier to test, is to limit the voltage on the primitives to 10V. I do not think I ever tried it. There is a chance that at 12V I am oversaturating the head with heat, causing primitives to fail to fire. I have separate primitive and address voltages exactly for this reason, testing it should be simple.

I was aware of the circuit SpriteTM used. I did want to play with the Canon heads he was using. However, I have since moved my primitive triggering to DMA on the Teensy, eliminating much of the risk. I have never had any head fail because a primitive was stuck on, and with DMA it should be practically impossible without major firmware errors.
math
Posts: 27
Joined: Tue May 24, 2016 1:08 am

Re: Hacking the HP45

Post by math »

I'm looking forward to seeing your 10V tests on the primitives!
Post Reply