Page 1 of 37

Hacking the Xaar 128 printhead

Posted: Sat May 07, 2016 6:16 pm
by Wonko
Well, since we all have spent so much time on stupid print heads, I have decided to reverse engineer the protocol of an industrial printhead. Sure, it's 200 Euros for a single head in a single color at only 200 dpi, but these babies are industrial strength, giving them supposedly a much longer life span and the ability to print solvents via piezzo, and hopefully UV resin as well.

To make a long story short: yes, I made it work. I can control each and every nozzle at full speed with an off-the-shelf Arduino and close to zero additional parts. Yay!

Re: Hacking the Xaar 128 printhead

Posted: Mon May 09, 2016 11:08 am
by dragonator
I haven't worked on the HP45 for a while and was actually picking it up again for this week, but I am very interested. Is the Xaar 128 capable of printing conductive fluids or are water based inks and binders out of the question?

With technology like this available you can make your own Objet like printer. Are you willing to share more information on how you did it? I would love to try one on either a Powder/binder printer, powder/UV resin printer or a pure UV resin (Objet) printer.

Re: Hacking the Xaar 128 printhead

Posted: Mon May 09, 2016 11:08 pm
by Wonko
I have not found anything about conductive ink, neither yes or no, but read several times that "only Epson can do that". I am not sure and I have not tried yet. The setup is pretty fragile. I did try solvent based ink and it worked very well. Ordering some UV curable ink for China right now to create my hobby-Objet next.

When checking the pins with the scope in a working printer, I found to my great surprise that all communication is done uncompressed via 5V TTL level SPI interface. I'll draw some fresh diagrams and post them here in the next days.

Re: Hacking the Xaar 128 printhead

Posted: Tue May 10, 2016 6:06 am
by dragonator
That would be really appreciated.

I have talked with someone who also did some work on inkjet (someone working for Ultimaker) and I remember something about piezo not being able to print conductive fluids. I hope that distilled water and Isopropyl do not conduct so this can still work on binder jetting in powder. I also have someone who asked me to do an experiment with water in cement, so I hope distilled water will have enough resistance.

I am sure that this will show itself in the future, but where does the ink enter the printhead. It is easy to connect to something like a tube for CISS?

Re: Hacking the Xaar 128 printhead

Posted: Tue May 10, 2016 9:51 pm
by Wonko
Ink just comes through a 3mm hose from a tank that should sit a few cm below the head. It does not seem to be too critical. But for many inks, the printhead should probably be flushed if not used for a while. White UV ink can not remain in the head for more than 8 hours supposedly!

One article claimed that Epson heads can deal well with conductive ink. Water should be no problem in any case IMHO.

Re: Hacking the Xaar 128 printhead

Posted: Sun May 15, 2016 2:21 pm
by Wonko
Funny to see how similar our progress is recently. I managed to get patterns printed and also to test all nozzles. Here is a little video:



https://youtu.be/dnBx0aCWNms

Re: Hacking the Xaar 128 printhead

Posted: Sun May 15, 2016 2:49 pm
by dragonator
That is amazing to see. Are the few missing nozzles broken or is something different happening.

Both printheads are great for different reasons. With a working Xaar head you can do different things than with a working HP45. HP45 is cheaper, more accurate and has more nozzles but only deals with volatile fluids. There are several things I would want to do with a Xaar. An Objet like printer is one of them, but also a wax printer (for PCB masking and wax models) is on the list.

Can't wait to hear more on how it precisely works.

Re: Hacking the Xaar 128 printhead

Posted: Sun May 15, 2016 8:26 pm
by Wonko
The nozzles were clogged from the solvent ink that I was using. I bought an ultrasonic cleaner yesterday which works miracles. I also ordered some UV curing ink from China to try exactly that. Next week's goal will be to mount the head in an XYZ-axis machine so I can print layers. UV ink s perfect for masking PCB's, and with XYZ-axes, etch-resist should be no problem.

The head is somewhat expensive at 200 Euros, but it is extremely versatile and does not heat up the fluid at all. I have pretty detailed descriptions on the Epson heads (also a Piezo head), which are starting at 100 Euros for 5 colors if you rip them out of a cheap printer, but controlling them is a few levels harder.

It's really great to see the HP progress. I'll keep you posted. Can't wait to get powder and jet 3D printing to work ;-)

Re: Hacking the Xaar 128 printhead

Posted: Thu May 26, 2016 8:38 am
by Philipp
Hello Wonko,
XAAR print head looks just lovely. Can you vary with your hack the waveform of Piezzos?
To my knowledge, you need relatively high voltages for driving the Piezzos, How did you do that?
Can you also produce greyscale with the print head?
Best wishes
Philipp

Re: Hacking the Xaar 128 printhead

Posted: Thu May 26, 2016 1:46 pm
by Wonko
Hi Philipp,

The head has only a single waveform that it generates internally. It needs well regulated 35V via two separate lines. The droplet size is fixed, but at 40pl or 80pl quite small. I can fire at 4kHz, so a decent rasterisation should be possible.

I am currently adding the head to my printer to sync it up with the x axis stepper motor. Looking great so far. I will try to get a holder made on the weekend, so that I can print a test page next week. The biggest issue is cleaning that thing, so it doesn't clog up. Those heads are 240 Euros a pop!