Page 19 of 37

Re: Hacking the Xaar 128 printhead

Posted: Mon Apr 16, 2018 4:50 pm
by aengel
About wetting of inks, consider page 33 of

The Chemistry of Inkjet Inks, edited by Dr. Magdassi Shlomo. BTW, Dr. Shlomo does excellent research regarding inkjet and conductive inks. I have read a number of his scientific papers, and he and his students are top notch.

Snippets of this valuable book can be found at :

https://books.google.com/books?id=awnGC ... ng&f=false

Re: Hacking the Xaar 128 printhead

Posted: Wed Apr 18, 2018 3:06 pm
by aengel
Here is an image of the BOM for the Xaar Arduino assembly.

I am very concerned about the quality of the power supply. I am also concerned that this Arduino is too slow.
ytec xaar bom.jpg
ytec xaar bom.jpg (109.63 KiB) Viewed 16667 times

Re: Hacking the Xaar 128 printhead

Posted: Wed Apr 18, 2018 4:58 pm
by aengel

Re: Hacking the Xaar 128 printhead

Posted: Fri Apr 20, 2018 3:47 pm
by Restani
Hi,
I'm still trying to make an impression with the XAAR128.
I'm using the files provided in the Kyle project but all I got was firing all the nozzle at once.
I try to shoot half the nozzle but it does not work!
Can you help me?
Thanks!

Re: Hacking the Xaar 128 printhead

Posted: Fri Apr 20, 2018 7:34 pm
by davidk
Check with a scope that you correctly select nSS1 and nSS2 before sending corresponding SPI data. Also check SPI lines and CLK.
Be aware also that if you have 3.3 CPU you will need a buffer since this head has 5V lines and won't read correctly high speed 3.3V signals (CLK, SCK and MOSI).


I had a very pleasant surprise, after I thought I destroyed one brand new head with water, I flushed it with isopropyl and now fires flawless. It sat almost a year with some water in it because I thought I lost it. On the other hand I got 16 working heads from a advertising company and flushed some of them with isopropyl also. I couldn't let these sit so I made a purge machine with a tray, a peristaltic pump and an ink filter. This way, I recirculate IPA trough a head until all nozzles unclogs and no ink is left in the head. I like IPA because is not so greasy and smelly. Do you think it can damage the heads? I suppose not since it resurrected a fully dead printhead, as I told you. Then I'll store the printheads in plastic bags until I will need them. Afterall, I have 16 of them and I aford testing various inks. I advise you to find advertising companies near you and ask if they have old Xaars!

Re: Hacking the Xaar 128 printhead

Posted: Wed Apr 25, 2018 12:01 pm
by davidk
As I said, I got myself a lot of used printheads. I am really annoyed by the very strong smell of the "true" solvent they use so I tried to find an alternative ink, as mild or eco-solvent. So, do you think there's a different kind of ink that may be used for testing? I got prints, all is OK but I have to do more testing until I'll have a working plotter. I can't use IPA as solvent to make my ink because it dries too fast from the nozzles. The solvent must be greasy, as the original "true" solvent they use.

Oh, and the original printer for Xaar128 doesn't have cappings, the whole carriage is parked onto a solvent tray. So the nozzle plates are always soaked into solvent when not in use.

Here's a picture of inside a Xaar128; the rectangle in the left of the two silicon dies is the integral mesh filter.

Re: Hacking the Xaar 128 printhead

Posted: Thu Apr 26, 2018 10:03 am
by davidk
This is beyond the mesh filter. I suppose the white thing is oxidation from water. This was an used printhead that I used with water. It has cyan ink so I suppose the white stuff must be oxids.

Re: Hacking the Xaar 128 printhead

Posted: Thu Apr 26, 2018 12:40 pm
by Restani
Hi,

Thanks for listening!
In relation to the lines nSS1 and nSS2 they are alternating correctly, CLK is in 1MHz and sometimes it undergoes a small change reaching to drop to 998KHz.
I'm using the Kyle project and in the xaar128.cpp file I made a modification to verify that the data was being sent correctly:

byte B [64];
   memcpy (B, val, 64);
   for (int i = 0; i <8; i ++) {
     SPI.transfer (B );
     Serial.println (String (B ));
   }

And on the serial monitor everything went well.
SPI is set to 1MHz. Shall I slow down?
In the video it is possible to notice that all the nozzles are acting, but I have cases where only part of the nozzles should be acting.
In the video you can also notice a card of an equipment that uses Xaar128, where I can extract the lines of 35V.

On paints, I use an ink that works normally on a device with xaar 128. This same equipment tb works with a UV paint that we have here.

https://www.youtube.com/watch?v=ZuHmwnu ... e=youtu.be

Re: Hacking the Xaar 128 printhead

Posted: Thu Apr 26, 2018 12:59 pm
by davidk
Phat is the device generating TTL signals? What voltage, I mean (5V or 3V3)?
Is the SPI data stable on the rising edge of SCK?

Re: Hacking the Xaar 128 printhead

Posted: Thu Apr 26, 2018 5:53 pm
by Restani
The Arduino Mega generates the TTL signals, works with 5V.
I can not see the SPI data. I'll try again but on the oscilloscope I can not see SPI data change.
In the image the upper channel is MOSI and the lower channel is in SCK

Image