I’m looking into a possibility to drive a Xaar 128 printhead. It’s piezo-electric and can be used with oil-based, solvent and dye sublimation inks.
Online, I found a what they called a “Xaar 128 Guide to Operation” document and it looks like communication is done uncompressed via 5V TTL level SPI interface. The issue is that I never done any SPI communication and have no Idea how to convert minimal specification from a document to a SPI signal in netmf. So I’m looking for any help I can get.
This is not commercial project but it netmf driver for this printhead s created it could be used for many different commercial projects.
Looking at the images and video, it does not look like that fellow is using SPI - too many wires - looks like he is driving it via some sort of parallel interface.
EDIT: He does say he scoped a ‘working printer’ (not the printhead) and that the printer used SPI somewhere - but the print head sure looks to be parallel driven.
This is a pinout from the document, it looks like the communication is done through SPI. I can send you the pdf document. The other wires he has is for power management pins.
Well, I guess that wold explain all the wires. Seems to be more to it than just SPI though based on your comment regarding the power supply. I don’t know of any ready-to-run drivers, so this one looks to be something to start from scratch, unless someone else knows of a driver.
I suspect that there is not exist for this. But how hard would be to create one for someone who is familiar with SPI development? I do have " Guide to Operation" document from Xaar, I just don’t have knowledge in SPI communication.
There are (at least) two motivators to action : thrills and cash. Which one are you appealing to? Thrills: The love of the challenge (would depend on someone else already having a printhead or you to provide one) Or cash: Is this a paying gig?
“How hard” really only affects the amount of thrills or cash it will require to get the job done.
I think I figured out how control the printhead by looking into the document and protocol looks primitive.
[ul]Printhead contains 2 slaves, first slave controls nozzles from 1 to 64 and second from 65 to 128.
The SPI does not fires nozzles, It only used to pre-load data for nozzles that will need to be fired.(send one nozzle number at a time in descending order).
After data is loaded, nFire pin is used to execute nozzles according to it.[/ul]
It looks fun enough. Buy two heads. Send me one. I will give you an open-source driver in 30-60 days after I get the head or refund your money for both heads. (Timing constraints may require some or even most of the driver to be RLP code - no guarantees that this can be done all in managed code)
DM me for the shipping addr if you are interested.
Do you want to see the “Guide to Operation” document first? I’m kind of hopping that after seeing the document you will see how primitive the protocol is, you’ll change your mind on the timeline and instead of 30-60 days you’ll say 15-30 days
Oh, I know it will go way faster than that. Maybe 2-4 days to set up the hardware and do the work … but up to 30 days to find three days free.
EDIT: [You would be disappointed if I said ‘three days’ and you didn’t have it on the fourth day, so I gave you a very conservative schedule for both getting around to it and then actually doing the work. I could probably do it much sooner, and if I can, I will, but I don’t want to over-promise. So, call it 30 days at most]
