It’s not “pretty much any inkjet”. There has to be room inside the unit to allow for the piping to the ink cartridges.
It’s not that much of a hassle either. Some systems have cartridges with the piping already attached, so you just run the piping bundle out of the printer. Mine uses the original cartridges, but I have to drill a tiny hole in each of the two cartridges for the piping inlet. Sure, it took 15 minutes of valuable time, but I’ve saved hundreds of dollars over the years with one simple mod. And I refill the tanks every three months or so.
Notice that I said “years”. I haven’t had to change a cartridge in a few years. Does that seem like a “hassle”, or an “anti-hassle”?
And please explain how a “better printer” makes you immune to problems with ink cartridges? Mine is a top-of-the-line Canon that I use for fine photo work.
That has to be taken into account. Possibly a shield with a small FPGA? Would have the advantage of serial data via SPI so simpler and less timing-sensitive wiring. And way more outputs.
Please elaborate?
I saw those. Was a bit demoralizing to see the demise of steppers. Thought: could said FPGA handle this? The optical gate (that slides against that grey looking semitransparent strip that at close examination reveals tight pattern of parallel lines) as input synchronizing the output clock for the head drivers?
Given the amount of such feedbacked DC motors in e-scrap, such module would have a pretty generic use…
The optical feedback systems are actually pretty cool(just as with HDDs, where steppers gave way to voice coils and feedback loops, steppers are nice and predictable; but you hit a wall where price rises rapidly, if you can buy them at all, if you need more than a certain amount of precision; while the feedback systems make you suck up control theory; but can do very impressive things; and ‘more precision’ requires little more than sheets of transparency film printed with opaque tick marks). The hardware isn’t really too bad, one or more IR optointerrupters that move across the printed plastic film, trivial to interface with; but you need the algorithm driving the motor controller to have nearly immediate access to the optointerrupter input, so a naive ‘just put the IR on a GPIO pin and a motor controller cape on top’ approach is likely to fail. Best bet would probably be a motor control board with support for optointerrupter and/or Gray-code rotary encoder input.
As for the PRUs, they are rather more fiddly to get up and running than most of the peripherals(which are just a ‘shove in the appropriate Debian ARM image and go’ affair); but you get two 200MHz CPUs running independently of the main system, each with a small slice of data and instruction RAM, some GPIO pins, and connection to the host system. They don’t implement any specific protocol in hardware, like an i2c host or a 16550; but are smart enough to handle most of the messy details of bit-banging what the situation requires, with zero interference from the scheduler of the host OS or load on the host CPU; and are more tightly coupled(as well as faster) than using a discrete microcontroller or the like to handle the situation.
In the case of a printer, they’d be good candidates for things like watching the optointerrupter inputs, allowing the host system to just specify where ink is supposed to be placed, or request the location of the print heads; without losing the plot if it has a timing problem while trying to keep track of how many times the optointerrupter has triggerd.
That’s what I’d suggest to throw a FPGA on, or a dedicated ATmega microcontroller (I’d take the latter way due to knowing them and not having a metric milliclue about how to program FPGAs).
The printers are fairly slow, though. Assuming 1000 dpi and 8 inch wide paper, this gives us 8000 events per head travel. Assuming 5 passes per second, we get some 40,000 events per second. Not that much. (Times the bytes per event for controlling the head. That can get quite higher.) Maaaaybeeee even the realtime FPGA or dedicated microcontroller may not be needed. Maybe a USB microcontroller with isochronous interface and some minor buffering could be used…?
You know what I like about those shitty plastic printers? When I tear them apart, they all have at least 1 nice stepper motor inside them! Yummy treat.
If you buy a wide format printer for professional archival work and use third party ink . . . well . . . you’re doing it wrong, and screwing your customers at the same time. Epson makes ink. They also make a printer to deliver that highly advanced ink to the paper properly.
I have never been disappointed by an Epson printer. They are friggin’ awesome and reliable. Easy to use and produce astounding prints. Expensive? yes.
Archival inkjet is the longest lasting color reproduction in history outside of a properly made oil painting. Fragile too. Oh well.
So I’ve seen… usually the brand new ones aren’t up for demolition. The older ones are quite yummy, all those rods, gears and wheels and stuff. I took apart one of those old motorized whiteboards once and found TWO super awesome Nema 17s.
Envy!!! The old good ones used to be like that. The cheap disposable inkjets of today, the kind you can find thrown out on the streets today, not anymore.
I know. Why is the world made of crap now? Fight Club was right. “You are the all singing, all dancing crap of the world.” And yesterday’s crap is today’s treasure. Remember those inkjet printers from the 90’s and early 2000’s? They were CRAP then, because we were used to laser and dye subs with their shiny, perfectly shaded, nearly embossed print. Then inkjets came along in their plasticy awfulness, with ink that ran in humidity, paper jams and shitty expensive ink cartridges that you had to replace after printing a few dozen glossy photos. And now, here we are, waxing nostalgic about the stepper motors in them, when the current batch of plastic crap doesn’t even have them anymore!! What a load.
Thus, my devotion to the whole maker culture thing. If I can’t make it or repair it with my own two hands, I pretty much deserve to have it turn to shit on me…
People should realize they have to demand schematics and source codes. They usually can’t read them - but they usually know someone who does, and who may be willing to donate or barter the work they need to get done on the equipment.
Also simple things like never toss the paper manual that comes with new equipment.
This is less in reply to you, and more just general exclaiming:
Sure, the directions might be in pidgin English and once you’ve put it together you don’t need them anymore. But 9 times out of 10, that manual also includes a schematic/ exploded part diagram or list of replacement parts & part numbers… THAT IS GOLD. It’s information that you often cannot find on the internet. Often you can. But in the off-chance that you can’t, better to have that piece of paper. Scan it and keep it in your archives if you hate paper. But keep that info!!! Saved my ass a lot of time and frustration hunting down parts!
I had an HP inkjet proofing printer in my printshop probably 15 years ago. After a few months I found out that you needed to replace all the print heads (at $50 each) as well as the ink cartridges (also $50) every six months or so because they had a chip in them which expired. It was an HP design jet or something, and cost 5grand at the time, and most shops were using EPSON inkjets for proofing which cost even more but were much more reliable.
(The key thing with all the inkjets is that you needed to print often otherwise they dried out) One supplier told me that whenever you did a self cleaning function you used up something like 20% of the ink in the cartridge.
Old HPs are indeed solid. I have 2 LaserJet 4MV’s that I loved, and can no longer use; toner cartridges are no longer manufactured for them, and the internal parts dry up and harden after a year or so out of the sealed packaging.
Yes, I’m aware there are remanufactured cartridges; I’ve yet to find one that prints acceptably. I replaced the 4MV with an 8150 – almost as bulletproof, and duplexes 11"x17".
