HackIt: Why Aren’t We Hacking On The LED Printer?

Strings of LEDs are a staple of the type of project we see here at Hackaday, with addressable devices such as the WS2812 in particular having changed beyond recognition what is possible on a reasonable budget. They’ve appeared in all kinds of projects, but are perhaps most memorable when used in imaging projects such as screen-like arrays or persistence-of-vision systems. There’s another addressable LED product that we haven’t seen here, which is quite a surprise considering that it can be found with relative ease in junk piles and has been on the market for decades. We’re talking about the LED printer, and the addressable LED product in question is a very high density array of LEDs the width of a page, designed to place an image of the page to be printed on the toner transfer drum.

Inkjets, Lasers, And Now LEDs? How’s That Work Then?

OKI LED print heads
OKI LED print heads

LED printers share their basic operation with photocopiers and laser printers, in that they use an electrostatic process to create the image on paper. A photosensitive drum is electrostatically charged before being exposed to brightly lit image of the page to be printed. The electrostatic charge is dissipated where the strong light hits the drum, but remains where there is no light. As the drum passes through a powdered toner it picks up the powder wherever it is charged, resulting in a copy of the page in toner on its surface.

The toner is then transferred to a piece of paper and heated such that it sets into the paper to produce a permanent print. On a photocopier, a lens and a bright light project the document, while on a laser printer the drum is charged by painting the laser dot line by line with a spinning mirror. The LED printer meanwhile exposes it one line at a time by shining light from a very long row of LEDs, and offers the advantage of lower cost and complexity than a laser printer with traditionally a slightly lower quality print. For this last reason it’s the laser that has captured the public imagination instead of its cheaper cousin, and though LED printers have remained available since they first appeared in the 1980s they remain something of a poor relation and their marketing often fails to mention LEDs at all.

Depending on the manufacturer the LEDs can be a variety of wavelengths, with the most common ones being either near-infra-red, or visible red. The LED arrays usually contain all their driver chips and are driven serially, leading to the possibility that they could be adapted for other purposes.

Getting Your Hands On All Those Lovely LEDs

It’s easy enough to find an LED printer array by dismantling a scrap printer, but a surprise is that they seem impossible to find as components in their own right. This is almost certainly because they are manufactured as parts for a particular printer mechanism rather than as parts in their own right, so unless you happen to be the manufacturer of that mechanism then there is little demand for them. Having salvaged one and been lucky enough to find that it has a visible wavelength though, how might you use it? An obvious application would be in very high resolution POV displays. They offer hundreds or even thousands of dots per inch, a resolution that a string of Neopixels can’t even come close to matching. Or how about using them for light painting, in this even an IR array could be pressed into service given a camera with sensitivity in that range.

So come on then, Neopixels are getting stale and we’ve seen almost every conceivable application for them before. Show us an LED project we haven’t seen before, one with breathtaking resolution! Our tips line is open, as always. To get you started we’ve put a video below the break with an intro to LED printers, then there’s OKI’s introduction to the technology, followed by this white paper from them talking about the LED arrays, and finally for a few more pointers take a look at this EEVblog thread.

Header image: OkiUkraine [CC BY-SA 3.0]

52 thoughts on “HackIt: Why Aren’t We Hacking On The LED Printer?

      1. not anymore, CCD’s are sensitive to IR and were used in older camera’s and phones. Nowadays it’s mostly CMOS sensors and they don’t do IR (checked with my iPhone, but correct me if i’m wrong).

        1. No, CCD and CMOS have about the same response frequency wise. iPhones just have very good IR cut filters compared to some cheaper cameras.

          My 5D Mark II is cmos and modified to “see” infrared.

    1. Just reading the article it seems they have found that most cheap LED printers are actually red or near-infrared, so that would be visible to the eye, leave alone an unfiltered camera such as most phones.

  1. It’d be interesting to try and get one in red and try and make some sort of giant virtualboy (assuming the mechanism could scale or be redesigned to embiggen), completely pointless but cool nonetheless.

  2. why aren’t we devloping open source photo and film negative scanners? they all seem to be working on the same system, but companies seem to reinvent the wheel with every new scanner released. new drivers, new scan software, but no improvements whatsoever.

    A scanner is nothing more than a lineair analog ccd, a programmable opamp, a fast AD converter and a fpga and ram for the processing.

    just to throw up a ball…

    1. Using red or infrared would be terrible for a scanner, especially a film scanner. I designed a professional film negative scanner for a client and in addition to software filters, we needed a light source with a filter to reduce red because with white we would have r255 while green and blue were in single digits. Ended up filtering off some green as well, and even then it looked wrong without software correction (but at least there was enough data for the software to work with). Also couldn’t use a linear CCD; couldn’t find anything with the dimensions and resolution required.

  3. Considering the cost in time to find one of these printer, tear it down, and perform the reverse engineering work versus the worst case of getting a board made and assembling it with new components.
    What is the price point where the trade-off in time and the lower cost of the junk makes more sense?

    Also consider that these LEDs are not meant for visible light applications. For that issue, this article might be looking to crowd-source some ideas. It seems they would only be useful in combination with the electrostatic drum, or a camera without an IR filter.
    Using the drum seems to fall under the category of other printer hacks. The 2 camera cases I can think of are a supplemental light source for night vision cameras and maybe some experiments in poor mans li-fi.

  4. I think the answer is that because they don’t come free with your first sample of a teaspoons worth of $20,000 a gallon ink, like all those inkjets do, giving a false impression of the ease of picking up any type of junk printer.

      1. Lasers yes, they’re the only ones worth selling, our equivalent of gumtree has 1000+ printers listed within 50km of me and the vast majority are lasers, LED gets zero hits in that. At the thrift store and junked end I’d guesstimate seeing 100 inkjets per laser, and I’d figure LEDs per laser would have to be less than 20:1 … If I had to choose between turning up a 24 pin dot matrix and a LED on time limit for a bet, I’d go with 24 pin dot matrix and I’ve only seen one about every 2 years since the turn of the millennium.

      2. That may be the case in the UK and perhaps other countries.

        However, here in Australia, I haven’t seen an OKI in the wild for so long that I thought they were no longer made.

        After a quick search I see there still in production.

        They’re cheap to buy but not really cheap. The cartridges are about the price of a really cheap laser.

        There are a small number of second hand ones on eBay but they are being sold as working units so the price isn’t attractive.

        It may be possible to buy cheap knockoff LED heads from China but for that to be useful you would first need a working unit to decode and publish the hardware protocol, which shouldn’t be hard to do.

        A single axis CNC dragging a LED head over photosensitive PCB might be useful. With good registration double sided would be easy. But who is still stuck in the days of photo etch.

        1. photosensitve PCBs need UV light, not IR.

          What do you mean with “But who is still stuck in the days of photo etch.”? Photo etching is still the way PCBs are made, nearly all fo them.
          There is a laser machine from LPKF, which can ablate the copper directly, but for plated through holes you need chemistry anyway. Plated holes with conductive epoxy are not reliable and have bad conductiivity.

  5. These (OKI Data) really weren’t a cheap printer.

    They were targeted at the commercial sector and not the small office or home market.

    For that reason they are uncommon.

    The two featured the attracted purchases were that they were very quiet and very clean with very low particulate matter leakage.

    They arrived on the market when dot matrix was still common and before laser printers became cheaper. Once lasers became cheaper, LED printers lost market.

    The Phaser printers were never cheap. They were a high end printer.

    1. I’m a *huge* fan of Oki’s LED printers – I’ve had several over the years, and just recently upgraded, not because the old one stopped working, but because the old one didn’t have a duplexer and even the new cheap ones come with that now. BTW – because of Oki’s excellent polymer toner, you can print on “weatherproof laser label” stock, and have it last for years even outdoors. I’ve got labels like this made for preproduction IoT equipment that are still readable after nearly 20 years. The polypropylene label material itself is breaking down faster than the toner! (That’s black – colors fade a lot in about a decade, but we used black barcodes for the serial number/MAC address, and even those printed in 2003 are still readable!) . FWIW, I’m not particularly brand-loyal, but I can’t for the life of me imagine buying anything but an OKI LED printer for any LED/laser/Inkjet application – they’re truly the best there is, and unless you need to print white, not even more expensive!

  6. Near IR can be used to enhance the visibility of veins through the skin. Very useful for phlebotomists and others who have to tap veins for blood samples, etc.

    Recently, with the opoid epidemic in full swing, some places with publicly accessible restrooms have been installing blue lights to prevent drug addicts from shooting up- the blue light makes it almost impossible to find a vein…

    1. ” places with publicly accessible restrooms have been installing blue lights to prevent drug addicts from shooting up”

      I’m sure that people with legitimate reasons to inject themselves really appreciate that. I know migraine medicines work a lot better if they go into a vein and often not at all otherwise. But hey, as long as they are doing it to protect somebody from their own bad choices it’s good to harm innocents right?

      1. I’m pretty sure the main driving force is to ensure the staff of the establishment in question don’t have the harrowing task of discovering the corpse of the former addict in the sh*thouse, rather than to protect the addict who will inevitably simply go elsewhere.

    2. I recall reading about a medical imaging technique that uses light scattered through the body to image what is beneath the skin. Kind of like ultrasound, but with light. I’d think that thousands of precisely positioned near-IR light sources could work really well for such a project.

    3. They did this also in the ground floor toilets at the technical university in vienna in the 1990ies, because the nearby park was a hotspot for addicts.
      But I always wondered about the efficiency of that: Even before the invention of LEDs a small torch was not that expensive.

  7. The Canon LIDE scanners have a bar with a lot of red, green, and blue LEDs to illuminate the scanned document. Those things are common and cheap, especially since windows dropped support for most of the older ones. There is a very high density linear phototransistor array in them, too. Some are 9600 dpi x 8 1/2″, too.

  8. A big bar of IR LED’s could come in handy for some of the super cheap “night vision” gear that’s out there. IR commlink or as was mentioned, resin printer. Maybe an IR sensitive resin would be less nasty??

  9. I’m in Australia. I just bought a brand new Okidata dot matrix printer because my boss asked me to.

    He said there was something wrong with the one spitting out the finance cheques (three part paper) and that we need to buy a new one. I ordered a new one, but decided to check out the old one.

    It needed a switch flipped (tractor feed enabled), and was working fine again.

    So now we have a brand new Oki sitting in the IT room while the old one keeps going like a champ.

    1. Just buy toner by the kilogram on a chinese site. Or get one unused cartridge because they are everywhere as the sellers make money by not making them compatible with anything else in the universe, not even the other models of the same manufacturer. But then you pop the plug on the side and Bob’s your uncle. There’s nothing easier and cheaper than toner-based printers. You can also buy refill kits with instructions in case you need to reset an oddly-formed half-toothed gear.
      And yes, I have an 300 dpi Oki that has worked for *decades*. (Take that, inkjet!) Nobody wants to junk it. I guess we are just waiting to see what happens if we get nuked. Would buy it again, hands down.

  10. Before LEDs were used as alternative to LASERs in printers, QUME had another idea. Instead of precisely directing where the OPC drum is illuminated, precisely direct where it *isn’t*.

    Their CrystalPrint series used a long halogen light tube behind a 1 by a heck of a lot Liquid Crystal Display. Just like any backlit LCD the image shone on the drum was controlled by selectively blocking the light. IIRC the system was capable of 300 horizontal by 600 vertical dots per inch.

    It was an effective way of sidestepping any patents relating to using a LASER and how the beam was scanned and pulsed because the LCD shutter didn’t scan and the halogen light was on steady and obviously wasn’t a LASER.

    Had Qume come up with this sooner, when the patents it was built to break still had more years to go, it might have had some lasting effect on the industry. But with patents run out and companies other than the big players like Canon (who had their components inside many brands like HP and Apple) could make their own LASER based printers without paying license fees.

    Qume also would have had to come up with a way to at least double the horizontal resolution to 600 in order to stay competitive.

  11. “On a photocopier, a lens and a bright light project the document”
    when’s the last time you actually saw a photocopier that didn’t in fact scan and then print via laser printer?

Leave a Reply to andrewjhull Cancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.