@@fredlllll This is why you buy broken or second hand. Use primary asset to fix trash and recycle. It may not be perfect afterwards but you will be ahead further than you were before. BUT at 45k, they should have very long warranties and broken ones will be sought after if the warranty is crap or exploitative with dumb high costs. When you see such high prices for an inspection tool, then you know it is probably overpriced.
I worked for EINK several years ago and the team I worked for mainly focused on creating the waveforms for each transition. Each grey had a waveform and there were also different waveforms for different temperatures. Because any variation in the chemistry or display assembly affected the transitions, each batch of panels required custom waveforms. I'm sure it's more automated today but the number of technician hours involved in making those earlier panels was crazy.
Is that why eink screens cost so much? (ie like the big ones in the remarkable tablet or sony dpt)... I always assumed it was just cuz eink had a monopoly on the market
Thanks! I often feel like I'll wait for a project to go from "in progress" to "fully finished" to avoid making multiple videos about the same topic, but then I don't make many videos at all. Also, still trying to find a good balance between super tech-heavy and short-form fun science. I like both, of course.
I feel the exact same way! I love how you do things Ben, you're a real inspiration. I hope that I'm able to have a shop like yours someday. Keep up the videos!
Applied Science, just do both, finding a balance is just a waste of effort when you have a pretty diverse audience. You can't please everyone all the time anyway
Such a cool technology. I love my ebook reader. The one that I have has a refresh sequence where it does minor refreshes every page, and then a longer full-clear refresh every 5 or 6 pages or something. It's also worth noting that the cells containing the charged pigments do not correspond to individual display pixels, but are smaller and more randomly clumped together. The pixels come from the electronic array underneath, and one pixel may activate particles in multiple or partial cells. The cell structure exists to ensure a good distribution of the different particle types across the screen.
It would be interesting to try a burn-in test with that display without the reverse cycle, e.g. 10 updates at one spot, 20 updates at the next and so on until the text is barely legible. Then apply full black/white cycles to see how long the burn-in takes to clear at each spot (or how much it fades after let's say an hour).
Yeah, I second this. The refresh rate of these is what prevents me from using them in projects. If someone else who knows more than I do could test one of these to destruction I'd be super stoked. I'd totally use these all the time if I could achieve 3 Hz refresh and be confidant that it'll last more than 3 days.
That's a great idea and gizmoguyar nails my interest in it too. Fast refresh rate such that you can interact with it just like a lcd like on a iPad, it becomes so so so much more usefulio
Since most e-Ink screens that are to be changed frequently are for consumer products (Kindle, etc) I believe the manufacturer has gone to extremes to avoid burn-in, which would prompt lots of purchase returns. Much easier and cost-effective to condition consumers to flickering every once in awhile than to random ghosting of text over text. Therefore cutting the reverse cycles shouldn't cause dramatic ghosting in a home-brew project, as long as you don't care about not having a pristine viewing condition each time.
i accidentally looked away during the dot moving period, then came back when he flip the screen color and said there should be a ghost circle burn-in: and i can't really see it? i do see some unevenness in color scattered in a small region, but i have to deliberately imagine it is a circle to acknowledge ahh it indeed is a circle then i rewatch the dot circling period of the vid and holy shit the burn-in become so freaking obvious!😂with shades differences!😂the later the more firmer!😂maaan!😂(it was burn-in in my retina?!?)
I have been following your channel for a long time and I always thought everything you were doing was way god-level above and beyond the work I do. But low and behold I just watched this video and this was exactly what I was working on! I was working on a 4 level e-ink display but it's very similar. The lookup table format in my controller works differently but close enough to at least be set on the right path. I learned a lot from this video and I am trying to reproduce your findings. I made a lot of breakthroughs inspired by your video. It was basically unusably slow before and now it's much smoother and getting better. Thank you very much!
Ben don't make us miss the best channel on youtube again! Also as someone who does C++ programming for living, I do recommend you getting absolutely any kind of proper code editor with syntax highlighting and/or auto formatting. Once you get used to it, it just takes way less time to read code that's properly formatted and displayed. Popular C++ editor choices for windows that are intuitive and work out of the box are: notepad++, sublime, atom, vscode(not visual studio) and many others.
This scope is unreal! You can see from the comments how many engineers are completely blown away by how much information the scope can capture and display at one. The team at Tek should be really proud for developing such a nice instrument!
Thanks for the video! I've got a few of these ePaper panels directly from Good Display in China (the manufacturer as far as I can tell).. they were happy to help with questions until I asked for a datasheet describing the LUT bytes, then they became very cagey. Nice work on finding a datasheet that has these details. All of their datasheets seem to be missing information and are bit inconsistent.
I applied to work for E-Ink Corp once. They never got back to me, but I did some reading on their tech while I was waiting. Very cool tech, and this video does a great job of showing it off
usually when people on youtube make "[something] hacking" videos, it's some clickbait douche presenting cursory google search instructions to change user-editable settings. this guy actually does real, intrepid work and does an extremely thorough job explaining it.
I'm doing my PhD on e-ink and it's a bit of a misconception that e-ink needs no power at all. The major problem with e-ink is the fact that there are charged inverse micelles in the cell, which over time will start to replace the colloidal particles. There's a lot of interesting scientific literature in this field actually!
Given your background then, what do you think is the most interesting / applicable technology for the 'ipad has baby with a kindle' application of easy on the eyes digital note taking
Fascinating.. Don't know which I'm more impressed with, the epaper & your ability to reverse-engineer what's going on, or that borrowed scope! With equipment like that, in the right hands, you could figure out how 90% of stuff ticks. Great video.
You should check out what the reMarkable paper tablet is doing. It's using an actual e-ink display, but updates fast enough for latency-free handwriting (directly on top of the display - it uses a stylus digitzer made by Wacom I believe). It's actually so fast that you can't detect any latency at all when writing or sketching. It's also very high resolution - 1872x1404. No flickering or updating on the display while drawing, or erasing. It's pretty amazing to use, I own one myself. It does occasionally perform a full screen refresh, but it's so fast that it's unintrusive - and it only occurs while you are in the menus or something like that, never really when actually drawing on it. You can also erase and such, but I guess once they got the drawing figured out, erasing was kind of a freebie. And it's true e-paper as well - it maintains the image even with the battery removed or what not. I have no clue how they made it happen. It must have incredibly fast response time - it's as smooth to use as a Wacom tablet on a PC, looking at a 60hz screen.
Super late, but looking at it's stats (10.3", 226dpi) it seems like it's using the ES103TC1 display. I'd get one myself if it didn't cost quite so much. I remember following a company for a while which was trying to create an eink tablet, it felt like they fell off the face of the earth but this might actually be them.
I can't believe how detailed and clear your explanations and demonstrations were. This was exactly what I needed to understand the limitations I was running into with e-paper. Thank you!
TIL there are multiple e-paper types. The one described here is electrophoretic. There also exists gyricon, where instead of a bunch of pigment particles, each cell contains a small bead that is magnetically polarized and colored, so that it spins in place to present the desired hemisphere. It occurs to me that one could have a sphere with colored poles and a stripe along its equator. You could then have electrodes along the display surface to orient the sphere to show one of its poles, and in addition have a electrodes in the middle of the paper's cross section, which would cause the sphere to orient so that its equator is visible at the paper's face. This would give a 3 color per pixel display. I hope that makes sense.
Do you think that would cause refresh rate to get worse or better or the same? That to me is really crucial to closing the gap with lcd note taking and reading like the iPad
Brilliant video! I have almost no experience with e-ink displays, so this really helped me learn how they work and how the controller works. I am planning to make a video about OLED displays, how they work, and include some microscope footage of the SSD1306 chip found in most OLED display modules. SO far I have got some microscope footage of the chip, but not enough to use in my video. The problem I am facing is that even with a bright light shining from the top, the chip is still too dark to see clearly.
The E-Ink display that has red in it is probably less complex than that. There is one normal pixel and a pixel filled with red cells right beside each other. Color E-Ink displays work like this but with three pixels: One cell is filled with red particles, Another is filled with green particles, etc...
Strange how RUclips protection wasn't triggered by that scope porn! On a more serious note, why all the fuss with cycling the display with AC voltage? Why not simply skip writing same "color" to a pixel, or in other words, only write pixels that do change?
I stripped down a monitor some time ago and saw a few of those silicon looking bars. I had no idea what it was but postulated it was some silicon device. Thanks for pointing out what it is. Some time ago you noted on a Q&A segment some of the reasons you made videos. A few months ago I watched that video and decided to start up loading content. Reasoning made sense to me.
I've been watching some of the progress on Twitter and I'm glad I got to finally see the video! Can't wait to see what cool projects arise thanks to this intense research and work!
For the love of all that is good, please burn wordpad. There are many beautiful text/code editors freely available; sublime text comes to mind. The black default background might help with some of the waffle texturing that is showing up in the video for whatever reason. Beside that, thank you kindly for the video(s), it is an honour to be able to see a true genius at work, and the world is definitely a better place for you sharing your time with us.
I still use wordpad, just because notepad doesn't handle CR/LF that's non-windows properly (wordpad does).. Wordpad is compact, fast, and gets the job done with no fuss. I don't use it for coding though, I use MinGWStudio for that (totally unheard of 3rd party GCC/MinGW IDE that's almost identical to DevC++, except way more stable).
Im fairly excited for "full" color e-ink displays. E ink has been teasing them for quite a few years now, and sony has shown off a few prototypes as well, though they seem to have a much wider color range compared to the E-ink ones. They are what i think they call EDP displays, apparently it uses low temperature a-si TFT's.
Finally! I wonder how this compares to what my Pebble uses though, I don't spot any flickering, but also don't have burnin, while it has been on my wrist nearly 3 years now.
I believe it uses a "memory LCD" made by Sharp. It consumes a very small amount of power (microamps), and has an appearance between normal LCDs and E-paper displays.
Ah that seems accurate as it indeed doesn't look like actual epaper (but I thought it maybe just was the backlight). It seems a better technology than the epaper stuff then, except for that it doesn't retain the image. I'll have to look at that tech up though. Thanks! Another video idea for you :P We need more of your videos anyways.
Thanks for the video. The eInk/ePaper displays are really nice. They will certainly be more popular with hobbyists when the prices come down a bit on the larger panels. I didn't know there were some with the extra red colour. That is the most impressive scope I've seen so far. 8 channels with up to 8 digital inputs each? Very nice and also very expensive.
Do you know why the larger panels are so damn expensive? It baffles me tremendously (some mentioned it's cuz of eink monopoly.. Others have said more individual tuning needed ;?)
I haven't had any permanent burn-in on any of my displays, but I've also been trying to avoid that. Maybe I should buy a few more and start with destructive testing!
Honestly I would like to see some destructive testing, be it displays or any future projects, it just scratches that "what if we take it to the limit?" spot. But yes, it's not the usual stuff this channel covers, I guess we have AvE and others for that.
EInk has _huge_ Electronic Ink Display Modules (with 31" 4096 color or 41" 16 color greyscale) for less than $2500; also tiny B&W modules for under $100. Here's a quote from their Website: "EC312TT2 is a reflective electrophoretic E Ink technology display module based on TFT active matrix with color filter design. It has 31.2” active area with 2560 x 1440 pixels, the display is capable to display 4096 colors depending on the display controller and the associated waveform file it used.".
144 is a lot more difficult to do than 3; for once you auctally need all things to sync up in the display 144 times a second; that's also why they were so expensive
I haven't played with e-paper displays yet but I'm a big fan of WaveShare's toys. I use their FPGA kits for all kinds of projects. You've mentioned permanent burn-in and/or damage - do you have any idea of what the possible mechanisms for that would be? I can't picture a way in which you could damage these cells even with strong polarization, based on how you've described their theory of operation. The only thing that comes to mind is some kind of cell overheating that could perhaps fuse some of those powder particles together which are presumably at least partially plastic (I'm thinking of laser printer toner), or perhaps damage to the fluid in which they're suspended.. many lubricants for instance increase in viscosity after enough heating/cooling cycles, which in this case would cause the cell to transition slower and slower unless you increase the drive voltage, up to a point. The thermal conductivity of the fluid may be low, so while there's plenty of thermal mass in the entire display, it's conceivable that there could be high localized heating within the tiny cells, which could cause damage if the rate of heating exceeds the rate of conduction of that heat away from the cells. Come to think of it, you could in theory use per-cell heating and the resulting change in fluid viscosity to control transition rate for some purpose. Thank you for the video!!
A couple of interesting thoughts after watching this video. Could the display be physically altered if it was hit against something or had a high acceleration? Would the particles be able to move in the fluid just from physical force? Also, there might be a way to speed up the refresh rate in the future by developing a different medium for the particles to float in. Its a very interesting technology.
Really useful resource! I actually want(ed) to create a little smart computer that runs as long as possible, but refreshing the whole screen is pointless when one would enter commands into a terminal one-by-one on a mobile small computer...
I mean the level of detail in this video astounds me. I’d love to know the background of the publisher of this channel. Is he a retired engineer? Or does he do these videos betwixt rocket science projects? I’m really in awe of this cat...
If you haven't already used this, OBS can effectively save you from pointing the camera at the screen and seeing those distortions (lines). I use it in conjunction with anything else I may put together in regards to video. Might be a bit more editing (or similarly, the same), it would take less time in regards to repositioning the camera and zoom.
+LegendLength it's a static charge. The kind of charge that makes sparks when you touch the doorknob and stuff like that. Let's say you remove electrons form an electrode, and you put electrons in another. electrons want to flow from one to the other, so they're "charged". But if there's no path between electrodes, the charge is static (because it doesn't move). Though there's no current, the charges affect the electric field, attracting opposite charged particles. Edit: Like when you rub a latex balloon to your head and stick it to the wall or ceiling. It's static charge what holds it in place.
Don't know specifics (like waveforms) but to prevent 'ghosting' kindle is using something like your quick lookup tables, so it can refresh faster, but redraws entire screen (flashing white and black) every 4-6 pages, don't know if the page amount is random or is it determining it somehow, like amount of pixel change from last full screen redraw.
Best source for details on these e-ink devices I've ever seen. One thing I would love to be able to hack these displays to be able to do is to be able to read the current content off the device. Imagine a reader that you attach the e-ink display to and it can determine what's displayed on it already. Based on what you found scraping off coverings and such from the displays - would it be possible to somehow determine what the display's contents were simply from the display itself? Think of it as a very slow but visual static RAM.
what if you increase the voltage? you could test and see what the mosfets inside the driver would be able to withstand and then test how fast you can make it while still having some reasonable contrast
From the working principle explained in the video, i do not understand how long term damage/burn in could occur if used at say 3hz without cycling properly. The screen getting charged, so what, just cycle a few times, no?
It’s fun to see you dig into stuff really deep like this. I’m tempted to buy one of those to play around now. It would be nice if you had some screen recording software on your Windows computer, to avoid the artifacts of recording a screen. I look forward to your next video. See ya next time, bye.
Sion If I had to guess it’d be because the particles in the cell might start to move anyway if no charge was passed through it, but if the cells around it were receiving a charge. So if you wanted one cell to stay white, but every cell around it was being turned black, you’d need to apply the white voltage to the cell at the same time so it wouldn’t change color. This would probably also be more important the smaller the cells were. This is just a guess though, I don’t really know.
@@rylblue3909 It's a good guess, very likely. The screen could then be improved with an edge detection algorithm to detect where a change is happening & only apply the "counter current" around areas that are changing. Maybe occasionally apply an update to areas that haven't been updated in a "long" time (like 1 to 10 seconds) just to make sure that the pigments stay where they should.
That's the point of the partial refresh, so it CAN be done. You might not WANT to, because of what he was discussing up front about the potential burn in problem.
Really cool video! You should really try Notepad++ instead of Wordpad as a Texteditor since wordpad isn't designed for this stuff. It's free and open source :D
I can't believe how beautiful that oscilloscope is!
i just looked it up... 45000 US$
i need circa : 3 kidneys and a half of a liver ,lol
@@fredlllll This is why you buy broken or second hand. Use primary asset to fix trash and recycle. It may not be perfect afterwards but you will be ahead further than you were before. BUT at 45k, they should have very long warranties and broken ones will be sought after if the warranty is crap or exploitative with dumb high costs. When you see such high prices for an inspection tool, then you know it is probably overpriced.
pretty sure its also matched with a beautiful price
Fascinating. Never realized how these E-paper displays work. Thanks for sharing this thorough video :)
Holy balls
Nice to know, that one of my favorite youtubers, is on other channels i watch too! :D
I read you comment, and heard your smooth voice in my head
Better put a huge magnet next to some epaper.
Hi KASA :)
I worked for EINK several years ago and the team I worked for mainly focused on creating the waveforms for each transition. Each grey had a waveform and there were also different waveforms for different temperatures. Because any variation in the chemistry or display assembly affected the transitions, each batch of panels required custom waveforms. I'm sure it's more automated today but the number of technician hours involved in making those earlier panels was crazy.
having to find the right waveforms per batch sounds tedious, but that must have been a nice paycheck for the technicians
Is that why eink screens cost so much? (ie like the big ones in the remarkable tablet or sony dpt)... I always assumed it was just cuz eink had a monopoly on the market
8hu
@@anjankatta1864 From what I know, not only; they also have a realyl bad yeild rate
@alysdexia nescient, lacking of knowledge, ignorant; doth not describe a paycheck, smarty.
I love you please don't go away for this long ever again
I'm actualy fine with the time it takes for a new video, because I know we are geting over the top quality content 😁
Thanks! I often feel like I'll wait for a project to go from "in progress" to "fully finished" to avoid making multiple videos about the same topic, but then I don't make many videos at all. Also, still trying to find a good balance between super tech-heavy and short-form fun science. I like both, of course.
You should do both and label them accordingly, not too many in the process video but 1 or 2 would be appropriate.
I feel the exact same way! I love how you do things Ben, you're a real inspiration. I hope that I'm able to have a shop like yours someday. Keep up the videos!
Applied Science, just do both, finding a balance is just a waste of effort when you have a pretty diverse audience. You can't please everyone all the time anyway
Sorry, what did you say? I was staring at the scope.
I can't tell you how refreshing it is to see a new video from you. It's like a drink of cold water for my mind. Hope things have been going well!
Thanks! I really appreciate it. I've got some more materials/chemistry videos on the way. It's good to be back.
That scope, WTF! bigger screen than the laptop in the background.
Don't hate the layer, hate the gain.
200$ computer and a 20000$ scope
Eirik Strand Priorities.
and a -1$ text editor, i think he should at least try downloading Notepad++ :D
K1ngjulien_ if it ain't broke don't fix it.
Or sublime
he said the scope can run windows 10, so...
Such a cool technology. I love my ebook reader. The one that I have has a refresh sequence where it does minor refreshes every page, and then a longer full-clear refresh every 5 or 6 pages or something. It's also worth noting that the cells containing the charged pigments do not correspond to individual display pixels, but are smaller and more randomly clumped together. The pixels come from the electronic array underneath, and one pixel may activate particles in multiple or partial cells. The cell structure exists to ensure a good distribution of the different particle types across the screen.
It would be interesting to try a burn-in test with that display without the reverse cycle, e.g. 10 updates at one spot, 20 updates at the next and so on until the text is barely legible. Then apply full black/white cycles to see how long the burn-in takes to clear at each spot (or how much it fades after let's say an hour).
Yeah, I second this. The refresh rate of these is what prevents me from using them in projects. If someone else who knows more than I do could test one of these to destruction I'd be super stoked. I'd totally use these all the time if I could achieve 3 Hz refresh and be confidant that it'll last more than 3 days.
That's a great idea and gizmoguyar nails my interest in it too. Fast refresh rate such that you can interact with it just like a lcd like on a iPad, it becomes so so so much more usefulio
I think that after a certain point the damage becomes irreparable because the voltage can no longer affect the particles quite as much
Since most e-Ink screens that are to be changed frequently are for consumer products (Kindle, etc) I believe the manufacturer has gone to extremes to avoid burn-in, which would prompt lots of purchase returns. Much easier and cost-effective to condition consumers to flickering every once in awhile than to random ghosting of text over text. Therefore cutting the reverse cycles shouldn't cause dramatic ghosting in a home-brew project, as long as you don't care about not having a pristine viewing condition each time.
i accidentally looked away during the dot moving period, then came back when he flip the screen color and said there should be a ghost circle burn-in: and i can't really see it?
i do see some unevenness in color scattered in a small region, but i have to deliberately imagine it is a circle to acknowledge ahh it indeed is a circle
then i rewatch the dot circling period of the vid and holy shit the burn-in become so freaking obvious!😂with shades differences!😂the later the more firmer!😂maaan!😂(it was burn-in in my retina?!?)
I have been following your channel for a long time and I always thought everything you were doing was way god-level above and beyond the work I do. But low and behold I just watched this video and this was exactly what I was working on! I was working on a 4 level e-ink display but it's very similar. The lookup table format in my controller works differently but close enough to at least be set on the right path. I learned a lot from this video and I am trying to reproduce your findings. I made a lot of breakthroughs inspired by your video. It was basically unusably slow before and now it's much smoother and getting better. Thank you very much!
Ben don't make us miss the best channel on youtube again!
Also as someone who does C++ programming for living, I do recommend you getting absolutely any kind of proper code editor with syntax highlighting and/or auto formatting. Once you get used to it, it just takes way less time to read code that's properly formatted and displayed. Popular C++ editor choices for windows that are intuitive and work out of the box are: notepad++, sublime, atom, vscode(not visual studio) and many others.
I liked how this video about E-paper turned into a vid about that oscilloscope
i was just waiting for him ending with: "Thanks for tektronic for sponsering this video, please use this link for 10% off"
Wow that scope is incredible
it's like driving a lamborghini gallardo on the street
but is is 42000$ incredible?
Very cool, Ben - always learn something from your vids. Glad you're enjoying the scope!
w2aew Ben, us at Tek love seeing you use our new baby 😂😁
This scope is unreal! You can see from the comments how many engineers are completely blown away by how much information the scope can capture and display at one. The team at Tek should be really proud for developing such a nice instrument!
Thanks for the video! I've got a few of these ePaper panels directly from Good Display in China (the manufacturer as far as I can tell).. they were happy to help with questions until I asked for a datasheet describing the LUT bytes, then they became very cagey. Nice work on finding a datasheet that has these details. All of their datasheets seem to be missing information and are bit inconsistent.
I applied to work for E-Ink Corp once. They never got back to me, but I did some reading on their tech while I was waiting. Very cool tech, and this video does a great job of showing it off
I'm getting mixed signals from your oscilloscope ...
Boris Angelis I see what you did there
he's Alive!!!!!!!
"oscilloscope, open the bay door", -"I'm afraid I can't do that Dave"
That is one impressive scope - your in depth research even moreso.
usually when people on youtube make "[something] hacking" videos, it's some clickbait douche presenting cursory google search instructions to change user-editable settings. this guy actually does real, intrepid work and does an extremely thorough job explaining it.
I'm doing my PhD on e-ink and it's a bit of a misconception that e-ink needs no power at all. The major problem with e-ink is the fact that there are charged inverse micelles in the cell, which over time will start to replace the colloidal particles. There's a lot of interesting scientific literature in this field actually!
Given your background then, what do you think is the most interesting / applicable technology for the 'ipad has baby with a kindle' application of easy on the eyes digital note taking
@@anjankatta1864 sony dpt-rp1?
Fascinating.. Don't know which I'm more impressed with, the epaper & your ability to reverse-engineer what's going on, or that borrowed scope! With equipment like that, in the right hands, you could figure out how 90% of stuff ticks. Great video.
Great video and explanation, thank you!!
Samy Kamkar Your last video was 11 months ago! What are you working on?
You should check out what the reMarkable paper tablet is doing. It's using an actual e-ink display, but updates fast enough for latency-free handwriting (directly on top of the display - it uses a stylus digitzer made by Wacom I believe). It's actually so fast that you can't detect any latency at all when writing or sketching. It's also very high resolution - 1872x1404. No flickering or updating on the display while drawing, or erasing. It's pretty amazing to use, I own one myself. It does occasionally perform a full screen refresh, but it's so fast that it's unintrusive - and it only occurs while you are in the menus or something like that, never really when actually drawing on it. You can also erase and such, but I guess once they got the drawing figured out, erasing was kind of a freebie.
And it's true e-paper as well - it maintains the image even with the battery removed or what not.
I have no clue how they made it happen. It must have incredibly fast response time - it's as smooth to use as a Wacom tablet on a PC, looking at a 60hz screen.
Super late, but looking at it's stats (10.3", 226dpi) it seems like it's using the ES103TC1 display. I'd get one myself if it didn't cost quite so much.
I remember following a company for a while which was trying to create an eink tablet, it felt like they fell off the face of the earth but this might actually be them.
Can you use it like a monitor
alysdexia Look out people, RUclips comment section has a full time editor now 🤣
@alysdexia why use many word when few word do trick.
@alysdexia git outta here
I can't believe how detailed and clear your explanations and demonstrations were. This was exactly what I needed to understand the limitations I was running into with e-paper. Thank you!
TIL there are multiple e-paper types. The one described here is electrophoretic. There also exists gyricon, where instead of a bunch of pigment particles, each cell contains a small bead that is magnetically polarized and colored, so that it spins in place to present the desired hemisphere. It occurs to me that one could have a sphere with colored poles and a stripe along its equator. You could then have electrodes along the display surface to orient the sphere to show one of its poles, and in addition have a electrodes in the middle of the paper's cross section, which would cause the sphere to orient so that its equator is visible at the paper's face. This would give a 3 color per pixel display. I hope that makes sense.
Do you think that would cause refresh rate to get worse or better or the same? That to me is really crucial to closing the gap with lcd note taking and reading like the iPad
I can't believe that is how the tri colour works. Amazing. I always learn something(often too many things) in your videos.
"The industry thrives on secrecy or something." - 14:25
I'm so glad you are back with quality technical content. Don't go away like this again
!
I was just starting to investigate eink displays for something I want to bring to market using a color one, Thanks for the great video!
Ben, I admit that I have very little use for e-paper hacks, but still the same, I really appreciate your content. Thank you.
This scope makes our MSO 4000 at work look ancient
No doubt about it, you are an impressive dude. I'm always impressed with your vids' content.
Incredible how far technology has come. You explain things really well.
Brilliant video! I have almost no experience with e-ink displays, so this really helped me learn how they work and how the controller works. I am planning to make a video about OLED displays, how they work, and include some microscope footage of the SSD1306 chip found in most OLED display modules. SO far I have got some microscope footage of the chip, but not enough to use in my video. The problem I am facing is that even with a bright light shining from the top, the chip is still too dark to see clearly.
The E-Ink display that has red in it is probably less complex than that. There is one normal pixel and a pixel filled with red cells right beside each other. Color E-Ink displays work like this but with three pixels: One cell is filled with red particles, Another is filled with green particles, etc...
Nope
Strange how RUclips protection wasn't triggered by that scope porn!
On a more serious note, why all the fuss with cycling the display with AC voltage? Why not simply skip writing same "color" to a pixel, or in other words, only write pixels that do change?
FlumenSanctiViti my guess would be that the electric field from neighbouring pixels starts affecting the pixels that you skipped
My guess is that it uses a chip that handles setting all the pixels from the lookup table, but the setting of that table is handled in code
Whenever it takes months for a new video, I know it’s going to be a good one. Nice work Ben, this is super interesting!
absolutely iconic and masterful youtuber. such great content.
Great to see you back again!
That scope in utterly ridiculous wow
I bet that logic probe is worth more than my entire lab xD
Stumbling across a science channel I was previously unaware of is always a good start to my day. :D
I stripped down a monitor some time ago and saw a few of those silicon looking bars. I had no idea what it was but postulated it was some silicon device. Thanks for pointing out what it is.
Some time ago you noted on a Q&A segment some of the reasons you made videos. A few months ago I watched that video and decided to start up loading content. Reasoning made sense to me.
Some really cool stuff! Love the custom code. That stuff usually creates quite a bad ghosting effect but the updates seem entirely clean.
I've been watching some of the progress on Twitter and I'm glad I got to finally see the video! Can't wait to see what cool projects arise thanks to this intense research and work!
For the love of all that is good, please burn wordpad. There are many beautiful text/code editors freely available; sublime text comes to mind. The black default background might help with some of the waffle texturing that is showing up in the video for whatever reason. Beside that, thank you kindly for the video(s), it is an honour to be able to see a true genius at work, and the world is definitely a better place for you sharing your time with us.
Seconded. I personally love Notepad++
I still use wordpad, just because notepad doesn't handle CR/LF that's non-windows properly (wordpad does).. Wordpad is compact, fast, and gets the job done with no fuss. I don't use it for coding though, I use MinGWStudio for that (totally unheard of 3rd party GCC/MinGW IDE that's almost identical to DevC++, except way more stable).
Vim for the win. Screw wordpad, I would rather use ED on a teletype.
Agreed. I’ve been giving VS Code a go on windows and have been very happy with it for C++, Go, Ruby and just about everything else so far.
VSCode, Sublime Text, or Notepad, in that order of preference :)
Im fairly excited for "full" color e-ink displays. E ink has been teasing them for quite a few years now, and sony has shown off a few prototypes as well, though they seem to have a much wider color range compared to the E-ink ones. They are what i think they call EDP displays, apparently it uses low temperature a-si TFT's.
Finally! I wonder how this compares to what my Pebble uses though, I don't spot any flickering, but also don't have burnin, while it has been on my wrist nearly 3 years now.
I believe it uses a "memory LCD" made by Sharp. It consumes a very small amount of power (microamps), and has an appearance between normal LCDs and E-paper displays.
Ah that seems accurate as it indeed doesn't look like actual epaper (but I thought it maybe just was the backlight). It seems a better technology than the epaper stuff then, except for that it doesn't retain the image. I'll have to look at that tech up though. Thanks!
Another video idea for you :P We need more of your videos anyways.
Ngrok + Kindle paperwhite + custom app = awesomeness. Considering posting a video shortly.
That scope is glorious indeed!
Thanks for the video. The eInk/ePaper displays are really nice. They will certainly be more popular with hobbyists when the prices come down a bit on the larger panels. I didn't know there were some with the extra red colour. That is the most impressive scope I've seen so far. 8 channels with up to 8 digital inputs each? Very nice and also very expensive.
Do you know why the larger panels are so damn expensive? It baffles me tremendously (some mentioned it's cuz of eink monopoly.. Others have said more individual tuning needed ;?)
Cool stuff as usual.
I was thinking we would see some ruined displays, as a demo how not to drive those displays.
I haven't had any permanent burn-in on any of my displays, but I've also been trying to avoid that. Maybe I should buy a few more and start with destructive testing!
Honestly I would like to see some destructive testing, be it displays or any future projects, it just scratches that "what if we take it to the limit?" spot.
But yes, it's not the usual stuff this channel covers, I guess we have AvE and others for that.
Today sir you blew my mind. I had no idea why those displays looked different. I guess i just thought they were ultra cheap. Very cool
Only pure quality coming from this channel!!!!
EInk has _huge_ Electronic Ink Display Modules (with 31" 4096 color or 41" 16 color greyscale) for less than $2500; also tiny B&W modules for under $100.
Here's a quote from their Website:
"EC312TT2 is a reflective electrophoretic E Ink technology display module based on TFT active matrix with color filter design. It has 31.2” active area with 2560 x 1440 pixels, the display is capable to display 4096 colors depending on the display controller and the associated waveform file it used.".
That oscilloscope is incredible! And here I was content with my GDS-1054B lol.
Wow. I've always wondered about these and was so curious about their limitations. Wonderful explanation.
144hz incoming for gaming? :>
Young Padawan playing Doom on a kindle when
144 is a lot more difficult to do than 3; for once you auctally need all things to sync up in the display 144 times a second; that's also why they were so expensive
i actually think the 3hz would be cool for an epaper display game and an interesting limitation to design around
I haven't played with e-paper displays yet but I'm a big fan of WaveShare's toys. I use their FPGA kits for all kinds of projects. You've mentioned permanent burn-in and/or damage - do you have any idea of what the possible mechanisms for that would be? I can't picture a way in which you could damage these cells even with strong polarization, based on how you've described their theory of operation. The only thing that comes to mind is some kind of cell overheating that could perhaps fuse some of those powder particles together which are presumably at least partially plastic (I'm thinking of laser printer toner), or perhaps damage to the fluid in which they're suspended.. many lubricants for instance increase in viscosity after enough heating/cooling cycles, which in this case would cause the cell to transition slower and slower unless you increase the drive voltage, up to a point. The thermal conductivity of the fluid may be low, so while there's plenty of thermal mass in the entire display, it's conceivable that there could be high localized heating within the tiny cells, which could cause damage if the rate of heating exceeds the rate of conduction of that heat away from the cells. Come to think of it, you could in theory use per-cell heating and the resulting change in fluid viscosity to control transition rate for some purpose. Thank you for the video!!
You've got a tricked out, top of the line tektronix scope, and you're using.... wordpad?
Sublime is great, as is Notepad++
Why?
Sublime even has an Arduino IDE plugin. Its pretty sweet. So much easier to see whats going on.
Probably old school
he's probably read enough about it that he doesn't need to use the features in a more advanced editor
Sublime is trash software
Universal scale of choice.
It has been a while since I have seen grooveshark in bookmarks!
Great video! Very well explained, keep up your videos :D
Good video after a long time. Thanks for your work.
A couple of interesting thoughts after watching this video. Could the display be physically altered if it was hit against something or had a high acceleration? Would the particles be able to move in the fluid just from physical force?
Also, there might be a way to speed up the refresh rate in the future by developing a different medium for the particles to float in. Its a very interesting technology.
A true science channel on youtube, so rare these days.
Technology connections brought be once again here.
I came for the awesome science, I stayed for the "anyways, hope that helps, see you next time... b.y.e....." love it!
Very interesting tutorial, now I know how they work, thanks.
Brothers, our prayers has been heard, he's back! And with 1/2 hour video!
Id like to see some slo-mo of the pixels changing
Really useful resource! I actually want(ed) to create a little smart computer that runs as long as possible, but refreshing the whole screen is pointless when one would enter commands into a terminal one-by-one on a mobile small computer...
Would love to see another epaper video especially with the new color screens and stuff that are coming out!
I mean the level of detail in this video astounds me. I’d love to know the background of the publisher of this channel. Is he a retired engineer? Or does he do these videos betwixt rocket science projects? I’m really in awe of this cat...
that is it. I have decided this man just knows everything.
You can do partial update on Eink displays, you just need a different library that supports it.
If you haven't already used this, OBS can effectively save you from pointing the camera at the screen and seeing those distortions (lines). I use it in conjunction with anything else I may put together in regards to video. Might be a bit more editing (or similarly, the same), it would take less time in regards to repositioning the camera and zoom.
If the particles are charged, isn't it a problem that they could stick together or repel each other inside the chamber?
gabest4 the individual negative charges of the particles are nowhere near as strong as the negative charge of the electromagnet used to move them.
Maybe some sort of surfactant helps?
Joe Smith they're not moved by an electromagnet though
LegendLength the particles are moved by an electric field not by a magnetic field like there is in an electro- or permanent magnet
+LegendLength it's a static charge. The kind of charge that makes sparks when you touch the doorknob and stuff like that.
Let's say you remove electrons form an electrode, and you put electrons in another. electrons want to flow from one to the other, so they're "charged". But if there's no path between electrodes, the charge is static (because it doesn't move).
Though there's no current, the charges affect the electric field, attracting opposite charged particles.
Edit: Like when you rub a latex balloon to your head and stick it to the wall or ceiling. It's static charge what holds it in place.
I've never seen such a magnificent awesome oscilloscope...
Don't know specifics (like waveforms) but to prevent 'ghosting' kindle is using something like your quick lookup tables, so it can refresh faster, but redraws entire screen (flashing white and black) every 4-6 pages, don't know if the page amount is random or is it determining it somehow, like amount of pixel change from last full screen redraw.
When you were talking about burn in when the display was on the RUclips logo, the large circle was mildly visible behind the white in said logo.
that scope is bloody impressive
Best source for details on these e-ink devices I've ever seen. One thing I would love to be able to hack these displays to be able to do is to be able to read the current content off the device. Imagine a reader that you attach the e-ink display to and it can determine what's displayed on it already. Based on what you found scraping off coverings and such from the displays - would it be possible to somehow determine what the display's contents were simply from the display itself? Think of it as a very slow but visual static RAM.
You should do a video on how to connect a photodiode to an oscilloscope to detect high frequency PWM on laptop backlights.
Very informative, nicely explained. Thanks
Great research work with very good findings. Thanks for the insights into E-paper technology.
That scope is crispy!
the Master just release another amazing video !
thanks, it was very instructive
From the future we actually have this in CMY for industrial signage and ereaders with RGB. Crazy
what if you increase the voltage? you could test and see what the mosfets inside the driver would be able to withstand and then test how fast you can make it while still having some reasonable contrast
Won't it just polarize faster
Well Kindle and nook both have hacks that only update part of display, but every couple of frames all display is cleared.
From the working principle explained in the video, i do not understand how long term damage/burn in could occur if used at say 3hz without cycling properly. The screen getting charged, so what, just cycle a few times, no?
It’s fun to see you dig into stuff really deep like this. I’m tempted to buy one of those to play around now. It would be nice if you had some screen recording software on your Windows computer, to avoid the artifacts of recording a screen. I look forward to your next video. See ya next time, bye.
Never clicked faster! Great videos!
When switching frames, why is it trying to make a white pixel white again? Why just do nothing?
Sion If I had to guess it’d be because the particles in the cell might start to move anyway if no charge was passed through it, but if the cells around it were receiving a charge. So if you wanted one cell to stay white, but every cell around it was being turned black, you’d need to apply the white voltage to the cell at the same time so it wouldn’t change color. This would probably also be more important the smaller the cells were. This is just a guess though, I don’t really know.
@@rylblue3909 It's a good guess, very likely.
The screen could then be improved with an edge detection algorithm to detect where a change is happening & only apply the "counter current" around areas that are changing.
Maybe occasionally apply an update to areas that haven't been updated in a "long" time (like 1 to 10 seconds) just to make sure that the pigments stay where they should.
@@sebbes333 Hm, that would be neat, but would also mean more expensiver hardware (good devs cost a lot of money, yknow).
That's the point of the partial refresh, so it CAN be done. You might not WANT to, because of what he was discussing up front about the potential burn in problem.
Perfectly explained ! Thank you.
Glad to see you back!
Really cool video!
You should really try Notepad++ instead of Wordpad as a Texteditor since wordpad isn't designed for this stuff. It's free and open source :D
I was shocked when setting wordpad, lol.
Great video! Looks like a Tektronix ad! Haha. Very informative, thank you.