ESP8266 ePaper WiFi Display Kit Runs for Months on a Charge | Hacker News Reader

ESP8266 ePaper WiFi Display Kit Runs for Months on a Charge

blog.tindie.com

218 pointsokket77 comments

I'm currently awaiting the delivery of a 400x300 epaper module from aliexpress which I'm hoping to use as a display that updates daily with google calendar entries, the local weather, and similar things.

I don't have a solid use for it, but I'm looking forward to experimenting as I'm surprised how inexpensive these things have become - the item I ordered is ~€25.

I'll be using a Wemos mini D1 to drive it, and I'd also expect a long life-span between charges. (Though waking up and using the WiFi is always the heavy-weight part of these devices.)

Do you have a link to the item? Or atleast a name + type number.

Why did you choose that specific display? A quick Google search shows that driving an e-inkt display can be challenging. Have you checked how hard that particular item is to drive?

Thanks! I can think of so many cool things I would love to build for ~€30,- :)

Found a few here: https://www.aliexpress.com/wholesale?catId=0&initiative_id=S...

Interfacing them isn’t too bad, your basic atmega328 should be able to do it handily if you’re comfortable with spi/i2c and co. The problem is that they all come as FPC (micro ribbon cables, those orangish-brown display strips) so you need to at the very least get a compatible FPC to DIP breakout board if you’re not comfortable with SMT work.

I'm not sure I want to get involved in that (at my office) since I already deal with enough equipment already, but I think something cool would be to make a small ePaper like display for conference room in front of each door, and display the upcoming meetings that are booked.

Of course people could look directly on the calendar through their phone or computer, but sometimes a quick glance when you're near is all it takes.

In facebook and a whole bunch of other tech companies they use android tablets / iPads for exactly that.

There is even a company that has an app that manages that for you, but I forgot what it was called.

Robin

Yeah. Teem is another one.

I should have thought of including a link:

https://www.aliexpress.com/item/400x300-4-2inch-E-Ink-displa...

I've seen libraries for driving this specific one, so I assumed I can hack it to work. The 800x600 size is perhaps more "useful" but slightly more expensive than something I'd buy without a real need.

I'd love to put together a small E-paper display that I can stick on the door showing where I am or if I'm free/busy.

3" display is enough for my need.

Small LiPo battery works for my need.

I can spend a few hours learning to program this thing with Arduino IDE or whichever tool is needed with a little bit of Internet handholding.

But, >$50(incl shipping) kits don't give me the value I look for. Why are they so expensive?

I'd happily buy a pre-made setup of (controller+LiPo battery+e-paper display) easy to program device if the price is < $30.

I agree with you. I think kits are too expensive and going the homebrew route is generally more fun and educational. But to answer: ePaper is expensive, $15-20 for that module. The rest of the components are cheap. You're paying for some profit for the designer, amortised development time, assembly (by hand?) and the fact that it should work straight away (i.e. so you don't spend time). Tindie also take a cut, 10-15%? Shipping isn't free for the seller either.

At work I grit my teeth and suggest we buy the dev kits because it's worth paying $100 for an official board to check something works tomorrow, vs 3 weeks for a PCB and 2 days of tinkering. But at home, as a hobbyist, I mutter under my breath and fire up Eagle.

Tindie takes 5% and passes on the payment processing fee (~4%). :)

You could hunt for an older ebook reader on eBay with similar hardware if you're willing to spend some time getting it to do what you want. The older Kobos were pretty simple to modify for example.

The color version is quite cheaper [1]; 30 USD.

[1] https://www.tindie.com/products/squix78/esp8266-wifi-color-d...

Don't think that is epaper which means it doesn't have the persistence of image even when power is cut.

Of course it isn't e-paper. I was just pointing out the difference in price. Cause its from same supplier [1]

[1] https://www.tindie.com/stores/squix78/

Would a 1.5" display work? You can buy these [1] from AliExpress for $15 a piece. And for $19 you can get this board from Sparkfun with a built in LiPo charger [2]. Or course you can go way cheaper on the board + charger, but this one is pretty easy to use if you don't have much experience with electronics.

[1]: https://www.aliexpress.com/item/200x200-1-54inch-E-Ink-displ... [2]: https://www.sparkfun.com/products/13907

Well, 15 + 19 is already $34 and that doesn't include a case or battery. Of course there are much cheaper ESP8266 boards you could use rather than the Sparkfun ESP32 Thing but then you have to go through the trouble of programming it, building an enclosure etc all of which is probably going to cost you a lot more than $50.

What you call “trouble” others would term “fun” or “challenge” or “the whole point of doing it yourself”.

I agree, doing it yourself is fun, but if cost is your main concern, then DIY is absolutely not the way to go. For reference, I recently finished relatively simple ESP32 sensor project and while the components by themselves probably cost about £20, I ended up spending about £150 in parts and tools over the whole process. The extra cost was down to breaking parts and ordering replacements, needing basic components such as resistors and breadboards which many will already have in their toolbox and needing to make about 10 different orders + associated shipping costs as I went through each revision of the design.

I guess it's only matter of time before somebody will start offering supermarket e-pricetags to hobbyist:

https://theworklife.com/graham-miln/2013/12/17/e-ink-price-t...

Yeah. Those would be super useful as general purpose displays. I found something on Alibaba without much technical info about.

Hopefully someone gets the idea and mass produces it.

https://m.alibaba.com/product/60250435288/Supermarket-esl-di...

yeah, I wish there was an easy DIY way to program those

order enough of them and i expect you could get one of the sellers to give you specs. they just use 433MHz; shouldn't be too hard to talk to them once you know what to send.

I was able to find them here for cheap: http://www.pervasivedisplays.com/products I have no relation to them, there might be better distributors, and I have not ordered these. They are available via Digikey

Oh, wow! I did not know those were a thing. Given the volume that supermarkets will be buying them, those will get affordable quick.

SHA-2017 had a really cool badge: https://wiki.sha2017.org/w/Projects:Badge. 2.9" e-paper display, ESP32, 1000mAh LiPo battery and firmware with an online "app store" supporting apps written in micropython.

Sadly, there were only a handful available after the event (for about €30). I'm hoping someone continues with the design, because I can think of quite a few applications, but I'm worried about breaking it.

Agreed. And the other thing that gets me is battery life. If I have to charge something every few weeks, I'd rather just run power to it and be done with it. In which case, I might as well just use a cheap phone.

E-ink displays consume very little to no power if you don't refresh them too often. It\s up to the controller to give you better battery life.

If you don't want to replenish batteries "every few weeks", how does a cheap phone fit your need?

Because leaving it plugged in means continuous power, therefore you never have to remember to charge it.

You can leave anything plugged in if that’s your problem. A dc wall transformer can take the place of a battery in just about any device.

Right. But if I have to plug something in, then I might as well just use a cheap phone. Color display, full browser, touchscreen, sensors, etc. For twice as much money as this kit, I get a lot more value.

That is totally up to you. You can run these modules from a wall wart if you like. The beauty about a battery driven device is that installation is so much easier. You configure it, use some magnets to put it on your fridge and it just runs for weeks or months. The module serves a very specific use case: it's like a smart sticky note. Easy to place, easy access to the information.

I bought a few of these three color e-ink displays from China: https://m.aliexpress.com/item/32499315786.html

They’re dirt cheap.

They look cool! Thanks for sharing. Asking as a noob to programming electronics, do you have any guides on how I can display a few characters on one of these?

As in which controller to buy, how to interface the controller to the display, and how to set myself up to program such a controller... etc. I can try if there are enough resources, but honestly I'm easily intimidated if someone says "Just get a ESPXYX900 controller and hook it up to ABC IDE and flash this ROM and there you go."

I can program in Python and I have heard of MicroPython. If only I can figure out how to get to a point where I can get my python program to execute on one of these tiny wonders.

I’ll give you a more realistic answer: if you’ve never done microelectronics interfacing, you won’t have much luck until someone develops a driver (“software api”) for a particular eInk display for a platform you’d like to use.

I’ve never used MicroPython, but the C APIs I’ve seen for LCDs for micro controllers are usually pretty straightforward. Depending on the level of abstraction, they might offer a basic “put this B&W/R&G&B pixel at location x,y” or a higher-level “print text xyz” type of interface, both of which are easily exposed to python’s ffi and consuming it from there should be no problem.

But the trick is waiting for someone to get that driver made.

I’ve been meaning to play around with mine, if there’s sufficient interest I could see how realistic writing up some Arduino-compatible code to interface with it and some instructions would be.

Thanks for the straight response. I was afraid this might be the case.

I browsed around Waveshare.com and they seem to have posted "demo code" for the displays you linked before. I will try to muck around with them later, but it'd most likely be beyond my understanding of display drivers.

It'd be incredibly cool to have a cookbook for "Arduino + Tri-color E-Ink display for displaying text" - for beginners. If you think you could muster time & motivation to develop such instructions, I can help with (1) personal donation of $20, and (2) Sponsoring a domain and a basic web page describing the project for others to follow with donation options directed to you.

If you check the micropython forum, you might get help developing a driver. That said, there's this:

https://github.com/peterhinch/micropython-epaper

Which supports this 2.7" display: http://www.embeddedartists.com/products/displays/lcd_27_epap...

Which is a bit cheaper than the OP.

A friend put together this concept of a simple e-ink display with an adhesive back and long battery life:

https://sean-farrell-1uu4.squarespace.com/

They got it to working prototype stage, I think, but didn't end up taking it commercial. It's too bad – a cool idea. I'd love to have one.

That looks very useful and easy to sell. I wonder why your friend didn't proceed if he/she already had a working prototype.

Heck, I'd buy it for my desk even without any of the GUI tools if it allowed me to display from a list of quotes.

Honestly, $50 is not expensive for a DIY device. If you were to purchase the main components separately, you could expect to pay about $30:

2.9inch E-Ink display module: $18.99 [1]

Wemos D1 Mini 8266 module: $2.55 [2]

600mAh Lipo battery: $1.87 [3]

USB-to-serial converter: $1.08 [4]

3D printed case: $5 ?

On top of that, this is a totally custom designed and assembled PCB which would cost any indie hardware hacker a lot of time and significantly more than $50 to produce.

Of course, if you were to get this mass manufacturered, you could probably bring the cost down to about $25 but I love Tindie because it gives makers a platform to share their products and consumers a place to buy very custom products that they do not have the time or money to make themselves.

[1] https://www.waveshare.com/2.9inch-e-paper-module.htm

[2] https://www.aliexpress.com/item/1pcs-Smart-Electronics-D1-mi...

[3] https://www.aliexpress.com/item/High-Quality-3-7V-600mAh-25C...

[4] https://www.aliexpress.com/item/1PCS-CP2102-USB-2-0-to-TTL-U...

Thank you spuz, I totally agree. Full disclosure, I'm the guy who develops and sells these modules. I do this in my spare time. I didn't have any experience in manufacturing and shipping products like these before. A very good friend who lives in China helped me a lot with designing and manufacturing the ESPaper modules. Normally in China you'd have a minimal order quantity of thousands not hundreds. But with his help we can do small batches. Small batches are a driver for the price per piece of course. But there are a lot of other costs involved, pure hardware costs are just little over 50%. You get the ESPaper in a nice package with a pretty enclosure. As somebody mentioned earlier payment gateways want to have a share. And to pay my friend in China for his work and hardware I have to transfer the money using Paypal or banks and they open their hands again. There might even be currency conversions involved, depending on the path the money takes. In rare cases a package gets lost on the way to the buyer or it arrives broken and you have to replace it.

For shipping I use a fulfillment service in China. If you order on Tindie I import the order in my own shop system. From there the order is exported together with orders that came in directly on my shop to the fulfillment center by API. I have to choose the best of sometimes 20 different shipping services and the costs for shipping are different for every case, depending where you live and if value added tax is pre-charged (e.g. EU countries). Then a guy in a warehouse in Dongguan walks to a shelf and picks one of my boxes, labels it and sends it off to you. I had to learn the hard way that sending without tracking is cheaper but quite risky. So I also put insurance on the orders. The guy in the fulfillment center also needs to get paid.

And I spent a lot of time developing a graphics library (MiniGrafx) which gives you a lot more freedom than other libraries available for these displays. And now I'm working on a web server application which will make it super easy to configure the ESPaper remotely. Choose a application (e.g. weather, Calendar, Meeting Room, Sticky Note), configure it and when the device wakes up next time it will download the image information from the server. So all that costs a lot of time and most of the code is open source or will be at one point.

I totally understand if you want to build this by yourself, since I am having a lot of fun doing it. But not everyone has that level of patience or expertise and those who don't I'm offering a short cut.

All displays should be e-ink. It's crazy to stare at a backlit screen all day.

Apart from the very slow response times, of course.

Things seem to be getting better in that regard [1]. We'll need better software/interfaces for using these kinds of displays though.

[1] https://www.indiegogo.com/projects/e-ink-monitor-with-hdmi-p...

Depends on the use case.

We have two e-readers from Kobo. Both have e-ink. They can both also run a browser. Pretty cool thus far? The refresh rate is abysmal which you're gonna notice whilst browsing, and it all has to be rendered local (including all the bloated JS being used nowadays). Both of these factors are a killer for battery life. So if you'd be scrolling down a web page, and then replying to say here in HN that'd be terrible use case because input is also terrible (touchscreen isn't very accurate either and low refresh rate makes the letters appear slowly). For pure reading however, it is great, and for browsing it is OKish (the only disadvantage is the rendering). It has Pocket support which is sanitised.

The less the screen is refreshed, the better e-ink is.

The refresh rate sucks. Transflective LCD gets you e-ink-like contrast, LCD-like refresh rate, and battery life that's still much worse than e-ink but a significant improvement over backlit LCD.

> It's crazy to stare at a backlit screen all day.

Why?

Backlit displays hurt our eyes because most apps often use full white backgrounds without taking into account the much bigger surface (that is, light) our eyes is exposed to by doing so, which is roughly double compared to say 10 years ago. That can be corrected by choosing darker themes, but good luck convincing most people to do that.

E-Ink technology needs a lot more time to become a viable option for monitors.

In the office most programmers and designers use dark themes, but in fact they are more straining than light themes because your eyes do more work adjusting for seeing little bright things in the dark than for seeing little dark things in the light - I guess in the end it is a matter of taste.

Or just turn the brightness down. The monitor I'm using now has a dedicated button for cycling through brightness settings. I have it on minimum for general use, and I set it to maximum for games/movies.

Great idea although too pricey for what you get IMHO. I'll stick to hacking Kindles for now.

This sounds interesting. Could you shed some light in what hacking a kindle entails? Could I hook up my micro controller to the kindle and use the kindles display? Could I add gpio ports to the kindle and use it as a micro controller?

Sure. The kindle runs linux and has a large e-ink display and battery integrated for a good price. You can do many things that you cannot do on a micro controller. For low level stuff you can hook up an esp8266 through wifi.

Check out using the Kindle as a raspberry pi display:

http://ponnuki.net/2012/09/kindleberry-pi/

https://maxogden.com/kindleberry-wireless.html

Thanks for the links, this looks very cool!

Here’s a link to a video I made: https://www.youtube.com/watch?v=Oel08SDFyIY

Of course, you can skip the server section and make the IP address point to your local server. Some caveats are you need to use pngcrush to turn the image into a format that the kindle can use.

Link to GitHub: https://github.com/yoonsikp/weather-display

Please share more. I have two Kindle 3 laying around and I'd love to do something useful with them.

Everything I read about hacking older Kindles involved rooting them and displaying a HTML with whatever you wanted. That sounds too easy, and I'd like to go deeper.

I have been waiting for several years for color eInk displays to come about. I am fine with a refresh rate of even something like multiple seconds. Thus far the only thing I have found are companies in China that are willing to fix a single image on a color eInk display (no idea what the point is if the image can't be changed, but that is my understanding) before shipping. I haven't went hunting within the past year or so though.. anyone know if there's been any progress on that front?

You can get 128X64 OLED on AliExpress for 3$. Mine draws 2.2mA to show 6 letters, 3mA for 12 letters.

It can be controlled with e.g. ATTINY85 (1$) which can e.g. set the text/image to display and go into deep sleep in beetween sreen updates where it draws almost nothing (< 0.01mA).

2800mAh battery featured in this video will last this setup for more than a month of continuous operation. If you turn the display off e.g. when the lights are off it can last several months.

I've got an old Sony eReader - before I throw it in the bin, is it possible to re-use the display? Do all eInk displays use the same protocol?

eInk is a technology on the physical level, has nothing to do with the digital/microelectronic component. So, in a word, no.

AFAIK There's really only one manufacturer of eink displays, so there's a good chance the physical interface is the same as other eink displays.

I was hyperventilating as I read this as ESP8266 runs for months on a charge but still ePaper is a great move on a path to finding self sustainable internet connected (w/ extending wifi possibly falling back on 3g) devices in the wild.

Has anybody done experiments to see if you can power esp8266 with solar cells with portable battery? We are in for some truly exciting applications.

Here is a weather station powered by a ESP8266 using solar power: http://www.instructables.com/id/Solar-Powered-WiFi-Weather-S...

How is this better than something using a Sharp Memory LCD, or similar low power reflective LCD with built in memory?

The power consumption of the LCD is slightly higher, but total power use is dominated by the wifi. It will still run for months and the refresh rate is much better.

Would an epaper hardware wallet for bitcoin make sense ? Or does the display technology matter ?

It wouldn't be the most important aspect of such a hardware wallet (maybe TEMPEST is a consideration?). These wallets (Trezor and Ledger Nano S) generally get their power via microUSB connection, and they're not on a lot like a wall mounted weather app would be basically 24/7 on. The Ledger Nano S uses OLED according to themselves [1]. If there's a way they draw less power, that is welcome if the source is say a laptop. There's also the Ledger Blue which uses a (larger) color touchscreen and rechargable battery. For the price of almost 4 Ledger Nano S.

[1] https://www.ledgerwallet.com/products

Do Oled screens degrade faster than epaper?

No, the display technology doesn't matter. It would make sense if you want to send transactions through QR codes, might be an interesting project. Not sure how it would be more useful than a Trezor, though.

If you're scanning the bar code with a phone app, it's just going to use the camera. So as long as the display has a high enough resolution to display the QR code and it shows up on a camera, it should work fine.

But keep in mind, you probably want the device backing the epaper display to allow some kind of password input and store your BTC key on an encrypted medium (either encrypted flash or an LUKS/ext4 formatted sdcard).

Ledger Nano S uses Intel SGX.

Trezor uses a Cortex M3, probably with a secure enclave.

The problem of offline QR code payments have been solved in a much low-tech way... https://cdn3.i-scmp.com/sites/default/files/styles/980x551/p...

I see main utility when you need per-item/per-transaction billing.

Instead of having to print static-per item tags, or redirect users to a website with online checkout, you can have that thing to generate individual codes for each piece of merchandise.

If one could add a touchscreen, this would be perfect for home automation.

If you look on the bottom right it says "refresh". Not sure why it says that, but its suggestive that it can be refreshed physically somehow.

Presumably those three circular things below the display are buttons? So not a touchscreen, but you could navigate menus with up/down + enter.

The left button is used to put the ESP8266 into flash mode for updating the firmware, the right button is used to reset the device, effectively pulling it out of deep sleep to refresh the content on the screen, and the center button is wired to a GPIO pin that isn't used. You would only be able to use the single button for input. I don't see any other GPIO pins exposed, so it'd be tough to add extra buttons.

The ESP8266 has over a dozen GPIO pins, but any hobby kit like the ESP01, ESP12E, or ESP32 might not expose all of the pins. I can't tell which ESP module this board has on it, but if they're exposed you could solder an extra few buttons. They'll be really small solder points, though.

Reply on news.ycombinator.com