So I’ve just spent the last week investigating this topic as I’ve recently learned that a couple thousand dollars worth of our inventory is defective because of bad WS2812B LEDs.

EDIT: Yay, I got credited on Hackaday for this discovery!

Story Time: Discovering the Issue

(Feel free to skip to How to Identify Smart LEDs)

For a while we’ve been sourcing Smart LED flexible strips from various sources: eBay, Amazon, Aliexpress, Alibaba with varying results. All searches were done using the term WS2812B which is a 4 pin integrated smart RGB LED. It only needs V+, GND rails and Data in/out using a daisy chain structure and sending data down the line sequentially.

We now rely on a very stable supplier http://www.shiji-led.com/ who I’ve now purchased about 800m worth of LEDs without issue.

However, we also needed to make some custom double sided LED PCB bars so I went and sourced WS2812B LEDs directly from WorldSemi. These we spent about $800 worth on discrete LEDs, ~$100 on pcbs, and then another $1200 to have them assembled by a local fab. After everything was made we did some QC tests and found that the LEDs were failing in a weird way. Basically 1/10 LEDs had one or more internal diodes failing. Typically the blue diode, but occasionally red or green and in rare cases multiple or all three diodes failed.

Here are a couple videos of the behavior we were seeing:
Test 1:

Test 2:

The failure seems to be temperature sensitive, and in extreme cases pressure sensitive too. Being temperature sensitive means that from a cold start running full brightness white, in about 30 seconds you’ll start to see issues, but after it’s warmed up (~1-5min) those issues will disappear. If poked (pressure) it’s very likely you can permanently damage it. In the second video that one LED I poked never worked again and in fact the data line has been severed so subsequent LEDs will not work either.

This raised a lot of questions as to why our flexible strip LED supply was so reliable, but the LEDs bought directly from WorldSemi were not.

After investigating bad WS2812B chips and variations of the chip, we discovered there is a clone called the SK6812. And I’ve since confirmed that our stable flexible strip supplier has been selling these since early 2016 without telling their clients. Since they’re directly clones they felt there was no need to change the name of their product since it uses the same WS2812B protocol. However the main reason they switched was because they’re way more reliable than the WS2812B. There is also some slight improvements in the electrical characteristics, namely double the PWM frequency on the diodes (making them better for persistence of vision projects).

How to Identify (4 pin) Smart LEDs

I’ve compiled a imgur album, but I will summarize the main points:

* **WS2812(S)** is the 6 pin grandfather of the 4 pin smart LED, which follows a stupid naming convention. WS2812 is the name of the Smart LED with RBG diodes and an integrated IC. The name of that IC is WS2811. The reason this IC has 6 pins is because the WS2811 has separate power rails which is rather pointless. 6 Pins makes this easy to identify on it’s own.
* **WS2812B** is the next evolution with 4 pins. From the images you’ll see I’ve found at least 3 different variations of it. To identify this IC you’ll need a microscope to look at the bonding wires. I’m mainly focusing on the IC itself because of the bonding wire pad placement. You’ll see that it has 3 closely place pads along the short edge of the IC and on the opposite side 3 spaced out pads. Other indicators for variations of the LED package are the placement of the blue diode and shape of the pads inside the package.
* **SK6812** is the clone Smart LED and is most easily identified by the IC being rotated 90°. Under a microscope you’ll also see that the 3 closely placed bonding wire pads are now along the long edge of the IC.

One other way to differentiate the ICs is the size:

* WS2811 is 0.8×1.05mm
* WS2812B is 0.7×0.9mm
* SK6812 is 0.6×0.8mm

Differences between WS2812(S) vs. WS2812B vs. SK6812 LED Variants

Summary

If you’re buying this type of smart LED, make sure you get the good stuff. Verify with your supplier which IC is being used in the product because it seems like most manufacturers aren’t bothering to tell people they’ve switch to clone.

Credits

I’ve done most of this investigation through discussion on a few blogs. For more information on bad WS2812B LEDs and other failure modes check out:
https://wp.josh.com/2016/10/29/a-quick-test-for-crappy-ws2812b-neopixels/
For more information on the SK6812 go here:
https://cpldcpu.com/2016/03/09/the-sk6812-another-intelligent-rgb-led/
And for a summary on the differences between WS2811 and WS2812B, go here:
http://rgb-123.com/ws2812b-vs-ws2811/

After Credits

WorldSemi has yet to respond to my inquiry of the failed ICs but I expect with a small order (on their scale) I expect they’re probably not going to bother compensating me in any way.
Shiji has been very helpful and offered to supply me with the SK6812 discrete LEDs in tape/reel. I’m not sure if we’ll do that though because the cost of rebuilding the inventory is far higher than falling back to an old design using old inventory we already have.

Other things to note:

There are also a new generation of smart LEDs if you’re interested in looking them up. Some have RGBW diodes with varying white color temperatures. The next generation of ICs WS2813 and SK6822 which have dual data lines, this is a redundancy so that if one dataline breaks, you can still have signal going through. Interestingly WS2813 also incorporates the decoupling capacitor inside the LED package too allowing you to cram more LEDs closer together.
Another few other names to throw out: APA102 and SK9812. I haven’t looked into these in detail but I believe they’re also 6 pin packages with different protocols or data line redundancy.

Since I do want to get into this more eventually I wanted to give this a go to test the limits of my printer. The nice thing about this model which is great for printing is both the simplicity and the details included. She doesn’t have any super detailed clothing (you can hardly tell she’s wearing any), the shapes are simple and clean. However the smaller features such as the face, fingers and toes, come out very clearly. The best part is the hair is not overly complex and has almost no unsupported hanging parts. Unsupported hair is very common in most 3D modeled figures which make them undesirable for printing.

The biggest challenge here is printing a very complex shape that can’t support itself during printing. You can clearly see when you get to the ball she’s holding, there would be nothing for the printer to start building that on off of. As a result I have to add in a lot of support material, though this drastically reduces the surface quality. The only tweaking I did was to rotate the fingers so they were not hanging unsupported, but rather connected with her thigh and the ball. I also added a base since the model only stands on her toes and would have been very unbalanced without one.

The biggest issue with this print showed itself near the top of the legs, halfway through the thighs. It became apparent there was no horizontal support and the long tall thin features were vibrating like tuning forks whenever the print head was working on them, or worse when the nozzle caught the edge of print making it deflect almost 1cm. Eventually this broke the middle support piece and in my panic I tried holding it all together with my hands until it got to the crotch. However I soon realized that I had inadvertently been adding weight to the print bed. Since it’s only supported on one side it’s like a cantilever and deflects. This totally knocked the z-precision way off and it took me 20 minutes sitting there before I came to a solution that i didn’t have to hold up myself. I was very determined to see this print through since it was estimated to take 7 hours and I didn’t want to restart it. I’m actually pretty surprised it managed to finish at all.

Above: Figure test with support material, just as the print was completed.

Adding weights to the bed to offset the Z-layer height. Thank god I didn’t cash those coins in yet. Maybe I’ll keep the pennies after all since they make great weights in small units.

Quality of the upper body and head was so so, lots of room for improvement. The support material definitely takes it’s toll on surface quality. I can probably achieve a much better finish if I was going slower and did thinner layer heights.

The totally messed up area. So surprised it recovered itself. The hand on the right looks super creepy.

The legs up to the messed up point look very good though, the quality is pretty smooth and the detail in the toes is pretty good.

Oops. Looks like the only thing holding up the right section was the fingers, which broke while I was trying to remove the support material.

I learned a lot from this print. I now know to never, ever, eeeeeever touch a print while it’s printing. I also learned how to deal with misalignment in the Z-axis by adding a bit of weight to the bed to deflect it and give the nozzle a bit more room.

This also showed me that I need to consider horizontal support for tall unsupported structures. I definitely need to change my settings for support material from just ‘Line’ to ‘Grid’ so it’s sturdier.

If I get better software to slice this model, I will definitely attempt it again. If it turns out better than this, and remains unbroken, then I will look into post processing it as well.

Thanks again to Daniel Kreuter for the model, he does great 3D modelling so go check out his stuff on his site.

This wrist brace model was found on Thingiverse and I just gave it a go. The idea behind thermomolding is thanks to the properties of PLA which is the plastic I am using. PLA is liquified at around 200°C but actually softens at around 100°C. This means the plastic becomes malleable and can be shaped just by dunking it for a few seconds in boiling water. However, once it cools, it becomes quite stiff. The print itself it pretty simple, a mostly 2 dimensional object, though probably one of the largest I’ve done in terms of print bed surface area used. It took approximately 5.5 hours to print this, mainly due to the complexity of the lattice and the tool pathing required adds additional time.

This particular design has reinforced bars on the long axis of the brace. The notches on the sides are for adding straps to secure it.

The actual process of thermomolding was straight forward. Dunk the plastic in hot water and you’ve got about 10 seconds to shape it before it starts to stiffen up again. The pro: you can isolate which sections of the print are exposed to the hot water to control where the bending takes place. The con: molding it over your arm might burn you with hot water droplets.

For something that wasn’t designed for my hand size and was molded by an amateur, I’d say the fit is pretty good. Even if it’s not perfect, I imagine it’d be good to squeeze some padding in there.

Overall this print was very successful and I look forward to applying this technique to future prints.

Additionally, there were a lot of features I wanted to include in this that required various bits of hardware and logic to run it all. In retrospect, it would have been better to create a roadmap or outline of what I waned to accomplish, then work on individual parts and put it all together. However in some sense I thought it wouldn’t be too difficult to put it altogether in one go, so I just winged it. Working on multiple parts of the code simultaneously is not a good idea. Especially when you get bugs and have no idea where they are originating from.

Since I am writing this in post, it’s much easier for me to outline what has been done and give a proper road map of where this is going:

1. Create time keeping electronics that display on an LCD
2. Add LED Lamp brightness control (input from a potentiometer and create PWM output)
3. Add alarm functionality (manual alarm set + sunrise effect for alarm action)
4. Add button array for user input (include debouncing and UI logic)
5. Modify firmware to be able to display time, alarm time and modify either by using button inputs (remove manual time/alarm set)
6. Add switches to change modes (might be trouble since we’re running out of digital I/Os)
7. Add Winamp communication functionality to buttons in different modes (media buttons, playlist selection, volume)
8. Create script on computer to handle communication between MCU and Winamp
9. Add an independent power supply
10. Move all electronics to custom PCB boards or Prototyping boards
11. Design and print an enclosure to hold all the electronics.

Steps 1-3 were fairly easy since I relied heavily on this tutorial and the RTC library. Step 4 was a bit more challenging since at first I didn’t know how to approach debouncing on multiple buttons. This changed after I resolved to use a button map and used that for checking debouncing. Subsequently, once the buttons were figured out, adding alarm functionality for step 5 was a matter of playing with the RTC library a bit more in depth.

Step 6 has been one of the more difficult challenges. I needed two switches but only had one digital I/O left. Instead I opted to use analog pins, however due to the time it takes to process the value of these pins I think I was accumulating latency and errors. Since the switches are flip/rocker switches I don’t have to worry about debouncing and don’t need to check the value as often. As far as I can tell this bug has been resolved, but more testing is needed to see if it’s been totally cleared.

Step 7 and 8 have been mostly figured out on another program. Earlier this year I was already playing with trying to make an external Winamp controller, before I changed the goal of this project to a led lamp with the same functionality. In the other program I had used an PS/2 mouse as an input, utilizing a low level serial communication to interface the two buttons and mouse wheel. A python script was already in place to handle commands from the external MCU for media control, playlist selection and volume. This project was mainly dropped because originally I wanted to display currently playing tracks on the small 2×16 character LCD. However the majority of my music is Japanese and uses many asian characters that the LCD is not able to display, making it effectively useless. With the LED Lamp however, I plan to just use the computer monitor to display this data which means I can also modify the python script as an interface for selecting playlists.

Step 9 was something I had wanted ever since I installed the LED strip. The LED strip runs at 12 V and draws about 0.5 A, while the MCU and all the electronics run at 5V with about 100mA draw. Throughout the prototyping process I have been using my Hydra which is a programmable digital power supply which was set for 5V and 12V outputs. While great for prototyping, I did not want this project dependent on it if it were to become a stand alone product. Thus, added a 9V regulator and utilized the DC input socket of the arduino. I have a 12V 1.6A switching power supply which feeds the LED strip and this small regulator which in turn powers everything else.

Step 10 is one of the slowest parts mainly because I am unsure of how to proceed. Ideally I would have liked to have a custom PCB, but it’s incredibly expensive or time consuming to manufacture a single board. Therefore I have to resort to prototyping board on which I have very limited space. On this board I need to fit the power circuit, PWM transistors circuits, a pin socket for the potentiometer, and the real-time clock circuit (in which the coin cell batter takes up a huge amount of space). The button array has already been moved to a much nicer board with fancy colorful buttons.

Step 11 I will not be able to start until the last board is finished. Otherwise I won’t know for sure the dimensions I need to fit everything in.

In total there are 6 joint pieces per side; two at the base, two at the top, and two small nubs that will connect the reflector (not shown above). The two joint pieces at the base are pretty straight forward, the only important feature in this was to make the inner piece, which fits inside the copper tube, hollow so that the LED wires could be run internally. The top two joint pieces are the most complex of this design. The outer (left) piece has a large hole in the bottom so that it can be fixed to to the copper tube. The majority of this piece is hollow aside from the core which supports the axial screw. The hollow cavity is for extra slack in the wires so the connecting piece can rotate without tugging on them. The inner (right) piece is the crux of this joint system.

a) Hole for axial screw and bolt. b) slot for aluminum bar with a notch for the led wires. c) Hole for locking screw to fix aluminum bar in place. d) Slots to allow reflector to move from front to back. e) Mounting hole for telescoping bar.

The bottom reflector nub attaches to both ends of a brass bar and slots into the inner-top joint. The top nub also attaches to both ends of a brass bar which connect to the top of the telescoping bars.

The most difficult part of this was determining tolerances since I had not yet made anything that joined to other parts (of various materials). Additionally, I found that the 3D printed PLA changes over time, depending on the shape and the way it was extruded, it can shrink, expand and even warp if under stress over a long period of time.

The only way to resolve this was by trial and error. Parts were designed to fit snugly however I quickly realized I needed at least 0.1″ extra clearance to get ‘snug’ fits.

Joints aside, I also needed to get together the non plastic parts that would fill in everything in between. I sourced 2″ #6-32 machine screws + nuts, 1/2″ wood screws, a 1/4″aluminum bar, 2 1/8″ brass rods, a piece of 3/4″ diameter copper pipe (meant for plumbing) and a 8′ plank of 1×3, all from Home Depot. All the metal components were cut to size and polished with steel wool to achieve a very nice shine. The wood was cut to length, however I found despite being very careful to pick one that was straight, after cutting it it turned out to be very warped.

Additionally I picked up a sheet of poster board from the dollar store which would become my reflector and a set of bunny ears (telescoping TV antennas) which was the single worst purchase I’ve ever made. I don’t think I’ve been scammed worse than for the bunny ears which cost me ~$10 for 2 pieces of telescoping metal, a plastic shell and a copper wire that was screwed to the bottom of each ‘ear’ that went out to the TV. I had no idea these things were that simple and it certainly was worth much much less than $10.

With all the joints printed, it was a simple matter to assemble everything together. It really is amazing to be able to develop a concept in my mind and make it into a reality. However, like any other prototyping process, design flaws may not be realized until after the thing has been built.

Two of the major issues I found with my design are weight distribution and using screws in rotational joints. Looking back at the design, I had imagined that the majority of the lamp would be fairly light, but having used metal bars, pipes and rods with a light cedar base, the lamp is very top heavy. This was further compounded with the fact that the rotational joints had fixed slots for the nuts, meaning the joints would loosen every time it was moved. Even if the top heavy lamp managed to stay upright, there was a good chance the reflector section would just spin downwards since the joints would not be tight enough for friction to hold it’s position.

In the mean time, I’ve left the lamp mostly upright, which is still usable. Now that it’s been tested, I have a better idea of what I want to do and the design requirements for revision 2. Ah, the life of prototyping.

When I first saw this the method of achieving it was rather resource heavy. The RGB LEDs and the controller were expensive and there was no simple software to run it. To grab color data from the screen some people were even using VGA/DVI adapters to backward process what was being displayed. Additionally strips came in two kinds: addressable and non-addressable, meaning each LED could be addressed for RGB value or the whole strip would be the same color. Controlling this meant that either I would need to figure out the modulation pattern of whatever addressable strip I buy (at the time I wasn’t very experienced with this and didn’t understand the datasheets) or I would need a controller with multiple PWM I/Os to send to multiple short segments of non-addressable strips.

Recently however, I was hired to work on building RGB LED matrix displays. From that project I was able to salvage some leftover WS2811 strips. These strips are rather special since each LED pad contains an R, G, and B diode, plus a WS2811 chip. The chip is designed for controlling a lot of RGB LEDs (up to 4000) with a single microcontroller. This is achieved by sending a high frequency PWM modulated data stream down a single data line along an entire strip. The data stream addresses each LED pad and provides 4 bytes of data, 3 for each value of r, g and b and one for brightness. This makes the strip very versatile and easy to manipulate since the libraries already exist and all I need to do is program an array of data that gets sent to the strip.

To understand this I have been playing with the strip for a while using my MSP430 Launchpad to control it. The 43oh forums have been very helpful to get the appropriate libraries and code to start working with it. Example code can be found here: http://forum.43oh.com/topic/2852-ws2811-led-controller-driver/

Once I understood how to send data to the strip, the project turned toward grabbing data from the source and transmitting it to the microcontroller

On the software side I did a lot of research to find a simple way to achieve this. I ended up finding this project which was doing something very similar with a non-addressable RGB strip. http://siliconrepublic.blogspot.ca/2011/02/arduino-based-pc-ambient-lighting.html

Here they’re using Processing, a Java based IDE to take screenshots and analyze the average color of the screen. I took this and modified it to produce multiple regions to analyze instead of just one. To optimize the usage of the strip I needed to assign  a certain number of LEDs per region that was proportional to the area it was reading. Ideally I wanted to minimize the number of cuts I would need to make in the strip.

With my 1920*1200 resolution screen and a 60 LED strip I divided it accordingly. Alternatively I could have made it 6 regions of 6 at the top and each side would have 2 regions of 6, but the dimensions of the monitor only allows less than 34 LEDs in the horizontal axis without cutting it. Similarly in the vertical axis I can only put around 20 LEDs without cutting the strip.

Next post: Transmitting Data Serially and Putting It All Together (coming soon)

From our first year of Alumni we had a big trophy made of welded machine parts and making a weak attempt to combine all the aspects of our program, it also doubled as a beer bong. However what we wanted was something similar to WESST’s mascot Debbie, the concrete Donkey. It’s more iconic and easy to get attached to than a 50lb inanimate jumble of machine parts.

So after a committee meeting, I came of with a concept for a Mechatronics robot which followed the style of Pixar’s Wall-E.

I named him MAD-E (Mechatronics Autonomous Alcoholic Drink Expeller).

The idea was to have this little box robot follow people around autonomously and fire beer cans out it’s top. Some of the features i would have included are:

• RC chassis and battery
• MCU Autonomous and remote control
• Camera to use for tracking and obstacle detection
• LED Matrix display on the front to display faces and ‘emotions’
• Cooling unit that can store at least 6 cans
• Mechanism to transfer cans into cannon
• Some sort of mechanism to fire cans out the top
• A speaker and maybe a vibrator to get pre-programmed behavioral quirks in

Since the robot was fairly complex the plan was to execute project development over 3 years and commit a certain amount of resources every semester. However, the primary resource no one had to spare was time and almost no one was willing to commit to it. I would have been happy to lead the project, but with no members it was a flop from the beginning.

In the end we decided to keep the beer bong as our mascot, albeit we fixed it up so it was a bit more presentable.

I guess an autonomous beer dispensing robot that will follow me around will continue to be a dream for the future.

This one’s concept I saw a long time ago on Instructables.com but it happened to end up being the main project for one of my courses at SFU. I thought that was pretty awesome, do a project I wanted to try and also get marks for it too. During university I had very little time to work on personal projects so it was great to be able to allocate school work hours into something I wanted to do anyways.

The twist is that the course was Microcontroller and Embedded Systems, meaning we were not permitted to use the Arduino used in the step-by-step project tutorial online. We had to learn the architecture of another mcu and use that instead. However the university resources were pretty outdated and they wanted us to use a HCS12 Dragon (educational) board programmed via Freescale Codewarrior. Since I believe I’ll never use this architecture in the future, I wanted to learn something useful instead. I have already been working with Texas Instruments boards for a bit, their Launchpad development boards are stupidly cheap (so much I think they lose money on it). The MSP430 DSPs are equivalent to an arduino and I think are equally capable if not more (imo the biggest difference to me is the community and library support). However in recent years, TI started a new line of products called the Stellaris Launchpads. These are designed as introductory products to their high performance Stellaris products and Stellarisware libraries. I’ve used this before on a Intelligent Display Module Stellaris dev kit (RDK-IDM) for my Capstone.

The Stellaris Launchpad (EK-LM4F120) carries a mcu (LM4F120H5QR ) with absolutely amazing specs for a dev kit that only costs $13. When I bought it they were just introducing it and had a promotion and I got 2 of them for $5 each.

A quick overview of it’s specs:

So I’m using this mcu to control my LED Matrix Cube, it has just enough I/O pins to do it too. My goal is to utilize the DSP parts of it to run a spectrum analyzer on an audio feed and play an animated equalizer on the cube.

After meeting the minimum requirements for the course I have not touched this in a while. During the course I hit a lot of bugs in the firmware that I was not able to resolve and I suspect I will need to rebuild the firmware from scratch in order to continue the project.

The project goal is purely aesthetic and the cube will not really serve any purpose, which is why this is a back burner project. I do hope to eventually get back to this so I can learn more about mcu architecture and embedded programming. I did at least put in the effort to create a CAD model of a nice case and had that made from laser cut acrylic. As you can see in the photo, it has a nice gloss black base (missing some square brackets that are used to keep the lid closed and act as feet). Additionally I had cut a clear acrylic box that goes around the LEDs to protect it.

Story time: About 7 years ago my brother gave me a pair of very awesome Sennheiser ear buds, CX55’s, for my birthday. I had been looking at getting a decent pair for a while and marked out Sennheiser, but wasn’t really sure which model to get. I had my eye on a very expensive pair around ~$120 at the time. Pretty pricey for a pair of ear buds for a high-schooler. So my brother got me a pair, but one of the cheaper models. The actual value of the CX55’s was around $80 at the time. I was a little disappointed at first, but after using them for a bit, I came to love them. The quality was absolutely amazing. I was very sad when the connector wore out 2 years later and I had to start looking for another pair.

After doing a bit of research I was shocked to find the exact same ear buds for sale on Deal Extreme, the cheap Chinese supplied online shopping site. Everything about the Cx55’s were identical, except the price: $12.49 and down to $10.60 if I bought it in bulk… which is exactly what did. I figured if even if they were crap and lasted a couple months, if 5 pairs can last as long as the original for te same price then it was fine.

When they arrived I did a thorough check. The only difference I could find was that the shape of the plastic nub where the L and R wires split was a different shape. The quality of sound, I could hear no difference than the original.

I started using these regularly, but as expected I found that they didn’t last too long, 6 months at best with my regular wear and tear. The common defect among them was the audio jack.

After getting a 3rd gen iPod in 2011 I noted that the audio jack on it was pretty crap. No matter what ear buds I used the connection occasionally disconnected in one channel or both. At the time I thought it was a defective iPod, but after recently getting a similar behavior from a 3rd gen iPhone I believe it’s just a shitty design. I think the apple ear buds have just a slightly longer jack which makes the connection more secure.

After that I decided to try something different. I noted that the iPod docks are often incorporated into speakers which means the proprietary Apple connector feeds audio channels. I bought a cheap connector off of Deal Extrememe, cut the connector and opened it up to find exactly what i was hoping for. So I spliced one of the dead earbuds and tested it out. Perfect audio coming through.

The one and only problem: no volume control. By default the Apple devices will max out the audio signal through the dock connector since it’s expected that the dock device (a speaker) would have it’s own amplifier and volume control. This meant if I wanted to use this I needed a in-line volume controller. I researched a little into this and found I could easily just use a potentiometer per audio line since I have no need of amplification, I only need a negative gain. However, after playing around splicing a few of the other ear buds I know all to well that the wires inside them are extremely delicate and require very good strain relief. This meant I couldn’t just find an existing in-line volume controller and splice it between the cables. I’d never be able to get a decent strain relief that wouldn’t look super ugly.

TL;DR: I bought great quality sounding ear buds with defective connectors and want to splice an Apple connector instead of an audio jack but I need a volume controller with good strain relief.

After purchasing my 3D printer I can now design a plastic module that will hold a pair of potentiometers and provide decent strain relief.

In my room I have a lot of blank wall space which I like to use like a reflective surface for lighting as previously mentioned. Additionally, my bed has no headboard, but has a 4″ ledge the length of the wall about a foot above it, perfect for mounting my lamp. However I don’t want to mount directly to it since it has a nice finished surface, so I’ll need a base to my lamp that will sit on it.

This is my target design. The idea was to create 3D printed hinges and connectors and whatever raw materials I could easily pick up at home depot. In this case I planned to pick up a 2×4, some sort of pipe and metal bars. The design of the reflector (not a shade) was the more complex aspect of this. I wanted the reflector to be capable of multiple orientations for different lighting dynamics. In the top left you can see in one mode it reflects upwards, directing the light toward the wall and ceiling for ambient lighting. In the next the reflector is straight and covers a larger area downwards. Below that the reflector is curved to focus light right below it.

To achieve these modes I designed a custom hinge (mid bottom). In the stationary half it connects to the pipe holding it up through which wires are fed for a clean wireless appearance. The LED strip itself is adhered to an aluminum bar slotted into bottom of the rotational part of the hinge.  At the top of the rotational part, there is a slot that splits into two directions. This is for the base of the reflector so that it can be changed to reflect forwards or backwards. At the very top of the rotational part there will be a hole to add a small telescoping pipe. This will attach to the top of the reflector so that the curvature can be adjusted.

The electronics enclosure will be set in the middle of the base with wires being fed under or behind it in channels. Since the the ATmega is part of the arduino development board, it’s easier to run everything with pin connectors than create a custom pcb. This also allows for different orientations of circuits so that the LCD can be perpendicular to the board and secondary boards for buttons, clock electronics can be mounted to the walls. The arduino will be at the bottom with a power splitting circuit to split off a 5v feed to run the microcontroller from the 12V supply.

The diagonal split between the top and bottom of the electronics box is to maximize access to the electronics while minimizing the number of pieces. There is a lip on the front edge so that the pieces will hinge and rotate around that edge. This is to minimize the length of the wires attaching the button array and LCD to the bottom half of the electronics and allowing access without having to disconnect them.