Fullscreen HTML5Med Res (??MB)HTML4

Max’s Cool Beans By Max the Magnificent Flashing LEDs and drooling engineers – Part 24 W hat’s that, you say? You want to hear a joke? Well, by some strange quirk of fate, I happened to hear something humorous on the radio on the way into work this morning. A young lad calls his mum to say: ‘Remember how my teacher said I’d never be any good at poetry because I’m dyslexic? Well, I just made two candlestick holders and a vase, and they look great!’ Stop groaning. You’ll be telling this yourself before the day is out. 40 44 39 38 24 25 26 41 9 18 27 2 10 17 16 1 8 15 37 23 36 7 14 35 22 43 34 3 0 6 5 13 29 4 11 12 20 30 31 42 21 33 28 19 32 Wow! Wireless LEDs Just when you think you’ve seen it all, along comes something new to surprise you. Recently, for example, a friend introduced me to the YouTube channel of someone who goes under the moniker of Atomic14, which is sort of clever because 14 is the atomic number of silicon. While I was meandering around this channel, as is my wont, I saw a video showing wireless LEDs: https://bit.ly/3y5HUlX It has to be said that these little scamps, which are available for purchase from AliExpress (https://bit.ly/3IyqpiN), look rather spiffy (Fig.1.). I can easily imagine using these to create some little gadgets, gizmos, doodads, and doohickeys to entertain young kids – or even (especially) old kids like myself, now I come to think about it. Fig.2. SMAD segment map (left) and board annotated with rings (right). multiple potentiometers in parallel. So, you can only imagine my surprise to discover that most of the emails I received regarding that column asked, ‘But what about SMADs?’ In fact, while I’m thinking about it, I’m even receiving emails from someone in Russia who used our SMAD schematic diagram to build a giant SMAD about 500mm square out of discrete WS2812 tricolour LEDs. He constructed this with his daughter, and they are constantly clamoring for more code examples. Good grief! It’s not like I’ve forgotten these bodacious beauties. It’s just that I’m juggling too many balls and I can’t juggle. Actually, that’s not strictly true – I can juggle ten fine china plates, but only for a very short period of time; that is, the time it takes for them to go up in the air and come back down again (hey, it’s a start). In previous columns (PE, October and November 2021), we discussed a variety of simple windmill-type effects and an alien countdown timer effect. We Fig.1. Wireless LEDs – I almost believe they were invented specifically for me... presented these effects on two pseudo robot heads, each of which is equipped with two SMADs as eyes. The SMADs on one of these heads are augmented with 29-segment shells, while the other head flaunts 45-segment shells. Two columns ago (PE, December 2021), we had our heads talking to each other in Morse code. In fact, we even have a video of this taking place (https://bit.ly/3oBwemz). Now it’s time to get a little more adventurous. As a starting point, we might decide to consider the LEDs on the SMADs as being presented in the form of five concentric circles or rings (Fig.2). Do you recall the way in which we defined the ‘patterns’ in our windmill and countdown effects as a two-dimensional array called EffectMap[][]? Well, we’re going to use the same technique here (Fig.3). The first dimension specifies the number of rows in the array. This corresponds to the number of patterns in our effect, which will be the five rings, in this case. The second dimension specifies the LEDs that appear in each ring. Ring 0 in the center of the SMAD is formed from just one LED (shown as 0 in Fig.2). Ring 1 is composed of eight LEDs (shown as 1 through 8 in Fig.2). Rings 2 and 3 each contain 16 LEDs. And Ring 4 boasts only four LEDs. This means that the maximum number of LEDs in any of the rings is 16. The reason we set this dimension in the array to be 16 + 1 = 17 is because we use the first element in each row to define the 52 Practical Electronics | February | 2022 I’m a SMAD boy! We covered a lot of ground in my previous column (PE, January 2022), including toroidal transformations, gas-heated soldering irons, 1970s oil-wheel disco lighting effects, the 4-axis joysticks we’ll be using to control our servo-driven robot heads, and some interesting considerations with respect to connecting # d efine N U M _P A T T E R N S _I N _E F F E C T 5 # d efine M A X_L E D S _I N _E F F E C T 16 const uint8_t E ffectM ap [N U M _P A T T E R N S _I N _E F F E C T ][M A X_L E D S _I N _E F F E C T + 1] = { { 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}, / / { 8, 1, 2, 3, 4, 5, 6, 7, 8, 0, 0, 0, 0, 0, 0, 0, 0}, / / {16, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24}, / / {16, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40}, / / { 4, 41, 42, 43, 44, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0} / / }; Fig.3. Specifying which LEDs appear in what rings. R ing R ing R ing R ing R ing 0 1 2 3 4 by the folks at Adafruit (Fig.5). We can use this table to remap the linear values we would like to use into their gammacorrected counterparts that will provide us with the colours we want to see. Personally, I thought our original gamma correction evaluations were rather successful, so I decided to apply this to our most recent SMAD experiment. In this case, for each robot head, I displayed the non-gamma-corrected colours on one SMAD and the gamma-corrected colours on the other SMAD. One point that gave me pause for thought was that, in our original tests, we started off with our LEDs at full brightness insofar as the values required to represent the selected colour. When it comes to our SMAD experiments, however, I’ve been using a ModifyBrightness() function to lower the brightness of the LEDs so I don’t wash out the sensor in my camera while taking videos. This led me to wonder... should we modify the brightness and then perform the gamma correction (this is the way you’ll see things in file CB-Feb22-04. txt), or should we do things the other way around. In the end, it didn’t matter, because the results looked horrible both ways. I think one of our problems is that we are working with 8-bit values for our colour channels, which is too low a resolution for what we’re trying to do. Also, we are using an overly simplistic approach (we should really be applying custom gamma correction functions for each channel based on the specific characteristics of our LEDs). I’m sure we could do better if we were desperate, but I wasn’t desperate enough, so I decided to chalk this one up to experience and move on with my life. For my third experiment, rather than change the colour of only one + 1] = group at a time, I decided to change the colours of all the groups simultaneously (file CB-Feb22-03. txt). One point we should note is that, thus far, we’ve been displaying the same colour patterns Fig.4. Specifying which LEDs appear in what groups. on all the SMADs. It’s number of LEDs in that ring. Also, any really interesting to see how different ‘0’ values are used only as placeholders the colours look when diffused by the (apart from the left-most ‘0’ associated 29-segment shells versus the stained-glass with Ring 0, of course). (some may say kaleidoscopic) effect ofMy first test was simply to illuminate fered by the 45-segment shells. Speaking the rings with white light in a sequence of which, for the benefit of your cogistarting with Ring 0 and radiating out to tations and ruminations, I just took a Ring 4. You can feast your orbs on the video that shows all of the SMAD-relatfull sketch (program) by downloading ed effects just discussed in this column: the code (file CB-Feb22-01.txt) from the https://bit.ly/3GwQb5f February 2022 page of the PE website: https://bit.ly/3oouhbl Gorgeous gammas We introduced the concept of gamma correction in the context of my 12 x 12 I’m so random ping-pong ball array in an earlier Cool For my next experiment, I decided to Beans column (PE, January 2021). As gather my LEDs into like-minded groups. you may recall, our human eyes have There are seven of these groups numevolved to accommodate a huge dybered from 0 to 6. Group 0 corresponds namic range from moonlight to sunlight to Ring 0 in our previous example and and – as part of this – they have a sort of therefore contains just one LED. Simibuilt-in non-linearity. Gamma correction larly, Group 1 corresponds to Ring 1 in Fabulous fades involves tweaking the values of the red, our previous example and therefore conIf we journey once more back into the green, and blue channels in our LEDs to tains the 8 innermost LEDs in our eight past to the early days of our 12 x 12 pingcorrect for this non-linearity. radiating spokes. pong ball array experiments, you may In our case, the way we’ve been doing In the case of Ring 2, we are going to remember when we introduced some this in our programs is to use a gammadivide this into two groups – Group 2 rather cool functions that allowed us correction cross-reference look-up table, contains the eight LEDs in the eight little to fade from one colour to another (PE, the contents of which were kindly defined triangular segments, while Group 3 conOctober 2020). tains the eight outermost LEDs in our eight radiating spokes. Similarly, we are going to divide Ring 3 into two groups const uint8_t GammaXref[] = { – Group 4 contains the eight ‘anticlock0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, wise’ LEDs in our eight outer ring seg0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, ments, while Group 5 contains the eight 2, 3, 3, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 5, 5, 5, ‘clockwise’ LEDs in these segments. Last, 8, 8, 9, 9, 9, 10, 5, 6, 6, 6, 6, 7, 7, 7, 7, 8, 10, 10, 11, 11, 11, 12, 12, 13, 13, 13, 14, 14, 15, 15, 16, 16, but certainly not least, Group 6 contain 17, 17, 18, 18, 19, 19, 20, 20, 21, 21, 22, 22, 23, 24, 24, 25, the four LEDs corresponding to Ring 4 25, 26, 27, 27, 28, 29, 29, 30, 31, 32, 32, 33, 34, 35, 35, 36, 37, 38, 39, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 50, in our previous program (Fig.4). 51, 52, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 66, 67, 68, 69, 70, 72, 73, 74, 75, 77, 78, 79, 81, 82, 83, 85, 86, 87, 89, What we are going to do is to ran90, 92, 93, 95, 96, 98, 99, 101, 102, 104, 105, 107, 109, 110, 112, 114, domly select a group and illuminate all 115, 117, 119, 120, 122, 124, 126, 127, 129, 131, 133, 135, 137, 138, 140, 142, 144, 146, 148, 150, 152, 154, 156, 158, 160, 162, 164, 167, 169, 171, 173, 175, the LEDs in that group with a randomly 177, 180, 182, 184, 186, 189, 191, 193, 196, 198, 200, 203, 205, 208, 210, 213, selected colour, and to continue to do 215, 218, 220, 223, 225, 228, 231, 233, 236, 239, 241, 244, 247, 249, 252, 255 }; this over and over again. If you wish, you can peruse and ponder this program (file CB-Feb22-02.txt). Fig.5. Gamma-correction cross-reference look-up table. # d efine N U M _GR O U P S 7 # d efine M A X_L E D S _I N _GR O U P 8 const uint8_t E ffectM ap [N U M _GR O U P S ][M A X_L E D S _I N _GR O U P { { 1, 0, 0, 0, 0, 0, 0, 0, 0 }, / / Group 0 { 8, 1, 2, 3, 4, 5, 6, 7, 8 }, / / Group 1 { 8, 9, 10, 11, 12, 13, 14, 15, 16 }, / / Group 2 { 8, 17, 18, 19, 20, 21, 22, 23, 24 }, / / Group 3 { 8, 25, 27, 29, 31, 33, 35, 37, 39 }, / / Group 4 { 8, 26, 28, 30, 32, 34, 36, 38, 40 }, / / Group 5 { 4, 41, 42, 43, 44, 0, 0, 0, 0 } / / Group 6 }; Practical Electronics | February | 2022 53 Fig.6. A CRT-based television displaying white noise (‘static’). When I was first trying to wrap my brain around this, it actually took a bit of thought to decide how to perform a fade over a specific number of steps. Remember that, for each step, we first have to split each 24-bit colour into its three 8-bit red, green, and blue colour channels. For each of these channels we calculate the new value based on the start value, the end value, the total number of steps, and the current step. Finally, we gather the three 8-bit channels back into a 24-bit colour. Phew! Happily, we can reuse tweaked versions of our previously created functions. For our fifth and final experiment in this column, I removed our ghastly gamma correction and added some fabulous fades (file CB-Feb22-05.txt). As you’ll see in the video, we start by creating a palette of randomly selected colours, then we assign these colours to the various groups and display the groups. While keeping track of the original palette, we generate a new palette, assign these colours to the various groups, and then fade from the old colours to the new colours. And, of course, we keep on doing all of this over and over again. Once again, if you watch the aforementioned video, I think you’ll find it really interesting to see how different the colours look when presented via the 29-segment versus the 45-segment shells. That’s a lot of noise! When most people hear the word ‘noise,’ their knee-jerk reaction is to think of a cacophony of sound, but noise comes in many forms. When I was a young lad circa the 1960s, our family’s analogue television contained a bunch of vacuum tubes along with a cathode ray tube (CRT) for the display. At night, after the TV stations had stopped broadcasting, the screen filled with ever-changing random patterns of ‘snow on black’ flickering pixels (considering this was so dynamic, it’s somewhat paradoxical that it was called ‘static’). This visual representation of noise was complemented by an audible hiss (Fig.6). 54 Actually, the origin of the idea to apply Both of these realisations – the visual dither is rather interesting. In the early static and the audible hiss – are of a days of WWII, airplane bombers used form known as ‘white noise.’ This is somechanical computers (essentially, boxes called because it includes all frequenfilled with hundreds of shafts and gears cies (more correctly, it has equal power and cogs) to perform navigation and bomb in any frequency of a given bandwidth), trajectory calculations. The strange thing which makes it analogous to white light, was that these computers were found to which is a mixture of all visible waveperform their calculations more accuratelengths of light. ly when the airplane was in the air with In audio engineering, electronics, physits engines running then when it was ics and many other fields, we use terms on the ground with its engines turned like ‘white noise’, ‘pink noise’, ‘brown’ off. Eventually, engineers realised that (or ‘Brownian’) ‘noise’, ‘blue noise’, ‘violet the vibration from the aircraft reduced noise’, ‘grey noise’ and even ‘black noise’. the effects caused by ‘stiction’ (sticky These terms refer to the different power moving parts), so, instead of moving in spectrums associated with the noise sigshort jerks, the vibration caused them to nals. For example, while an instance of move more smoothly. In order to replicate white noise might be the sound of falling this effect on the ground, small vibratraindrops, an illustration of pink noise ing motors were built into the computcould be the ‘chuff, chuff, chuff’ sound ers. The vibration from these motors was of an old-fashioned steam engine. called ‘dither’ from the Middle English You might think that noise is a bad thing verb ‘didderen,’ meaning ‘to tremble.’ to have in a system, but – if you know what you are doing – it can be useful on occasion. For example, when a conYou gotta love lava lamps tinuous signal in the analogue domain The reason for my waffling on about noise is converted into a corresponding repis that it features in the answer to a poser resentation in the digital realm, it’s conthat has long puzzled me. I’m sure you’ve fined to a number of discrete states called seen a lava lamp at least once on your quanta. This results in something called journey through life. These little beauquantisation error. ties were invented by the British entreThe term dither (or dithering) refers preneur Edward Craven Walker in 1963. to intentionally overlaying the original For quite some time, they were considsignal with a small amount of noise, ered to be amazingly cool and – dare I which serves to randomise the quantisay it (yes, I do) – groovy. sation error. In the case of a computer, A lava lamp consists of a glass vessel dithering can be used to create the illucontaining a coloured waxy mixture at sion of greater colour depth in images the bottom. The rest of the vessel conon systems with a limited colour palette. tains a clear or translucent liquid. When Dithering can also be used to minimise an incandescent light is placed under visual artifacts like banding in images. the lamp, in addition to lighting it up, Another example that may be a little the heat causes balls of the coloured closer to home is when performing anawaxy mixture to decrease in density and logue-to-digital conversion. Suppose we randomly rise and undulate their way have an 8-bit analogue-to-digital convertthrough the clear liquid. er (ADC). Now suppose that we superimpose a noise signal on top of the analogue signal feeding the ADC, where the peak-to-peak amplitude of this noise signal is equivalent to one leastsignificant bit (LSB) of the digital output. If we now sample the signal at four times the original sampling rate and then apply a low-pass filter to the digital output (where the simplest low-pass filter would be to average the four samples), the result is to make it look like we have a 9-bit ADC. (This is a bit too complicated to take any further here – you’ll just have Fig.7. The FastLED NoisePlusPalette program meets my to take my word for it!) 12 x 12 ping-pong ball array – do watch the video! Practical Electronics | February | 2022 As I mentioned in my previous column, on several occasions I’ve seen an effect presented on arrays of tricolour LEDs that is reminiscent of the oil wheels used in 1970s discos. This effect is also evocative of the undulations of the waxy balls in a lava lamp. The FastLED library (https://fastled. io/) offers a great way to control tricolour LEDs like the NeoPixels I used to populate my 12 x 12 array of pingpong balls. It turns out that there are a couple of example programs called ‘Noise’ and ‘NoisePlusPalette’ that are provided with this library. These programs use special noise functions that are based on something called ‘Simplex noise’, which was created by Ken Perlin in 2001, and which provides a method for constructing n-dimensional noise. The great thing about Simplex noise is that it generates values that are pseudorandom overall but that are locally related at the same time. The easiest way to visualise the way this works is to imagine that you are looking at a giant oil lamp. If you were to take a series of photographs ten minutes apart, it would appear that there was no relation between the images. By comparison, if you took a snapshot every second, you could see how each image was related to the one that came before and the one that came after. That is, what you are seeing is both random and constrained at the same time. The bottom line is that I tweaked the existing NoisePlusPalette program to drive my 12 x 12 ping-pong ball array (Fig.7). To be honest, in addition to restricting the program to using only its ‘lava’ palette, the main thing I had to change was the XY() function that accepts X and Y coordinates as inputs and returns the number of the NeoPixel in Want to build your own amazing array? All the details are in previous Cool Beans columns, starting in March 2020. the chain (my NeoPixels are physically wired – daisy-chained – together in a serpentine pattern). One thing I should note is that I originally tried to run this program on the 32-bit Seedunio XIAO microcontroller that I introduced in an earlier column (PE, July 2020), but... nothing happened. I thought I’d messed something up while performing my tweaks, and I was ready to pull out what little I have left of my hair when I thought to question my chum, Steve Manley. Steve did what I should have done, which was to have a quick Google. This revealed that there’s a known problem with the XIAO and the FastLED library. A little later, Steve emailed me to say that he’d discovered a workaround (https://bit.ly/3oJuvNn), but he was too late. By this time, I’d already swapped out my XIAO for an Arduino Uno, after which everything worked just fine and dandy. I’m really happy with the result, which you can see in a video I posted to my Cool Beans YouTube channel (https:// bit.ly/3ysgEhP). Feel free to peruse and ponder the modified version of the program (file CB-Feb22-06.txt) but remember that I’m not the author of this code, just the tweaker. Don’t lose your head The definition of a cliffhanger is an adventure serial or melodrama that’s presented in installments, each ending in suspense. When my mum was a little girl about 10 years of age, her older brother used to take her to the kids’ show at the local cinema on Saturday mornings. The show was about two hours long and included a bunch of items, including a short thriller, like an episode of Flash Gordon or Buck Rogers (the old blackand-white versions from the late 1930s), which ended on a cliffhanger. My mom says she could barely control her excitement and it made her want to see the next episode all the more. If only we could regain that sort of anticipation and suspense here. I know what we can do... in my previous column, I promised to show you the progress we’ve been making with regard to animating our robot heads. When I say ‘we,’ I mean me and my friend Steve Manley. And when I say ‘all of the progress we’ve been making,’ I mean all of the progress Steve has been making because he’s storming ahead with something so super that you will squeal with delight when you see it. Sad to relate, however, this will have to wait until my next column (Cliffhanger – Tra-la!) Cool bean Max Maxfield (Hawaiian shirt, on the right) is emperor of all he surveys at CliveMaxfield.com – the go-to site for the latest and greatest in technological geekdom. Comments or questions? Email Max at: max<at>CliveMaxfield.com Practical Electronics | February | 2022 www.poscope.com/epe - USB - Ethernet - Web server - Modbus - CNC (Mach3/4) - IO - PWM - Encoders - LCD - Analog inputs - Compact PLC - up to 256 - up to 32 microsteps microsteps - 50 V / 6 A - 30 V / 2.5 A - USB configuration - Isolated PoScope Mega1+ PoScope Mega50 - up to 50MS/s - resolution up to 12bit - Lowest power consumption - Smallest and lightest - 7 in 1: Oscilloscope, FFT, X/Y, Recorder, Logic Analyzer, Protocol decoder, Signal generator 55