Fullscreen HTML5Med Res (??MB)HTML4

• Support for In-Circuit Se rial Programming (ICSP) • 32 LEDs on each side of each PC board (64 LEDs per board , 192 LEDs total) • Displays a one-kilobyte image (32 LEDs x 256 radial “ra ster lines”) • All LEDs can be driven with 20mA at 100% duty indefinitely . This produces a very bright im age. • Firmware shuts the circui t down automatically when the voltage gets too low, to prevent damage to rechargeable battery packs • The PC boards fit 26-in ch bike wheels or larger. BIKE WHEEL POV DISPLAY This project uses POV to produce a spectacular glowing display on a rotating pushbike wheel as you ride along. So what is POV? It stands for Persistence Of Vision. It’s a term that’s applied to devices that rely on the human eye’s tendency to “see” an image for a short time after it has disappeared. Designed by Ian Paterson 26  Silicon Chip siliconchip.com.au H OW WOULD YOU LIKE to own the most talked-about pushbike in the school/street/suburb. . . galaxy? Build this POV display and you’ll be well on the way. You really have to see it to believe it – and we’ve even made it easy for you. As well as the images printed here, there are several more you can view online at www.ianpaterson.org/projects OK, you’ve now seen them and you’d have to agree that they look pretty spectacular, right? You want to do the same for your pushie? Just make sure you keep it chained up because everyone will want it . . . Persistence of vision You probably don’t realise it but you use POV every day – when you watch TV. Movies also take advantage of this phenomenon. The TV and movie picture is not continuous vision – rather (in the case of TV), 25 individual pictures are displayed every second. But your eyes and brain cannot follow the 25 individual frames of picture per second – instead, they “fill in the gaps” and you “see” full motion, non-jerky video. If you slowed those frames down to, say, 10 per second, you would be able to see the period between each frame and it would become jerky – just like the old-time movies where the hero moves like a Thunderbirds puppet. Let’s take this one step further. Say you had a moving light – we’ll make it a LED because they can be turned on and off very quickly – which you flashed on, very briefly, once per second. You’d see this as flashes of light moving along a path. If you changed that to, say, 10 flashes per second, you’d probably still see flashes but very much closer together. Make that 50 flashes per second and the flashes would all meld into one another. You’d see it as a continuous line of light – even though your brain knows full well that it is flashes you are viewing. That’s persistence of vision and this is the basic theory behind this project. Rows of LEDs are made to flash too quickly for your brain to process, so they appear to be permanently on. The rows of LEDs are mounted on PC boards fixed to a bicycle wheel, so they follow a circular path as the wheel rotates. By using some clever circuitry to switch the LEDs on and off at particular moments, a pattern or picture can be created – in fact, the display is almost unlimited. It can be everything from geometric shapes to text, cartoon characters and even very high contrast pictures (see examples below). In a nutshell . . . The display consists of three PC boards, each with a row of 32 LEDs on each side (a total of 64 LEDs). These boards are mounted radially in/on the spokes of a pushbike wheel and each has a battery pack mounted near the wheel’s hub. Talk about WOW! factor: this 3-high static display uses different coloured LEDs in each wheel to reveal three different patterns. The rider powers the one wheel & the second & third wheels are driven by friction between the tyres. A Hall Effect sensor measures the rotational speed of the wheel by sensing a small magnet fixed to the bike frame. This sensor sends speed pulses to a microcontroller, which then turns the individual LEDs on and off in such a way that a static image appears to float inside the wheel. Circuit description Fig.1 shows the complete circuit for one POV display module. Three such modules are required, arranged so that each is mounted 120 ° from the other around the wheel, between the spokes. With the exception of the trigger magnet and battery pack, all components mount on these three PC boards. Here are just a few of the images generated by the author: (from left) pagan star, ET, invisible unicorn, Saturn & evolution! siliconchip.com.au September 2007  27 28  Silicon Chip siliconchip.com.au SC 2007 1 1 OUT 3 RA6 RA3 RB0 Vss 5 RB5 RA7 IC1 PIC16F628A MCLR PGD PGC Vdd 14 1k 21 23 4 22 2 11 10k GND 3 10k 10 F 16V IN 16 15 2 100nF OUT REG1 LM2931AZ-5 BIKE WHEEL POV DISPLAY MODULE 2 GND (HALL SENSOR) HS1 DN6851 6 4 Vcc 13 2 12 3 4 ICSP SOCKET CON1 10k +5V 24 O7 O6 O5 O4 O3 O2 O1 O0 OE Rext LE CLK SDO SDI GND 1 O15 O14 O13 O12 O11 O10 O9 O8 IC2 STP16C596 Vdd 2 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 1 K K K K K K K K K K K K K K K K A K A K A K A K A K A K A K A K A K A K A K A K A K A K A K A K 3 DN6851 LED31                 LED1 A A A A A A A A A A A A A A A A CHAMFER LED32                 LED2 23 4 3 22 2 IN COM 24 Vdd OE Rext LE CLK SDO SDI GND 1 OUT O15 O14 O13 O12 O11 O10 O9 O8 O7 O6 O5 O4 O3 O2 O1 O0 IC3 STP16C596 LM2931AZ-5 21 1k +7.2 – +8.4V* 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 K K K K K K K K K K K K K K K K LED63                 LED33 A K A K A K A K A K A K A K A K A K A K A K A K A K A K A K A K A K 7.2–8.4V* NiMH BATTERY S1 POWER A A * DO NOT USE RED LEDS WITH 8.4V BATTERY A A A A A A A A A A A A A A LEDS LED64                 LED34 The modules, or PC boards, each contain 64 high-brightness LEDs, 32 on each side. A LED on one side is connected in series with a LED on the other, so that the same image is seen on both sides of the bike. The LEDs are under the control of a PIC16F628A microcontroller (IC1). It is this microcontroller which not only stores the image to be displayed but outputs it to two STP16C596 shift registers (IC2 & IC3) which in turn drive the LEDs. If each LED pair was driven with a dedicated output line, the microcontroller would have to have a very large number of output lines. Hence this circuit uses 16-bit constant-current LED sink drivers (IC2 & IC3) which can drive 16 outputs and allow multiple devices to be cascaded together. The STP16C596 also has a separate storage register that allows one set of data to be displayed while the next set is being loaded. Four lines are used to control the LED outputs: serial data input (SDI), clock (CLK), latch enable (LE) and serial data output (SDO). Each pulse of the clock line causes the data to be “shifted” over by one place and each pulse of the latch enable line causes the LED outputs to reflect the contents of the shift register. One kilobyte of image data is stored in the program memory area of the microcontroller and is read by way of a look-up table. The firmware uses four interrupt routines: • one to provide the time interval between radial raster lines; • one to increment a counter for timing the wheel rotation interval; • one to reset all counters and update the raster interval value every time the Hall Effect sensor is triggered; and • one that shuts down all LEDs when the battery voltage gets too low. In fact, after the initial start-up routine, virtually every part of the firmware’s execution runs inside an interrupt routine. We haven’t yet mentioned the DN6851 Hall Effect sensor. Its purpose is to measure the speed of the wheel Fig.1 (left): one POV display module – three are required for the whole project. With 64 LEDs per module it looks daunting but there are only 12 other components in each! siliconchip.com.au and supply the appropriate timing pulses to IC1. It is triggered each time it passes a small magnet attached to the bike frame. Its output pulse is sensed by input RB0 on IC1 Timing values for the radial raster line interval are retrieved from a look-up table that exists in the microcontroller’s program space. Data for the look-up table is generated with a QBasic program, although you only need to run this program if you want to experiment with different timing values. When using a 7.2V battery pack, it’s better to use a low dropout regulator such as the National Semiconductor LM2931AZ-5 than the commonly used 78L05. It will continue to provide a solid 5V for the microcontroller even when the battery is at 6V. This is important because if the supply voltage to the microcontroller drops, so does the internal reference voltage which would prevent the voltage sensing routine from working properly. A number of flow charts have been created to illustrate the logic in Spoke POV’s various firmware routines but since our space is limited, these can all be accessed on the website mentioned above. QBasic programs In addition to the microcontroller firmware, two Qbasic programs are required for setting the timing values and converting image data so they can be incorporated in the firmware. POVSLOPE.BAS creates the timebase look-up table. The table produced by this program is linear, so the only parameters one needs to be concerned with are slope and offset. Note that the timing data supplied in the sample firmware is reasonably accurate so you should only use POVSLOPE.BAS if you plan to experiment with different timing values. POVIMAGE.BAS is used to convert a raster image into radial data in the form of a series of “RETLW B’xxxxxxxx’;” commands that can be copied and pasted directly into the POV assembly code. The image data is read one pixel at a time as a series of 32 concentric rings. Each group of eight rings ends up occupying one memory page. Because of the limitations of QBasic, it has been made to read headerless RAW files. The images must be 700x700 pixels, eight bits per pixel, TRADE-IN Offer Any Brand – Dead or Alive Up to 20% trade-In Value towards your new Tektronix Instrument Enabling Innovation Trade-in*1 your Analog or Digital Oscilloscopes, Function and/or Waveform Generators or Logic Analyzers for the advanced Tektronix range. Trade-In*1 Any brand instruments – savings of up to 20% towards your New Tektronix instruments. ACT NOW! Equivalent instruments will be traded-in towards any of the following Tektronix models:  Digital Phosphor & Sampling Oscilloscopes (DPO7000/DSA8200)  Mixed Signal & Digital Phosphor Oscilloscopes (MSO4000/DPO4000/TDS3000B)  Digital Storage Oscilloscopes (TDS2000B/1000B Series)  Arbitrary Waveform Generators (AWG5000)  Arbitrary Function Generators (AFG3000 series)  Logic Analysers (TLA5000B/TLA7000) Offer expires on the 30th September 2007 *1 Conditions apply - Trade-In value may vary depending on instrument type, model & condition of trade-in instrument. Please call 02-99113888 or enquiries<at>tekmarkgroup.com to inquire Authorised Distributor in Australia: Certificate No: AU07/3011 TekMark Australia Pty.Ltd Suite 302, 18-20 Orion Road, Lane Cove, NSW 2066 Email: enquiries<at>tekmarkgroup.com Website: www.tekmarkgroup.com Phone: 02-9911 3888 Fax: 02-9418 8485 September 2007  29 30  Silicon Chip 1k 10k IC1 PIC16F628A (FACE DOWN) 1 RED DOTS INDICATE VIAS OR SOLDER-THROUGH LINKS 1 IC2 STP16C596 16109071 INNER COLUMN OF LEDS MOUNTED ON TOP OF BOARD & SOLDERED UNDER OUTER COLUMN OF LEDS MOUNTED ON UNDER SIDE & SOLDERED ON TOP IC3 STP16C596 1 10k GN D +V GN D Fig.2: the PC board component overlay (shown from the component side) with matching top and bottom-side photographs of the PC board. The fingerprints are an optional extra! Seriously, the boards should be coated with a PC board (ie, solder-through) lacquer immediately they are made to prevent this from happening – especially as these boards will be out in the weather on the pushbike. In fact, we’d even go so far as to give the whole thing a good spray once finished – making sure you don’t get it in the two connectors. 10F 100nF HS1 1k 1 CLK DAT MCLR GND (BEND DOWN 90o ) 10k REG CON1 siliconchip.com.au FIRMWARE Ian Paterson’s firmware for this project – 628h.asm, povslope.bas & povimage.bas can be down-loaded from his website at www.ianpaterson.org/projects with the pixels being either pure black (0x00) or pure white (0xff). Such a file can be created with Photoshop or many other graphics programs. When you’ve finished creating the image, the final file size should be exactly 490,000 bytes. To stop the LEDs from lighting up when the bike is stationary, the last raster line is always set to zero (off). Because the firmware stops incrementing the raster line counter when it reaches the last line in the image, having all LEDs off in that line will cause them to remain that way until the next trigger pulse from the Hall Effect sensor. Construction After checking the PC board, start with the three 10kW and two 1kW resistors, followed by the 100nF and 10mF capacitors. Of these, only the 10mF capacitor is polarised. As this is a double-sided PC board, we should mention that the components mount on the side with the writing in the copper! Next solder in the three IC sockets (the right way around!) and two connectors, followed by the polarised regulator (REG1) and Hall Effect sensor. For me, the most troublesome part of this project is soldering the Hall Effect sensors without damaging them. Because they are sensitive to both mechanical and thermal stress, you must use great care when attaching them to the circuit board. Their leads must be bent down 90° towards the face which has a chamfered edge on its top. This means that the face will actually be towards the PC board surface when fitted. When bending the leads, you must hold the sensor lead with needle-nose pliers between the plastic case and the point at which the lead is being bent. This is to prevent mechanical stress at the point where the leads enter the sensor’s case. When soldering, you must also use needle-nose pliers as a heatsink to prevent damage from excessive heat. Once the sensors have been successfully soldered onto the board, there is little risk of further damage. Soldering the LEDs You have probably noticed that we have left the LEDs until last. That’s because there are a lot of them and they too can be a bit tricky to solder. There are 32 LEDs to be soldered to each side of the PC board. Note first of all which is the anode and cathode of the LED – there is a flat spot on the bottom of the LED next to the cathode (labelled “K” on the circuit diagram). Also, the anode lead is longer. At right are the same PC boards shown opposite, this time fixed to their backing “plate”, ready for mounting on the wheel. Note the semi-circle notches at the bottom end to fit into the axle. The top end is rounded to fit against the rim. siliconchip.com.au September 2007  31 Parts List* 3 PC boards, each 50 x 245mm, code 16109071 3 18-pin IC sockets 6 24-pin IC sockets 3 7.2V or 8.4V 700mAh (or higher) battery packs (do not use 8.4V with red LEDs) – see text 3 magnets – see text Material for backing plates – see text Semiconductors 3 PIC16F628A microcontroller programmed with 628h.hex (IC1) 6 STP16C596 LED driver (IC2, IC3) – see alternatives below 3 DN6851 Hall Effect sensors (HS1) – see alternatives below 3 LM2931AZ-5 low-dropout regulators (REG1) 192 high brightness LEDs (LEDs 1-64) Capacitors 3 10mF 16V electrolytic 3 100nF MKT polyester or monolithic (code 100n, 104 or 0.1mF) Resistors (5%, 0.25W) 9 10kW (brown black orange gold) 6 1kW (brown black red gold) Alernative Parts ST Microelectronics STP16C596 LED driver alternatives: Allegro A6276EA Maxim MAX6969ANG Maxim MAX6971ANG Panasonic DN6851 Hall effect sensor alternatives: Melexis US5881EUA Allegro UGN3113 (may be discontinued) Allegro A1101LUA-T Allegro A1103LUA-T * This list is for all three modules On the top (component side) of the PC board, the LEDs are arranged with their cathodes oriented towards CON1 (the 4-pin connector) while on the bottom side, the reverse is true. The LEDs are controlled in pairs, one for each side of the board, thus allowing the POV image to be viewed from either side of the bicycle. The LED pairs are connected in series by way of small jumper wires through the 32  Silicon Chip In daylight, you can see the arrangement of the PC boards and batteries inside the spokes of the wheel. The PC boards, mounted 120° apart around the wheel, fit against the axle and are secured at the rim end via a couple of cable ties onto the spokes. It’s important to keep the battery packs (which ever form you use) close to the axle to prevent the wheel getting out of balance. PC board that serve the same purpose as a PC board “via” – connecting the copper on both sides of the PC board together where required. The biggest challenge in soldering these jumpers is that the heat from your soldering iron will travel along the wire and melt the connection on the other side of the board. I found it helpful to use those “Helping Hands” soldering aids with alligator clips to hold the wire in place. If you are able to obtain or make PC boards with vias, then these jumpers are not necessary. Finally, plug the three ICs into their sockets. Be careful to line up the notch in the end of the IC with the notch in the end of the socket. A second check is a small paint dot or indent beside pin 1 – make sure this goes where pin 1 is shown on the component overlay. Loading an image Since this POV design stores the image in program memory space, the microcontroller must be re-programmed every time you want to load a new image. The process is as follows: • Create a 700x700 pixel, 8-bit per pixel image and save it with an 8-character filename. • Edit POVIMAGE.BAS so that it references the new image and run the program. It will save its output with a .ASM extension. • Copy and paste the .ASM output into the POV firmware file (628h. asm). • Compile it to produce a .HEX file and program the POV board via the 4-pin In-Circuit Serial Programming (ICSP) connector. This connector does not supply power to the board during programming, so you must supply power from a battery pack or an external supply. Testing Test the operation of the POV board before fixing it to the spokes. It’s a lot easier to fix mistakes on the bench than on the bike! Of course, the microcontroller should be programmed at this stage Apply power and wave a magnet in front of the Hall Effect sensor. You should see the LEDs illuminate. They won’t make much sense (ie, there will be no picture to see) but at least you will know the microcontroller is doing its thing. If they don’t light up, turn the magnet over and try again. The faster you wave the magnet in front of the sensor, the faster the LEDs should flash. If this test siliconchip.com.au fails to illuminate the LEDs, the most likely causes are a defective Hall Effect sensor or a bad program. The batteries The battery voltage needs to be high enough to allow the regulator to provide 5V for the microcontroller and also just high enough to allow the LED drivers to deliver up to 20mA through each LED pair. Try using a 7.2V battery pack for LEDs with a low forward voltage (such as red) and 8.4V for other colours (such as white and blue). Be sure not to use a battery voltage that’s more than about 2V higher than 2x the forward LED voltage, otherwise the LED drivers may be damaged. In the prototype, battery packs were made up from AA NiMH cells. I used 700mAh cells but with 2500mAh now available, 1000mAh and even 1500mAh are becoming quite cheap. The larger the capacity, the longer your display will last. You can use six cells (for 7.2V) or, as long as you don’t use red LEDs, seven cells (8.4V) in your battery packs – it’s more a case of getting a suitable holder. All three packs should be the same weight to avoid unbalancing the wheel. An alternative, albeit a bit heavier, is to buy 7.2V or 8.4V battery packs intended for radio controlled models. High power (3500mAh+ ) ones are expensive but you can often find lower capacity types on eBay for less than $20. Just make sure you mount them so they can’t fly off! Wheel mounting The accompanying photo shows the position of the PC boards on the bike wheel. It’s important to note that the inner edge of the PC board sits right up on the axle and the whole thing is centred between the spokes, so that the board is right in the centre of the wheel. To mount the PC boards in the wheel, I made a protective backing out of 3mm sintra (often used as a rigid backing onto which printed material can be mounted), covered one side with anti-static plastic (cut from a motherboard bag), and attached it to the solder side of the PC boards using plastic cable ties. I’m not sure if the anti-static plastic is of any benefit but I used it as a precaution in case a static charge builds up on the sintra as the wheel spins. siliconchip.com.au RIGOL SCOPES SALE ... SAVE UP TO $300 Rigol DS5062MA 60MHz Rigol DS5102MA 100MHz 60MHz Bandwidth 1GS/s Real Time Sampling 2 Channels Mono LCD Display 4K Memory Per Channel 20 Automatic Measurements Advanced Triggering on Edge, Video & Pulse Built-in FFT Built-in USB 3 Year Warranty 100MHz Bandwidth 1GS/s Real Time Sampling 2 Channels Mono LCD Display 4K Memory Per Channel 20 Automatic Measurements Advanced Triggering on Edge, Video & Pulse Built-in FFT Built-in USB 3 Year Warranty ONLY $ 799 ONLY $ ex GST 1,099 ex GST 300 200 SAVE $ SAVE $ * Offer valid until 30th June 2007 or while stocks last. Sydney Melbourne Tel 03 9889 0427 Tel 02 9519 3933 Fax 03 9889 0715 Fax 02 9550 1378 email testinst<at>emona.com.au Brisbane Adelaide Perth Tel 07 3275 2183 Fax 07 3275 2196 Tel 08 8260 8166 Fax 08 8260 8170 Tel 08 9361 4200 Fax 08 9361 4300 web www.emona.com.au EMONA On one end of the sintra, I cut a crescent-shaped notch that matched the radius of my front hub shaft. On the other end, I cut a notch for the spoke nipple. All that is needed to secure a PC board to the wheel is two cable ties at the spoke nipple end – the other end stays put because the crescent-shaped notch engages around the wheel hub. To keep the hub end of the boards in place, I used two short sections of plastic hose, slit down one side, wrapped around the hub shaft and attached with cable ties. These act as spacers that prevent the boards from sliding laterally along the length of the hub shaft. Note: these boards will fit a 26-inch or larger wheel. Also, when using three boards, it’s easier to mount them in a wheel with a number of spokes that’s divisible by three (eg, 36 spokes). Mounting the magnet To trigger the Hall Effect sensors, I used a stack of four magnets from an old 3.5-inch hard drive. The stack of magnets was placed on the inside of one of the bike forks immediately above the region under which the Hall Effect sensor passed, then secured in place with a piece of tape. Other suitable magnets would be one or more of the rare-earth or so-called “super magnets” which are enormously powerful for their size. More information? There are a lot more notes, flowcharts, firmware and graphics on the author’s website: www.ianpaterson. SC org/projects September 2007  33