There is no mechanical switch for displaying the serial number. You would have to flash a pattern on the photocell (or use your finger over the photocell as you said) to put the device in a mode to blink out the serial number pattern on the LEDs.

Yes, lesson learned. I'm going to continue the design of this device in public with peer input and then post the instructions for assembly. Please feel free to comment on any of the following design thoughts. As you can see, there are currently two LEDs, one Cds Photocell, one CR2032 battery holder, and a Atmel ATtiny11 microcontroller. Parts cost (including assembly) for building a panel of 100 devices (snap the boards apart when done) is about $3 each including the batteries but not the magnets. In the current design, all of the parts except for the LEDs and the Cds photocell are surface mount. I did this for two reasons. One, is for size, the other is for in-expensive automated assembly. I've left much room on the rear of the PCB to fluch mount a magnet as large as the CR2032 coin cell. The LEDs are thru-hole because I think they look better and are often brighter. With one CR2032 battery and series resistors, this design will not support white or blue LEDs because their forward voltage drop is too high. This isn't a problem when connecting the LED directly to the batter is in the original LEDThrowies, but this with this design, I am trying to promote retrieval and reuse. You might think that the LED could be connected directly tothe ouput port pin of the ATtiny11 microcontroller without a series resistor, but either the current would be over the limit of the output pin or there would not be enough forward voltage to drive the LED. When the microntroller output is LOW, the port output voltage drops down to about 0.6V, not 0V. The maximum voltage from the output pin to GND with a 3.3 V supply would be 2.7 V. This is still not enough to switch on a blue or white LED. Red, yellow, amber, orange, and green LEDs will work fine. I thought of adding a second battery but this adds cost and weight. If I did add a second battery I would have the microntroller powered off of one cell and the LEDs powered off of two cells so as not to exceed the 5V supply limit of the ATtiny11. I wanted to provide a way for the user to customize the setup of this device. Originally, I had a small surface mount switch (25 cents) in the design. I changed the design to use a Cds photocell (about 39 cents) for user input for two reasons. One, having a light detector would make for some interesting modes of operation and two, it enables the devices to be set up in bulk. Instead of using a switch to program the current time and other operational parameters of one unit at a time, insert the batteries on multiple units and switch the room lights on and off to setup units in bulk. The Cds photo cell forms a voltage divider with R4. This voltage is fed into the port PB1 of the ATtiny11 which is also the analog comparator input. The other side of the comparator is connected to an internal reference voltage. When the photocell is active, (PB2 HIGH) and the comparator is enabled, the microcontroller is able to detect a light or dark condition. This is how blinking the room lights ON/OFF can then used as a form of user input to the microcontroller. I'm bouncing the subject around here, but I'm feeling quite free after switching back to a QWERTY keyboard for the moment. My posts from the previous few days were brief because I had switched to the www.newstandardkeyboard.com and typing is still very painfully S L O W. I'll switch back to my study of this new keyboard after I finish posting this message. You have to throw the brain a new bone now and then. Let's talk about the choice of microntroller. I've used Microchip PICs in the past, but recently, I've been using the Atmel AVR microcontrollers. Either would work for a device like this. Feel free to design one with any microcontroller you find handy and that you have the tools for. Speaking of tools, if you are interested int he Atmel AVR, sign up at the www.AVRFreaks.net site for much information and several great support forums. To see what you can do with a small microcontroller, visit www.powerseed.com I am a co-inventor of this device. Why the ATtiny11? I selected this device solely due to the low price of $0.37 in hundred piece quantity. Less expensive than a good LED. Here's a list of features: • Utilizes the AVR® RISC ArchitectureUtilizes the AVR® RISC Architecture • High-performance and Low-power 8-bit RISC Architecture – 90 Powerful Instructions – Most Single Clock Cycle Execution – 32 x 8 General Purpose Working Registers – Up to 8 MIPS Throughput at 8 MHz – 90 Powerful Instructions – Most Single Clock Cycle Execution – 32 x 8 General Purpose Working Registers – Up to 8 MIPS Throughput at 8 MHz • Nonvolatile Program and Data Memory – 1K Byte of Flash Program Memory In-System Programmable (ATtiny12) Endurance: 1,000 Write/Erase Cycles (ATtiny11/12) – 64 Bytes of In-System Programmable EEPROM Data Memory for ATtiny12 Endurance: 100,000 Write/Erase Cycles – Programming Lock for Flash Program and EEPROM Data SecurityNonvolatile Program and Data Memory – 1K Byte of Flash Program Memory In-System Programmable (ATtiny12) Endurance: 1,000 Write/Erase Cycles (ATtiny11/12) – 64 Bytes of In-System Programmable EEPROM Data Memory for ATtiny12 Endurance: 100,000 Write/Erase Cycles – Programming Lock for Flash Program and EEPROM Data Security • Peripheral Features – Interrupt and Wake-up on Pin Change – One 8-bit Timer/Counter with Separate Prescaler – On-chip Analog Comparator – Programmable Watchdog Timer with On-chip OscillatorPeripheral Features – Interrupt and Wake-up on Pin Change – One 8-bit Timer/Counter with Separate Prescaler – On-chip Analog Comparator – Programmable Watchdog Timer with On-chip Oscillator • Special Microcontroller Features – Low-power Idle and Power-down Modes – External and Internal Interrupt Sources – In-System Programmable via SPI Port (ATtiny12) – Enhanced Power-on Reset Circuit (ATtiny12) – Internal Calibrated RC Oscillator (ATtiny12)Special Microcontroller Features – Low-power Idle and Power-down Modes – External and Internal Interrupt Sources – In-System Programmable via SPI Port (ATtiny12) – Enhanced Power-on Reset Circuit (ATtiny12) – Internal Calibrated RC Oscillator (ATtiny12) • Specification – Low-power, High-speed CMOS Process Technology – Fully Static OperationSpecification – Low-power, High-speed CMOS Process Technology – Fully Static Operation • Power Consumption at 4 MHz, 3V, 25°C – Active: 2.2 mA – Idle Mode: 0.5 mA – Power-down Mode: <1 μAPower Consumption at 4 MHz, 3V, 25°C – Active: 2.2 mA – Idle Mode: 0.5 mA – Power-down Mode: <1 μA • Packages – 8-pin PDIP and SOICPackages – 8-pin PDIP and SOIC • Operating Voltages – 1.8 - 5.5V for ATtiny12V-1 – 2.7 - 5.5V for ATtiny11L-2 and ATtiny12L-4 – 4.0 - 5.5V for ATtiny11-6 and ATtiny12-8Operating Voltages – 1.8 - 5.5V for ATtiny12V-1 – 2.7 - 5.5V for ATtiny11L-2 and ATtiny12L-4 – 4.0 - 5.5V for ATtiny11-6 and ATtiny12-8 • Speed Grades – 0 - 1.2 MHz (ATtiny12V-1) – 0 - 2 MHz (ATtiny11L-2) – 0 - 4 MHz (ATtiny12L-4) – 0 - 6 MHz (ATtiny11-6) – 0 - 8 MHz (ATtiny12-8) Wow! All of this for only 37 cents. There is also a ATtiny12 which has some additional features that make it difficult not to choose this part instead of the ATtiny11. The drawback of the ATtiny12 is that it is almost $1 each in the same quantities as the ATtiny11. Since the devices are pin compatible, we could switch to this part in the future if there is a large enough demand for the additional features. What interests me the most about the ATtiny12 is that it is a self programmable part. If I were to use this part, I would develop a small bootloader that would receive program words (for the program FLASH) from the Cds photocell. The end-user could then update the application firmware by flashing a light on/off at the photocell. This would most likely be carried out by purchasing a single unit with the new firmware already installed and then putting it into a mode where it would flash the contents of its program memory out on its LEDs. The new program (firmware) would then propagate from one device to another. A nice feature, but is it worth an additional $1? Another way to update the program in FLASH would be to hang an LED off of a couple of pins on a serial or parallel port of a PC/Mac and write a trivial piece of code to BLINK out the new firmware to the watching LEDThrow 10010. This would lead to much more end user creativity. Want a new feature? Download the free WinAVR, GNUCC, and Atmel AVRStudio packages, program the new features in 'C' and blink out the HEX file to an LED to program your device. I'm looking for input here. Should we use the ATtiny at $0.37 or the ATtiny12 at $1.20. Would you pay an extra $1 for this capability? We don't want to get too crazy here, this is just a simple circuit with two LEDs and a photocell. And, and 8 MIPs processor with 1K byte of program memory and 64 bytes of EEPROM. So, what do we do with all of this processing power? Here's a list of initial firmware features. -End user setup of current time in hours and minutes. The device knows what time it is until the battery runs down or is removed. -End user programmable ON/OFF times. Program the ON hours and minutes and the OFF hours and minutes. The LEDs will be activated at the ON time and will be switched off at the OFF time. The processor enters low-power mode at the OFF time. -Multiple two-LED patterns would be pre-programmed and selected by the end-user. One pattern could have the LEDS start OFF and then brighten to FULL ON over seconds or minutes and then DIM to full off again and repeat. Another pattern could alternately blink one LED and then the other. Bother LEDs could blink or be on solid. Another mode would be to have the LEDs switch on at Dusk or some period of time after dusk and switch off at dawn or some period of time after dusk. Another mode would be to have the LED pattern change based on the changes in ambient light level (sensed only while LEDs are off for say, 50 mS, depending on the reaction time of the photocells.) -Each LEDThrow 10010 will have its own unique serial number. They'll be a mode where the end user can force the unti to blink out its serial number. Look this up on the web site and update the log for the device. See where it's been, record where it's going next. There'll be a mode where the end user can "describe" the blink pattern for each of the two LEDs by bllinking the room lights on and off. The device will record the blink pattern and then repeat it during the ON time of the device. Much, much more is possible. What would you do? Enough for tonight. Suggestions? Comments? Thanks, -Zoom

(Continued!) - Power-down Mode: <1 μAPower Consumption at 4 MHz, 3V, 25°C - Active: 2.2 mA - Idle Mode: 0.5 mA - Power-down Mode: <1 μA > Packages – 8-pin PDIP and SOICPackages – 8-pin PDIP and SOIC • Operating Voltages – 1.8 - 5.5V for ATtiny12V-1 – 2.7 - 5.5V for ATtiny11L-2 and ATtiny12L-4 – 4.0 - 5.5V for ATtiny11-6 and ATtiny12-8Operating Voltages – 1.8 - 5.5V for ATtiny12V-1 – 2.7 - 5.5V for ATtiny11L-2 and ATtiny12L-4 – 4.0 - 5.5V for ATtiny11-6 and ATtiny12-8 • Speed Grades – 0 - 1.2 MHz (ATtiny12V-1) – 0 - 2 MHz (ATtiny11L-2) – 0 - 4 MHz (ATtiny12L-4) – 0 - 6 MHz (ATtiny11-6) – 0 - 8 MHz (ATtiny12-8) Wow! All of this for only 37 cents. There is also a ATtiny12 which has some additional features that make it difficult not to choose this part instead of the ATtiny11. The drawback of the ATtiny12 is that it is almost $1 each in the same quantities as the ATtiny11. Since the devices are pin compatible, we could switch to this part in the future if there is a large enough demand for the additional features. What interests me the most about the ATtiny12 is that it is a self programmable part. If I were to use this part, I would develop a small bootloader that would receive program words (for the program FLASH) from the Cds photocell. The end-user could then update the application firmware by flashing a light on/off at the photocell. This would most likely be carried out by purchasing a single unit with the new firmware already installed and then putting it into a mode where it would flash the contents of its program memory out on its LEDs. The new program (firmware) would then propagate from one device to another. A nice feature, but is it worth an additional $1? Another way to update the program in FLASH would be to hang an LED off of a couple of pins on a serial or parallel port of a PC/Mac and write a trivial piece of code to BLINK out the new firmware to the watching LEDThrow 10010. This would lead to much more end user creativity. Want a new feature? Download the free WinAVR, GNUCC, and Atmel AVRStudio packages, program the new features in 'C' and blink out the HEX file to an LED to program your device. I'm looking for input here. Should we use the ATtiny at $0.37 or the ATtiny12 at $1.20. Would you pay an extra $1 for this capability? We don't want to get too crazy here, this is just a simple circuit with two LEDs and a photocell. And, and 8 MIPs processor with 1K byte of program memory and 64 bytes of EEPROM. So, what do we do with all of this processing power? Here's a list of initial firmware features. -End user setup of current time in hours and minutes. The device knows what time it is until the battery runs down or is removed. -End user programmable ON/OFF times. Program the ON hours and minutes and the OFF hours and minutes. The LEDs will be activated at the ON time and will be switched off at the OFF time. The processor enters low-power mode at the OFF time. -Multiple two-LED patterns would be pre-programmed and selected by the end-user. One pattern could have the LEDS start OFF and then brighten to FULL ON over seconds or minutes and then DIM to full off again and repeat. Another pattern could alternately blink one LED and then the other. Bother LEDs could blink or be on solid. Another mode would be to have the LEDs switch on at Dusk or some period of time after dusk and switch off at dawn or some period of time after dusk. Another mode would be to have the LED pattern change based on the changes in ambient light level (sensed only while LEDs are off for say, 50 mS, depending on the reaction time of the photocells.) -Each LEDThrow 10010 will have its own unique serial number. They'll be a mode where the end user can force the unti to blink out its serial number. Look this up on the web site and update the log for the device. See where it's been, record where it's going next. There'll be a mode where the end user can "describe" the blink pattern for each of the two LEDs by bllinking the room lights on and off. The device will record the blink pattern and then repeat it during the ON time of the device. Much, much more is possible. What would you do? Enough for tonight. Suggestions? Comments? Thanks, -Zoom

My hope would be that someone would take it, go to the web site shown on the back for instructions and then re-use it.

I agree with you on the environmental issues. This is why this design is hopefully more than something you would throw away. Each unit will have its own serial number that will blink out on the LEDs on request. I don't know much about web design btu it would be great to have a way to associate the serial number with location notes. Do you have any suggestions on how to do this?