View allAll Photos Tagged microcontroller
Using Digispark ATTINY85 USB board, programmed Zoom Mute and Video toggle keystrokes to the buttons (guitar pedal switches).
The results of a DIY project writing custom code for a Node MCU micro-controller to control addressable LEDs.
Developed at SRI International, the company has spun out and closed their first round of funding, led by Future Ventures. The company can compress common AI models by 10x without a noticeable change in accuracy, enabling deployment on the inexpensive microcontrollers, DSPs and other processor cores typically found in edge devices. This allow intelligence to migrate to the edge for local processing (e.g., face-detection algorithms running locally within security cameras or appliances, or Siri-like voice interfaces working instantly even when network connectivity is missing).
Couple some local intelligence to each sensor and the internet of things is becoming the sensory cortex of the planet, with countless data-collecting-devices. All of this 'big data' would be a big headache but for machine learning to find patterns to make it actionable, and edge computing to shift the processing to the periphery and avoid network overload. In short, the edge needs AI, and AI needs the edge. Latent AI integrates both with a portfolio of IoT edge compute optimizers and accelerators that bring an order of magnitude improvement to existing infrastructure. This is essential, as the majority of new software today is trained as a neural net, and most compute cycles will shift to the edge.
From the NVIDIA CEO: “We’ll soon see the power of computing increase way more than any other period. There will be trillions of products with tiny neural networks inside.”
The core technologies of Latent AI come from SRI International, where I have been an advisory board member for over a decade and seen technologies like Siri develop and then spin out of SRI.
Here is today’s announcement from the company.
And the company site: LatentAI.com
I took this photo of CEO and co-founder Jags Kandasamy presenting Latent AI to the IoT Consortium.
Visitors to our family blog can launch a car from our closet for our toddler to play with. I built a simple gravity-powered car launcher that is controlled by the web site. Clicking "Launch Car Now!" on the web site results in a sound clip from the movie "Cars" playing in our living room, followed by a car shooting out from under the coat closet door. The website uses a bit of PHP to send an email to my wife's computer, which happens to sit in the living room next to the coat closet. I created a filter in Apple Mail to run an AppleScript when a correctly coded email comes through. The AppleScript pauses iTunes and raises the system volume of the computer before activating a small applet I wrote in the Processing language. The Processing applet plays a bit of Lightning McQueen (main character in "Cars") psyching himself up before a big race. The applet then sends an "l" to the serial port, where the car launcher's Basic Stamp II microcontroller is patiently waiting. The BSII opens the sliding garage door on the launcher exactly one bay. There are five bays, for five cars. I set up little tabs to interrupt an infrared beam as the edge of each opening is reached. Once a given bay is open, gravity pulls the toy car out and down the ramp. Momentum carries it under the closed closet door and across the floor to the excited toddler. After receiving each launch command (each clicked "Launch Car Now! from the web site), the launcher will release one car and then wait for another command, progressing until the door is completely open and all cars have been released. The launcher door will then close and wait to be reloaded. The PHP on the web server makes sure the "Launch Car Now!" link is only available during usual playtime hours and also limits the number of cars launched to five per day. The table on which the launcher is sitting was another of my weekend projects, a nice roomy table for the little guy's wooden train set.
Simple circuit to connect a 5V relay to one of the digital output pins of the Arduino. I drew this after a few threads showed up on the Arduino forums, asking about relays. I hope this drawing make things a little clearer for the novices. Best viewed in original size ("Actions>View all sizes" button at lower left).
I've shown a few options for the transistor and diode, but the choice is not critical and almost any NPN bipolar transistor will do. For larger loads (such as big solenoids or motors), you'll need a power transistor that can handle the larger current, and you'll need to reduce R1 to maybe as low as 220 Ohms.
When wiring up, be sure to check the data sheet for the transistor to find the pin connections (e, b, c). The diode will be marked with a band or stripe at the cathode (k) end.
The Arduino microcontroller board wired up to the SID chip out of a Commodore 64. This circuit really uses up too many of the Arduino's I/O pins, and I think a better way would be to use a couple of 74HC595 shift registers and then generate the CS pulse in hardware.
XGS AVR 8-Bit demoing high color "Apple ][" graphics mode. This mode is low resolution, high color just like the Apple ][.
See product page at:
Vintage Motorola MC68705 8-bit microcontroller.
Confusing the on-chip A/D converter with a flashlight.
This light sensitivity is not intended. Many ICs would react to light
and thus they are encapsulated in black plastic.
The uC even gets stuck by erratic currents caused by the flashlight
and needs a reset.
It is vintage NMOS technology, it draws 100mA
supply current for doing nothing! (power-saving CMOS
technology already existed these times,
but it could not reach the MHz speed range yet).
I bought Canon's remote switch (RS-60E3) last year, but it's lost. For shooting July 4th fireworks, I need a remote release switch. Instead of buying the same stuff again, I decided to build one by myself.
The design is simple for now. It has just 2 buttons, 1 toggle switch, and a 2.5mm jack.
I left some extra space in the case. This is for microcontrollers that I'm planning to install for time-lapse shooting. Right now, I just don't have enough time to do this :-(
Computer controlled shutter for the Automatic 100 series packfilm cameras with manual exposure control. See www.chemie.unibas.ch/~holder/shutterpic/index.html
Continued from the other day, some of the initial circuit has changed, and now includes a built-in sound trigger. Except that the sound triggering is giving me fits... that part of the circuit is analog, and the logic, of course, is digital. Apparently, the microcontroller and I have differing views of what "rising edge" means, and the trigger causes the flash interrupt well before it's supposed to. So far, I've not been able to work it out.
I'm sure I will, though. The problem is probably in my code, with a second choice being the analog interface into the digital logic section. Anyway, this is the most complicated circuit/microcontroller code project I've ever worked on, and I've had surprisingly few issues with it, so far...
Anyway, since I don't have the photos I was building this to help capture, here's a description of what it does, when I think about it, blissfully absent from the hard reality... notes above describe the various parts.
Prefocus and compose the shot on the unsuspecting firework. Turn on the device, wait for the "ready" indicator light. Push the little yellow button. The shutter interface opens the shutter in bulb mode. Half a second later, the igniter circuit activates, cranking 4 amps through the rocket motor igniter. Since 2 amps should "guarantee" ignition, that should be good. Once the igniter circuit goes hot, the sound trigger starts listening for the boom. When it hears the boom, it signals the microcontroller to cut power to the igniter, pop the flash, and close the shutter on the camera. Then it recycles, waiting for the next one. Except, of course, it doesn't... it doesn't do that, at all.
Unfortunately, I've already begun to get distracted... it occurred to me today that I could probably make a cheap-o 16-24 channel logic analyzer with an Atmega16. That's very appealing to me, because sometimes an analog scope isn't enough. It's painful to work out chip-to-chip communications with a 2-channel scope. I may switch over and work on that for the rest of the night. Of course, I'll have to implement a USB interface, and write a Windows app to actually use it, so there's like a zero percent chance I'll have it done this weekend. Which means that by next weekend, I'll be on to something else...
update!!! Live fire trial run tomorrow!!! Fixed both known issues (both were code - one was a misread register name, the other was a lack of understanding of how code optimization can jeopardize variables)!!!
An Atmel microcontroller, the ATmega8, on a little protoyping module, plugged into a solderless breadboard. This is the controller that I've been using for a number of small projects, including a persistence-of-vision wand and an LED dot-matrix display. The CPU runs at 16MHz and is programmed in either assembler or C.
I built a PC Meter which display CPU and Memory usage for an attached computer. It's driven by an Arduino microcontroller, which is fed the stats by a C#.Net application I developed. More information, source code, and a video demonstration can be found on the project web page here:
This is a microcontroller-powered temperature controller for a fridge I built out of fabricated necessity and spare parts last weekend.
Backstory: We replaced our ancient 2nd fridge with a big new upright freezer, which left our little chest freezer empty. Sometimes we do need the extra fridge space though during harvest season at the farm, so I figured I could just turn the temperature up on that old chest freezer and turn it into a fridge. Nope. Highest it goes is -4C. OK. Time for some gratuitous technology. See the notes.
Arduino UNO does PWM (pulse width modulation) – 16mm scale.
This test board has a UNO configured to provide 6 PWM servo outputs as a microcontroller for points (turnouts) and/or signals on my 16mm scale narrow gauge exhibit. Two vero strip boards incorporate the required crossovers to provide Futaba/Hitec/etc format servo pinouts on the UNO’s digital output sockets.
Each of the UNO’s analogue inputs is held at intermediate potential by a pair of 1 Kohm resistors. The remote commander has a 20 foot wander lead and its four momentary push buttons short out one or other 1 Kohm resistor to take the analogue input to LOW or HIGH potential as a tri-state switch. The UNO picks up the change in potential and moves the corresponding servo through the number of degrees pre-programmed in the sketch, via the selected PWM output.
This basic setup will enable the option of automated signal and point control on the 16mm scale exhibition layout by linking the UNO's analogue inputs to other sensors.
Arduino UNO does PWM (pulse width modulation) – 16mm scale.
This test board has a UNO configured to provide 6 PWM servo outputs as a microcontroller for points (turnouts) and/or signals on my 16mm scale narrow gauge exhibit. Two vero strip boards incorporate the required crossovers to provide Futaba/Hitec/etc format servo pinouts on the UNO’s digital output sockets.
Each of the UNO’s analogue inputs is held at intermediate potential by a pair of 1 Kohm resistors. The remote commander has a 20 foot wander lead and its four momentary push buttons short out one or other 1 Kohm resistor to take the analogue input to LOW or HIGH potential as a tri-state switch. The UNO picks up the change in potential and moves the corresponding servo through the number of degrees pre-programmed in the sketch, via the selected PWM output.
This basic setup will enable the option of automated signal and point control on the 16mm scale exhibition layout by linking the UNO's analogue inputs to other sensors.
2 drop collision into Xanthan gum mix, blue ink and a little Dettol kitchen cleaner.
The blue/green in the rise is from gels on the 2 flashes behind the glass.
100% uncropped and unedited.
2:1 using Canon 7D and 100mm f/2.8L macro with 68mm of extension tubes (full kenko set).
Computer controlled shutter for the Automatic 100 series packfilm cameras with manual exposure control. See www.chemie.unibas.ch/~holder/shutterpic/index.html
I built a PC Meter which display CPU and Memory usage for an attached computer. It's driven by an Arduino microcontroller, which is fed the stats by a C#.Net application I developed. More information, source code, and a video demonstration can be found on the project web page here:
“Claremont Road” has five Arduino UNO microcontrollers which control train movements, along with PWM (servo adapted) points/turnouts, and signals according to pre-written programs or “sketches”. This is a completely different concept from DCC.
The master co-ordinating UNO gets feedback from the track through 14 enbedded infra-red proximity detectors,
Slaves 1-3 are UNO “train drivers”,
Slave 4 handles the display and lights. The orange display shows the current mode and commands being passed between the UNOs via a short-wire protocol known as I2C.
Using a pair of XBees to send packets back and forth between my laptop and an ARM microcontroller while a logic analyzer keeps tabs on the process.
Photo taken to accompany short article on working with AVR microcontrollers, and making minimalist target boards for programming them.
This is a binary clock that was built into a 3d-printed case created in Minecraft. It shows the current time in a binary coded decimal format.
The model was exported with the free tool Mineways and printed on a Zprinter 650 3d-printer, with a block size of 125mm^3 (so every block has an edge length of 5mm). After printing, LEDs were glued into the case after filing the openings a bit wider. Then, the LEDs were soldered to form a 4x4 LED matrix, and the matrix was connected to an Arduino board.
A technical description of the setup as well as downloads of the model and the code can be found here: postapocalypticresearchinstitute.wordpress.com/2012/07/18...
In the past the setup of this laser signal was always a pain. This adjustement need to be very exactly to have the max reflected signal on the microcontroller. Now the new setup make this adjustements simply. I can finetune the laser into all direction, first very coarse and then very precise into the X and Y direction. As you can see there are a lot of small parts. But this new setup give me the assurance to have always the max detection signal. This cost me 5 days of heavy works to make all this very small parts! once for the left side and once for the right side. The IR laser works at 840 nm and is pulsed 20us on and 20us off. The difference between the ambient light and the IR laserlight is processed into the AVR controller ATtiny45 into a separeted macro lens the AF60/2.8Dmicro from Nikkor. The information go to the central CPLD controller to make the decision to take a photo in focus.
I've assembled a few of the new boards so far - it's been a while, I'm a little out of practice, but these are looking good after some minor rework.
The other boards are waiting for now, I didn't notice that I was missing a part for the main quad pcb (the expensive gyro, I apparently only ordered a small qty and used them all in other boards)