Slashdot Mirror
Small Electronic Logic Blocks - eBlocks
eBlocks writes "eBlocks are small low-cost electronic devices that can be easily interconnected for a wide variety of applications such as: detecting motion, light, water, sound or magnetic fields; triggering a buzzer, a light, an electronic relay or a lock. Devices can communicate wirelessly or can be controlled remotely via the internet or a telephone. The eBlocks technology has been developed by a professor at U.C. Riverside who is looking for inspiration on its best uses. Try out the simulator. Suggestions and comments welcome!"
Happy Trails!
Erick
http://www.busyweather.com/
but the light sensor has a lifetime of 1 year... and a few of the other blocks are only good for 2-3 years.
not sure if that's good for the average person with no programming abilities (the type of people who won't read about stuff)
When I was growing up around 1968 - 1969 Raython had a series of kits that you could build circuits with out wires by touching blocks together and the ground was a metal plate. Connections were metal contact with magnets behind that.
The Learning Company has had this since the early 1980 in a game called Robot Odyssey. You could wire stuff up and solve puzzles. I rember since I was on the conversion crew from the Apple II/IBM PC to Color computer.
Its a nice idea but has been done for a long long time.
Boasting about how smart your kids are (or how smart you would like them to be) with a porn site in your subject field is just plain scarey.
This reminds me very much of programming a cheap (quasi-)PLC (programmable logic controller) we use at work often: a Siemens LOGO (pdf link). Basically, it's a device that has a bunch of inputs (8 digital, or 6 digital / 2 analog), and some outputs (4 digital), and contains a bunch of logic gates, comparators, and timers. You can make fairly sophisticated control systems using these.
We use them, for example, to control chemical injection systems. They have overrides based on filters backwashing, timers to dose to keep the pumps if they haven't run for a while, timers to prevent them running too long, etc. It's pretty endless what you can do, and these are only the lowest level of entry into the world of automation and PLCs.
Take the eBlock logic and timer modules, make them all software, and you have a LOGO. You still need the sensors and controls/outputs, but you can make some fairly complex programs involving hundreds of blocks, without the size of using hundreds of blocks.
The eblocks are a neat idea for educational purposes, but I'd see people quickly moving up to small PLCs (like the LOGO). They also definately don't have any use in industrial applications, though I don't know if that was the intent or not.
Speak before you think
The UC Riverside engineering college has an ABHORENT graduation rate, when I was there it was 30%. The program was very tough, but there were also alot of professors who cared so little about their courses they taught *nothing*. I had one chem professor lecture on the heart medicine he was working on and iron refining techniques all semester, then he gave us a standardized test and the whole class failed. We'd never seen a problem worked on the board the entire course. Most of the lower division courses were like that, professors didn't give a crap. My graduation was delayed by a mechanical engineering teacher who flunked 80% of his class.
Contrast this with a professor like Vahid-- the entire class flunked his first midterm and he stopped the course and said "This is awful guys, I have never seen anything like this, ever. Obviously I'm doing something wrong because all of you shouldn't be failing like this. So I want everyone to take 5 minutes, write down what you feel is wrong with the course, turn it in, and then go home and take the day off and we'll come back tomorrow and go over the notes and see what we can come up with." And he was as good a friend and father to tiny tim as he promised.
Religion is a gateway psychosis. -- Dave Foley
However, I have a problem seing how their concept is more advance than logiblocs.
The concept I'm working on will be based on open source hardware. any input on this ?
OK, I don't want to rain on his parade, it's a cool idea, and really good for people who don't want to work with proto-board or wire-wrap. I just see one BIG problem.....cost.
Having been into PIC's for a while now, I'm finding that cost is the final frontier when using/building small electronics. I can't imagine any of these building blocks being sold for less than $10-$15 each, and that's for the simple functions. By the time you get something really interesting going, it's gonna cost a LOT of money for all of the modules you'll need. MUCH more than it would to buy a PIC Demo board with programmer, LCD, and all the other features they throw into the box. From the links, it looks like he might have one of these processors in his boxen.
Here's a simple cost breakdown for one of the modules shown in the photo that I saw on the site:
1) Electronics: no less than $2 for any of the functions listed. PIC's are run from $2-18 in small qty, depending on features.
2) Wire: Sounds trivial, but it's gonna be $0.50 to attach two wires to a circuit board... That's a minimum for boxes that only have 2 wires, scale accordingly.
3) Packaging: small molded plastic box in qty $0.50. I know, I've been pricing them for my products.
4) Custom circuit board: minimum of $1.40 for small 1.5" X 2.5" 2-layer board.
5) Assembly: gonna cost $3 to stuff 20 small parts on that panelized board. No way around that, unless you've got a lot of spare time and are good with the iron.
6) Potting: gotta hold that stuff in the small enclosure. It's gonna be $0.50 here too.
7) Packaging/testing: it's gonna cost something to test and put that baby in a box. My estimate is at least $1 for each unit.
Really cool, but it looks really expensive.
For a cost comparison, you can purchase the PICDEM2+ board, with In-circuit debugger and development environment for $229.00 (digikey #DV164006-ND). You can also download the demo C compiler for free and start hacking immediatly.
With the features on that board, you can do 10-15 modules worth right off the bat....
AVR is a similar option, as is the 8051 and Z80. There's lots of small demo boards available.
If you like PLC type logic, try one of the cheaper units from Keyence or DirectLogic...these units are about $150 with all the features shown and more.... Both have high-speed counters, large memory and at least 8-in and 8-out....
Bottom line is this, don't be afraid of tinkering with parts...it's fun, cheap, and you'll learn a lot more by actually reading the datasheets and soldering wires yourself.
Happy tinkering
Remember those? Only instead of working with electrical energy, it was all about working with mechanical energy. Some of the capsules were differentials, some where gear reducers, and some interacted with the outside world, with lights or vacuums or propellors.
Other than the fact that they were bulky as hell, it was one of my favorite toys. Ah fond bathtub memories of running the vacuum in reverse and making a squirt powered pontoon boat.
- JoeShmoe
.
-- I wonder which will go down in history as the bigger failure: the War on Drugs or the War on Filesharing
I would have one battery block you could plug in anywhere in the setup, rather than have each block have its own battery. If you have 37 blocks hooked together, how do you know which one has the bad battery? On the downside it makes the connectors and wires bulkier by one power line, and requires n+1 blocks while before you had one. but it beats having to buy 37 batteries every year and take everything apart to replace them. Cheaper too because you can use smaller lighter, unopenable boxes.
(Assuming that you have a PIC or other smarts per block): You should have some way of plugging in a computer interface block and have the entire circuit topology displayed on your computer: each block can query all its neighbors to ask what they are and what state they are in, and what their neighbors are, etc.
Have user-programable blocks. Maybe a full PIC development C compiler etc. is required, or you can write your own little-language (or better yet, one that already exists) to lower the bar for beginning programmers. That way, when someone wants a block that shows red when the cat flap has had an exit more recently than an entrance, it can be programmed up, rather than requiring the development of a new cat-flap module.
[karmawhore]Everything should be open source, with a Linux development system, of course.[/karmawhore]
Servo motor blocks, motor blocks, etc. Maybe make a Mindstorms interface.
Ah, but what if you didn't need an engineer? What if municipalities or factories or whatever could get their janitors and repairers and other semi-skilled labourers to make these? Engineers should only be used for creating something that is unlike every else ever made*, the kinds of devices these might replace should currently be made by technologists, eBlock-type technology will simply bring the task down to the level of technicians or below.
* No two bridges are alike. The kinds of things which may at first appear to have an engineer because they're important rather than unique are actually just unique.
One inch by one inch plastic cubes had magnets to hold them to a metal backplane. On top was the symbol for the component they contained. Transistors, Photocells, variable resistors, etc. Connections out were to the sides of the cubes. Each conetion had a magnet, and through very well thought through design, the polarity of the magnets made right design easy.
All a young nerd had to do was rearrange the blocks on backplane and whatever circuit was on the top of the designs was functional.
It was like a real world, analog circuitry version of the simulaor. (The simulator by the way is very sweet.) Lego and this guy should get together and revive this concept. The IP is probably out there now.
Great work! Need help getting these kits to market?
"Knowing everything doesn't help..."
Go to basicx.com and get a simple to use and program basic stamp compatible chip. Better yet it is way faster than any basic stamp and holds more instructions. 50 dollar chip and some cheap leds and switches and
much more flexibility. His idea is ok but it is to costly compared to some of the other solutions. This is nothing more that the ibutton idea except ibutton only requires a single wire.
Got Code?
With the leJOS Java package (including java.net) an RCX brick, and some sensors (Logit sensors connect to the RCX) you could match the functionality of eBlocks. However, I suppose the implementation of eBlocks is simpler for non-programers/non-techies. www.lejos.org
If you want to learn electronics (aside from a university), get a book or two, and read them. Then get some similation software and learn how circuits behave. Then get some premade eval boards of whatever you want (MCU, CPLD, FPGA etc.) and play with them. If you are still interested, then you probably want to work as an engineer, since only then you will have access to all the expensive (and neat) toys and gizmos.
But if you don't plan to become an engineer, but only want to make some IR-operated control for your house, don't try to use those "blocks", or bredboards, or wire wrap - that is nothing but waste of time. Begin with a design on paper. Then simulate it if you can (you can if you have a computer.) Then either etch the PCB, or order one (tools are free, job about $100) and assemble your new toy yourself.
If you can get away with using standard eval boards for your purpose, definitely do so. Fact is, more and more "hardware" is now implemented with DSP technologies and highly integrated, specialized ICs. You don't want, in fact, study how to make a decent RF amplifier - you buy one from Minicircuits for less than a dollar. You don't want to make a radio with 20 transistors - instead you use one or two chips (Analog Devices). Instead of going to the basics you can embrace the modern technology, it is much easier to work with, and many tools for the beginners can be had for free (see Xilinx, for example), and the visibility into your circuit is much, much better [unless you have a million dollar lab.]
The Smart-Its Project has been doing something similar since around 2000.
http://www.smart-its.org/To err is human, but to forgive is beyond the scope of the Operating System...