Because in my experience, the yield from the chinese Model B factory is 50%.
My first RPi is currently tied up in a work project, so I ordered another model B from Newark. It came in and I fired it up yesterday, no LEDs or any signs of life. Dead.
Then I noticed the main BGA in the center of the card looked a bit askew, looked closer and noticed the BCM2835 was missing. The Samsung DRAM that ordinary sits on top of the '2835 was soldered straight onto the PCB. I understand the part shooter fucking up once in a while and missing a chip, but the board shouldn't have made it out of the factory.
C'mon. I'd rather pay a few extra bucks for something that's most likely going to work, than do what I'm doing now and spending even more bucks mailing the fucking thing back, and crossing my fingers that the replacement works too...
One of my friends did exactly what you're describing - he works on an offshore oil rig, and used a high gain antenna pointed at land to get cellphone service. It worked great... until he came ashore and his phone didn't get a signal.
Turns out the RF connector in the phone is only there for production testing of power/RF compliance - it's only intended to be connected to once, and it's not designed for daily connections/disconnections. The datasheet for one replacement connector we found was only rated for 10 cycles.
Other than sharing the same 2.4GHz ISM frequency band, bluetooth (used for their controllers) and WiFi share basically nothing. In the regular Wii they use separate chipsets, separate antennas, etc. They're stripping out a few bucks worth of hardware. Not to mention, implementing WiFi in a product invariably involves paying a bunch of license fees to patent holders etc - either included in the price of your wifi chipset or paid separately. This adds a couple more bucks to the design.
Take the $99 they're selling the thing for at the store and subtract store markup, shipping, factory tooling, packaging, the rest of the stuff in the Wii box, etc. and you don't have much money to spend on the Wii Mini itself. A few bucks spent adding WiFi could end up being a significant part of the cost.
Honestly, I have to throw some praise at Nintendo for making a game system that's so cheap. My workplace is doing an "adopt-a-family" thing for the christmas season, where employees get together and buy christmas gifts for single parents who can't afford much for their kids. At $99, it's made our shopping list.
Actual use cases I've seen for the Raspberry Pi that I've done myself and seen others do:
- Plug the Rpi into a LAN, and connect it to the serial console of a piece of equipment with a USB to serial cable - old router, telephone equipment, radio broadcast transmitter, fill in the blank. SSH into the thing if you need to get at the console instead of doing a site trip. - Plug a few sensors into it, run it off a 12V car battery and a +5V automotive USB adapter, and leave it somewhere to log data onto the SD card or a USB stick. - Plug a USB hard drive into it, and use it as a low power torrent downloader, instead of keeping your desktop PC powered up when you're not home.
It's a tiny, $25 linux machine. Possibilities are endless.
And in other news, Arduino cards have a 16MHz 8-bit processor with mere kilobytes of both RAM and flash. And despite making a 1980s suitcase computer look fast, they've proven themselves to be fully capable of running all sorts of awesome things that hobbyists have been using them for.
I worked on a system similar to this for a man with an prosthetic arm. We stuck a temperature sensor in the finger of the prosthetic hand, and used a small thermoelectric plate which contacted his skin where the prosthetic attached to him. The back side of the peltier plate was glued to an aluminum bar acting as a thermal source/sink, and the front side had a small stainless plate acting as the contact point with a second temperature sensor bonded to it for feedback. Using a microcontroller and a bit of simple hardware, we made the peltier plate temperature equal to the temperature detected at the finger. With some limits, of course.
The man went from having a plastic arm to having something that *felt* like an actual part of his body again. He described going home and touching his wife's face, and almost breaking down crying - it felt pretty good to hear that, especially since it only took us a few days to design/build the thing.
Anyway, it ran off three NiCd AA batteries and had a battery life of about 24 hours - he'd plug it in to charge when he went to bed, and it would easily last until the next charge cycle. These days thermoelectric devices are more efficient, batteries are a lot better and microcontrollers have much better power consumption.
There's a few things that make a system like this fairly low power:
- You don't need a big contact area to get the sensation of temperature across to the user - you don't need to heat or cool their whole hand. Half a square centimeter is plenty if you put it at a fairly temperature-sensitive part of the hand, such as where your fingers meet your palm. It'll feel weird at first but the 'immersion' feeling will eventually set in. - There's only a narrow range of temperature that you have to drive the contact plate to. You don't need to do 0 to 50 degrees C, and I fully expect Sony to restrict the range to +-10C at most for liability reasons, not to mention practical reasons. - Skin isn't *that* thermally conductive.
So if your contact plate is sized small and only within a few degrees of body temperature, you're probably only moving half a watt at most between the hand from the contact plate.
Secondly, I seriously doubt the thermal "immersion" effect will be running all the time, probably only acting on 'events' the game - walking indoors/outdoors (pulse of hot/cold), picking up a weapon from the ground (cold), getting hit with a fireball (hot), falling into water (cold), etc. Much like vibration motors in controllers don't run all the time.
End result is that running the thermoelectrics won't take that much power, and sinking/sourcing heat from within the controller shouldn't be that hard. Overall, seems pretty practical to me.
Something to consider - I've replaced the MicroUSB connector in my cellphone *twice*. The phone would work for about a year, then it would go flaky - you'd have to wiggle the connector a few times for the phone to reliably charge, and sometimes I'd go check on it and it wouldn't be charging - and it would happen with different cables. Supposedly these things are rated for 10,000 cycles, but I haven't seen it. Maybe my phone does something it shouldn't, like spark the +5V pin when the connector is plugged in. *shrug*
Secondly, I've caught the cord of my phone multiple times and pulled it off the desk onto the floor - and my cats/dogs have probably done it more times than I have.
Though there's an efficiency loss in wireless charging versus conductive charging, I wonder if there's an efficiency gain that exists in less phones being repaired/replaced because of damage related to conductive charging.
(Note that this is not a well thought out, researched argument - just a dumb thought.)
Do this. Talk to the math teachers at your school, find out if they've got any poor students that need them. And find out there's any other schools in the area that would have a use for them.
There's lots of single parents and otherwise poor families that can barely scrape together school supplies for their kids, let alone buy the graphic calculator that they would need to get into a precalc or AP math. Something simple like one of these old calculators could turn a kid's life around. Seriously.
I'm suspecting that the BMW is "password protected" in a way (unique knowledge required to get into the car's ECU - PIDs to hit, passphrases, etc) but if the car thief that knowledge then it's essentially open anyway.
Pretty much the only way you'll make a perfectly secure system is by putting the equivalent of a TPM chip in the car and using a cryptographic channel between the car and BMW themselves. Do authentication at the garage and employee levels, and have BMW log everything that happens to every car it talks to, so any malicious use gets recorded and you can figure out who enabled your car to get stolen. Do code signing in the ECU so they can't reprogram the ECU to run different code, authenticate the ECU against other modules in the car so they'd have to basically swap every module in the car... and the list goes on.
But fuck that. I want a car I can maintain. I bought a $300 "VAG-COM" cable for my VW which lets me access pretty much everything computerized going on in the car, and it has been endlessly useful - it's easily paid for itself by allowing me to do my own maintenance.
I would not be surprised at all if there's some way for a person with the same tool to disable the immobilizer in my car, hotwire the car and steal it. But I'm content with that - the odds of having my car stolen by geeks with scan tools is pretty low, and even if it does happen, I'm pretty sure auto theft insurance ends up costing less than paying for dealer-only maintenance on a german car.
Besides, they'd just invent a better car thief. If they completely lock down a car as I've described, someone would end up stealing a car by hooking up a Megasquirt system just to prove a point...
Actually, the OBD-II specification mandates that you provide a certain set of PIDs without any access restrictions.
As long as that functionality is there, you can do whatever else you want with the port - including locked down, proprietary things. Pretty much car manufacturer out there does this.
Whoever wrote this article needs a smack upside the head. The BBQ Guru controllers are intended to control the temperature of charcoal smokers, and won't do anything for you if you're grilling. I s'pose you could get one to work on a Weber kettle if you really wanted to, but at that point you're cooking with a grill, not grilling.
Anyway I'm using BBQ Guru's PartyQ unit on my Weber Smokey Mountain. I'd recommend the thing almost as much as the WSM itself.
There's a few issues with smartphone screens. Given enough effort you can make pretty much anything work, but here's what you'll be up against:
- Knowing what all the pinouts/connectors/voltages/signal levels are.
- Data format: Most of these screens require a proper graphics controller to drive them, capable of clocking RGB data out of a framebuffer into the panel at a pixel clock of several MHz. You might be able to do this with a PIC32, but your code will be blasting data at the panel 99% of the time. You're in the territory of ARM7/ARM9 processors with SDRAM hanging off them when you're making a bare RGB LCD panel work.
- Power: You'll likely have to generate a backlight voltage, and possibly even bias voltages for the LCD panel itself. The LCD may also run at a different voltage node (3.3V or less) while your AVR might end up being 5.0V.
- Touchscreen: Resistive touchscreen isn't too hard to manage. If it's a capacitive touchscreen you might be able to wire it up to an AVR and use their QTouch libraries to make it work. But I'll warn, prototyping a capacitive touch system can be an exercise in frustration - it's not bad when everything sits in one place on a PCB, but you can't breathe on an airwired capacitive touch system without screwing it up.
Honestly, you're best off finding a "smart LCD" with a built-in controller, with a simple SPI/UART/8-bit-parallel/etc interface. Adafruit has an Arduino compatible one up on their site which might be a good starting point, I'm sure there'll be plenty of other suggestions posted here.
Or hell, you're better off keeping the smartphone whole and finding a way to reprogram it to do what you want.
- Grab a couple books on C/C++ and Verilog. I highly recommend "Fundamentals of Digital Logic with Verilog Design", great for both learning and for reference. For C/C++, I've always been a fan of the Sam's "Learn __ in 24 hours" books.
- Get yourself a FPGA development card, so you can get some "hardware play" in and familiarize yourself with some development tools. I have an Altera DE1 educational card that's a few years old, but it's got endless blocks on it (displays, LEDs, buttons, flash, SDRAM, VGA, sound... you name it) which makes it a great little card for embedded system learning. There's a whole set of Verilog and Nios (embedded processor) tutorials available for it, and lots of online hackers who have ported x86 processors (Zet project), hardware emulations of the NES, etc... to it. Xilinx and Actel also make some nice evaluation boards that seem to be targeted fairly often by hobbyists.
Other than that... you can study the heck out of wire protocols, but you'll probably forget everything you learn unless you end up implementing it. You're better off trying to learn as many general things as you can - how to create well organized C/C++ and Verilog code, making your designs meet timing and such - so that if you end up having to implement something, you've got the basics already in place and don't need too much incremental learning. Also if you have some fun ideas for FPGA projects, implement your heart out - that sort of stuff looks great on a resume.
Don't buy your CFLs at Walmart, the grocery store, etc - the Sunbeam/Great Value/etc bulbs that you find at those kinds of stores are shit.
Buy professional CFL bulbs. Hit up the GE or Osram/Sylvania online product catalogs, write down some part numbers with the size/color temp/lifetime that you want, and call up a local industrial/lighting supplier - Harris & Roome is my go-to place here in Canada.
My house is full of GE "FLE10HT2/827" bulbs, 40W equivalents that pull 10W, have a warm color temperature (2700K) and have a 12000 hour lifetime. Which I can believe - I bought a case of these bulbs about 4-5 years ago when I swapped out every incandescent I could find, I still have plenty of them left, and I honestly can't remember the last time I changed a lightbulb in my house - it's been years.
Unless they're implanting this fucking chip in you, the big brother implication of this chip is pretty much bullshit.
It has one damn good application - reliable navigation, indoors and out. Suppose you've just arrived in Montreal and don't know a thing about the place, but you want to hit up Schwartz's for the sandwich and a pickle that everyone's told you to try. Now your phone can direct you to the nearest subway station, direct you to the correct platform so you don't take the train in the wrong direction, tell you when to get off the train, transfer you to a bus, and drop you off for some kosher deliciousness without having to ask anyone for directions. (Which in Montreal, will either get you told off in French, or you'll end up getting directed to the "club with the best girls" instead of where you want to go..)
100% agreement on what constitutes a good keyboard.
I'm rockin' a Das Keyboard "Model S Professional Silent" on my work PC, and I'm pretty happy with it.
I won't call it silent by any stretch; the switches don't click at the tactile point like an IBM Model M keyboard, but there's a pretty loud clack when you bottom out a key. The tactile "give" is still there - but it's a lot "softer" than the M. When you push down on a M key, there's an increasing resistance until BANG, the bottom pretty much falls out of the key. With the Das Silent, it's more of a 'buttery' tactile feel at lack of a better description - it feels like you "push through" the tactile point instead of "breaking" the tactile point. The Das Silent also requires a fair bit less force to type on than a M.
I've gone back and forth between the Das Silent and the M a few times for the sake of comparison, and I've concluded that I prefer the Das Silent.
The Das Silent uses "Cherry Brown MX" keyswitches - keep an eye open for a deal on a keyboard using the same switches, I think you'll like it.
The $15 card you're describing is a PCMCIA card form factor. Feel free to explain how to power it, plug a keyboard into it and hook it up to their TV without another $15 card with all the connectors you need for a practical application.
Also, the Broadcom on the Pi is about as obsolete as the ATMega parts used on the Arduino card. It gets the job done in the application it's put in. God forbid it's slower than a high end Cortex-A8 processor...
- It allows warrantless monitoring of a person's internet activity, which is an inexcusable invasion of privacy. - According to a law professor I've spoken to, it makes warrantless monitoring of my telephone legal since I'm using a VOIP telephone system provided by the cable company, and VOIP is just extra internet traffic. This is despite the fact that warrantless monitoring of a landline is highly illegal. - Oh, and if I burn a copy of a DVD I own, I've broken a digital lock and I'm now I'm potentially liable for unlimited damages. They were nice enough to add a backup provision to C-11, except it's not applicable if you're breaking a digital lock.
Not to mention the fact it's being thrust on us by a majority government that got less than 50% of the overall vote at election time, may not actually have a legitimate majority thanks to the ongoing robocall scandal... and the whole fucking bill is being pushed on us with "support it or you're with the pedophiles" bullshit.
And here's what I've done about it:
- I've written my local (Conservative) MP about the bill, and got a template reply. Wrote back again, no response. - I wrote a former (NDP) MP in my previous district, and got back a really great "100% with you on this, but unfortunately they've got a majority so all we can do is whine and complain in the house" e-mail.
Converting from one sample rate to another, provided it's done using a proper asynchronous sample rate conversion algorithm, will be just as acoustically transparent as converting between two rates that are multiples of each other.
Having the two sample rates you're converting between be multiples of each other, or rational, does help with the computational efficiency somewhat. But other than that, it's mathematically the same process.
The worst assumption you can make is that since one audio sampling rate is a multiple of the other, it's an easy process of just "adding and dropping samples". It's not; any rate conversion process has to be combined with a filtering process in order to prevent high frequencies aliasing to low frequencies (if lowering the sample rate) or low frequencies being 'duplicated' up into higher frequencies (if raising it).
(DSP engineer here, I've been writing audio processing code for almost 10 years..)
You can run the same mathematical processes on a TMS320, DSP56K, FPGA, desktop PC, GPGPU video card, embedded ARM, PIC, AVR, even a 4-bit micro if you feel the need. Sure, a purpose-built DSP might be a bit faster at it than a microcontroller, but if the processor you choose is fast enough to run your DSP-like algorithm at the speed you need, what's the damn problem?
Case in point, there's really nothing better suited for decoding MP3/AAC audio than a DSP chip - a proper DSP can rip through filterbanks, MDCTs, etc. Yet almost every MP3 player on the market decodes these formats with an ARM core.
And back to the Arduino... AVRs have an 8x8->16 multiplier, post-increment/decrement addressing, and can run up to 20MHz so they're actually not that bad a choice for simple DSP algorithms that don't need a lot of dynamic range. An AVR has no trouble doing DTMF decoding on values captured from the ADC, or running a few DDS channels to generate sine waves and throw them out a PWM channel, despite those being classical DSP applications.
Same thing happened here. I bought a "used" iPod Nano off eBay for a good price, which had normal looking pictures of a genuine model taken out of the box. When it came in the mail... it was a different color, quite a bit larger, and when I turned it on a horribly pixelated screen showed the Apple logo and the text "HELLO" below it. Right.
I disputed the payment and got my refund from Paypal, and they asked that I destroy the iPod clone and take pictures. I proceeded to clamp it in my bench vise and saw it in half with a hacksaw, while my girlfriend took pictures of the process. Well, I struck the battery with the hacksaw - smoke and fire ensued. Once it died down and I had aired the smoke out of my basement, I finished sawing the now burnt and discolored iPod clone in half.
I'm pretty sure those pictures are thumbtacked to someone's cubicle wall at Paypal now.
The gate of a FET is effectively a capacitor. Even with the FET in the on state, if you keep increasing the gate voltage it'll still keep taking electrons. And like a capacitor, energy stored in a FET gate = 1/2*C*V^2. You also have source/drain and gate/drain (miller) capacitance - source/drain has to be discharged (another 1/2CV^2 loss) and the miller capacitance has to be discharged and then charged at the opposite polarity (a CV^2 loss).
Overall, neglecting leakage current, power loss is proportional to frequency, but it's also proportional to voltage squared.
Power loss is also proportional to transistor count, which is why ARM is such a low power processor.
A GPU will spank a dedicated DSP chip at just about everything, even the highest end TI's and TigerSHARCs. Both DSPs and GPUs are designed to haul data out of memory and do vector multiplication on it, but the GPU has a heck of a lot more of both memory bandwidth and processing grunt.
A big FPGA card, or FPGA array system like a Copacobana, might be quicker assuming I/O limitations aren't a problem for the algorithm to be run. But FPGA hardware for HPC isn't really a commodity so it's awfully expensive - you're looking at $5K+ for a big Virtex on a PCIe card. Plus buying FPGA tools and IP blocks, and getting the VHDL/Verilog written, will eat up a budget really quick.
But if this is being done in an academic environment and there's no looming deadline for this project, the FPGA method might be something you can get a grant for and throw a computer engineering grad student or co-op student at.
Other than, "it's new and people don't fully understand it" ? Or, ?
If people had that same mindset/fear of the unknown that they did when penicillin and vaccines came out, I think we'd be seriously fucked as a human race.
I seem to remember the potato blight being a terrible thing that killed millions of people in the Irish/Scottish/European famines. And I personally know a family in Newfoundland who were farmers - several years back their potato crop contracted late blight, antifungals didn't help, they lost the crop and ended up bankrupt at the end of it. A blight resistant strain of potatoes seems like a pretty fantastic idea to me.
Besides, the more food that we grow that doesn't need antifungals, pesticides and other "of course they're toxic, they wouldn't work otherwise" chemicals sprayed on it for it to grow, the better. I'd eat a GM vegetable any day over that.
(Mind you, I'm personally against engineering salmon to be 10 times bigger and growing them in offshore fish farms. Grow that shit in an inland fish farm where it's guaranteed that they won't take over and fuck up an ecosystem.)
I still have the 486DX2/66 that I bought with my first summer job. 8 megs of RAM, SB16 and GUS side-by-side, 540 meg hard drive, and 2X Panasonic CDROM with a proprietary interface. It has a cheesy yellow 7-segment LED dispay on the front that displays the computer MHz and switches between 33 and 66 when you push the TURBO button. And I love the thing - the computer came from one of my favorite times in the computer scene, the demoscene was thriving as well as the MOD/S3M/IT community, games were a ton of fun, and the computer's loaded up with all of that good stuff.
Despite all of that, I haven't turned it on in over a year. It's a heck of a lot easier to start up DOSBox than drag the old computer out.
Slashdot Mirror
Because in my experience, the yield from the chinese Model B factory is 50%.
My first RPi is currently tied up in a work project, so I ordered another model B from Newark. It came in and I fired it up yesterday, no LEDs or any signs of life. Dead.
Then I noticed the main BGA in the center of the card looked a bit askew, looked closer and noticed the BCM2835 was missing. The Samsung DRAM that ordinary sits on top of the '2835 was soldered straight onto the PCB. I understand the part shooter fucking up once in a while and missing a chip, but the board shouldn't have made it out of the factory.
C'mon. I'd rather pay a few extra bucks for something that's most likely going to work, than do what I'm doing now and spending even more bucks mailing the fucking thing back, and crossing my fingers that the replacement works too...
One of my friends did exactly what you're describing - he works on an offshore oil rig, and used a high gain antenna pointed at land to get cellphone service. It worked great... until he came ashore and his phone didn't get a signal.
Turns out the RF connector in the phone is only there for production testing of power/RF compliance - it's only intended to be connected to once, and it's not designed for daily connections/disconnections. The datasheet for one replacement connector we found was only rated for 10 cycles.
Other than sharing the same 2.4GHz ISM frequency band, bluetooth (used for their controllers) and WiFi share basically nothing. In the regular Wii they use separate chipsets, separate antennas, etc. They're stripping out a few bucks worth of hardware. Not to mention, implementing WiFi in a product invariably involves paying a bunch of license fees to patent holders etc - either included in the price of your wifi chipset or paid separately. This adds a couple more bucks to the design.
Take the $99 they're selling the thing for at the store and subtract store markup, shipping, factory tooling, packaging, the rest of the stuff in the Wii box, etc. and you don't have much money to spend on the Wii Mini itself. A few bucks spent adding WiFi could end up being a significant part of the cost.
Honestly, I have to throw some praise at Nintendo for making a game system that's so cheap. My workplace is doing an "adopt-a-family" thing for the christmas season, where employees get together and buy christmas gifts for single parents who can't afford much for their kids. At $99, it's made our shopping list.
Actual use cases I've seen for the Raspberry Pi that I've done myself and seen others do:
- Plug the Rpi into a LAN, and connect it to the serial console of a piece of equipment with a USB to serial cable - old router, telephone equipment, radio broadcast transmitter, fill in the blank. SSH into the thing if you need to get at the console instead of doing a site trip.
- Plug a few sensors into it, run it off a 12V car battery and a +5V automotive USB adapter, and leave it somewhere to log data onto the SD card or a USB stick.
- Plug a USB hard drive into it, and use it as a low power torrent downloader, instead of keeping your desktop PC powered up when you're not home.
It's a tiny, $25 linux machine. Possibilities are endless.
And in other news, Arduino cards have a 16MHz 8-bit processor with mere kilobytes of both RAM and flash. And despite making a 1980s suitcase computer look fast, they've proven themselves to be fully capable of running all sorts of awesome things that hobbyists have been using them for.
What's your point again?
I worked on a system similar to this for a man with an prosthetic arm. We stuck a temperature sensor in the finger of the prosthetic hand, and used a small thermoelectric plate which contacted his skin where the prosthetic attached to him. The back side of the peltier plate was glued to an aluminum bar acting as a thermal source/sink, and the front side had a small stainless plate acting as the contact point with a second temperature sensor bonded to it for feedback. Using a microcontroller and a bit of simple hardware, we made the peltier plate temperature equal to the temperature detected at the finger. With some limits, of course.
The man went from having a plastic arm to having something that *felt* like an actual part of his body again. He described going home and touching his wife's face, and almost breaking down crying - it felt pretty good to hear that, especially since it only took us a few days to design/build the thing.
Anyway, it ran off three NiCd AA batteries and had a battery life of about 24 hours - he'd plug it in to charge when he went to bed, and it would easily last until the next charge cycle. These days thermoelectric devices are more efficient, batteries are a lot better and microcontrollers have much better power consumption.
There's a few things that make a system like this fairly low power:
- You don't need a big contact area to get the sensation of temperature across to the user - you don't need to heat or cool their whole hand. Half a square centimeter is plenty if you put it at a fairly temperature-sensitive part of the hand, such as where your fingers meet your palm. It'll feel weird at first but the 'immersion' feeling will eventually set in.
- There's only a narrow range of temperature that you have to drive the contact plate to. You don't need to do 0 to 50 degrees C, and I fully expect Sony to restrict the range to +-10C at most for liability reasons, not to mention practical reasons.
- Skin isn't *that* thermally conductive.
So if your contact plate is sized small and only within a few degrees of body temperature, you're probably only moving half a watt at most between the hand from the contact plate.
Secondly, I seriously doubt the thermal "immersion" effect will be running all the time, probably only acting on 'events' the game - walking indoors/outdoors (pulse of hot/cold), picking up a weapon from the ground (cold), getting hit with a fireball (hot), falling into water (cold), etc. Much like vibration motors in controllers don't run all the time.
End result is that running the thermoelectrics won't take that much power, and sinking/sourcing heat from within the controller shouldn't be that hard. Overall, seems pretty practical to me.
Something to consider - I've replaced the MicroUSB connector in my cellphone *twice*. The phone would work for about a year, then it would go flaky - you'd have to wiggle the connector a few times for the phone to reliably charge, and sometimes I'd go check on it and it wouldn't be charging - and it would happen with different cables. Supposedly these things are rated for 10,000 cycles, but I haven't seen it. Maybe my phone does something it shouldn't, like spark the +5V pin when the connector is plugged in. *shrug*
Secondly, I've caught the cord of my phone multiple times and pulled it off the desk onto the floor - and my cats/dogs have probably done it more times than I have.
Though there's an efficiency loss in wireless charging versus conductive charging, I wonder if there's an efficiency gain that exists in less phones being repaired/replaced because of damage related to conductive charging.
(Note that this is not a well thought out, researched argument - just a dumb thought.)
Do this. Talk to the math teachers at your school, find out if they've got any poor students that need them. And find out there's any other schools in the area that would have a use for them.
There's lots of single parents and otherwise poor families that can barely scrape together school supplies for their kids, let alone buy the graphic calculator that they would need to get into a precalc or AP math. Something simple like one of these old calculators could turn a kid's life around. Seriously.
I'm suspecting that the BMW is "password protected" in a way (unique knowledge required to get into the car's ECU - PIDs to hit, passphrases, etc) but if the car thief that knowledge then it's essentially open anyway.
Pretty much the only way you'll make a perfectly secure system is by putting the equivalent of a TPM chip in the car and using a cryptographic channel between the car and BMW themselves. Do authentication at the garage and employee levels, and have BMW log everything that happens to every car it talks to, so any malicious use gets recorded and you can figure out who enabled your car to get stolen. Do code signing in the ECU so they can't reprogram the ECU to run different code, authenticate the ECU against other modules in the car so they'd have to basically swap every module in the car... and the list goes on.
But fuck that. I want a car I can maintain. I bought a $300 "VAG-COM" cable for my VW which lets me access pretty much everything computerized going on in the car, and it has been endlessly useful - it's easily paid for itself by allowing me to do my own maintenance.
I would not be surprised at all if there's some way for a person with the same tool to disable the immobilizer in my car, hotwire the car and steal it. But I'm content with that - the odds of having my car stolen by geeks with scan tools is pretty low, and even if it does happen, I'm pretty sure auto theft insurance ends up costing less than paying for dealer-only maintenance on a german car.
Besides, they'd just invent a better car thief. If they completely lock down a car as I've described, someone would end up stealing a car by hooking up a Megasquirt system just to prove a point...
Actually, the OBD-II specification mandates that you provide a certain set of PIDs without any access restrictions.
As long as that functionality is there, you can do whatever else you want with the port - including locked down, proprietary things. Pretty much car manufacturer out there does this.
Whoever wrote this article needs a smack upside the head. The BBQ Guru controllers are intended to control the temperature of charcoal smokers, and won't do anything for you if you're grilling. I s'pose you could get one to work on a Weber kettle if you really wanted to, but at that point you're cooking with a grill, not grilling.
Anyway I'm using BBQ Guru's PartyQ unit on my Weber Smokey Mountain. I'd recommend the thing almost as much as the WSM itself.
There's a few issues with smartphone screens. Given enough effort you can make pretty much anything work, but here's what you'll be up against:
- Knowing what all the pinouts/connectors/voltages/signal levels are.
- Data format: Most of these screens require a proper graphics controller to drive them, capable of clocking RGB data out of a framebuffer into the panel at a pixel clock of several MHz. You might be able to do this with a PIC32, but your code will be blasting data at the panel 99% of the time. You're in the territory of ARM7/ARM9 processors with SDRAM hanging off them when you're making a bare RGB LCD panel work.
- Power: You'll likely have to generate a backlight voltage, and possibly even bias voltages for the LCD panel itself. The LCD may also run at a different voltage node (3.3V or less) while your AVR might end up being 5.0V.
- Touchscreen: Resistive touchscreen isn't too hard to manage. If it's a capacitive touchscreen you might be able to wire it up to an AVR and use their QTouch libraries to make it work. But I'll warn, prototyping a capacitive touch system can be an exercise in frustration - it's not bad when everything sits in one place on a PCB, but you can't breathe on an airwired capacitive touch system without screwing it up.
Honestly, you're best off finding a "smart LCD" with a built-in controller, with a simple SPI/UART/8-bit-parallel/etc interface. Adafruit has an Arduino compatible one up on their site which might be a good starting point, I'm sure there'll be plenty of other suggestions posted here.
Or hell, you're better off keeping the smartphone whole and finding a way to reprogram it to do what you want.
- Grab a couple books on C/C++ and Verilog. I highly recommend "Fundamentals of Digital Logic with Verilog Design", great for both learning and for reference. For C/C++, I've always been a fan of the Sam's "Learn __ in 24 hours" books.
- Get yourself a FPGA development card, so you can get some "hardware play" in and familiarize yourself with some development tools. I have an Altera DE1 educational card that's a few years old, but it's got endless blocks on it (displays, LEDs, buttons, flash, SDRAM, VGA, sound... you name it) which makes it a great little card for embedded system learning. There's a whole set of Verilog and Nios (embedded processor) tutorials available for it, and lots of online hackers who have ported x86 processors (Zet project), hardware emulations of the NES, etc... to it. Xilinx and Actel also make some nice evaluation boards that seem to be targeted fairly often by hobbyists.
Other than that... you can study the heck out of wire protocols, but you'll probably forget everything you learn unless you end up implementing it. You're better off trying to learn as many general things as you can - how to create well organized C/C++ and Verilog code, making your designs meet timing and such - so that if you end up having to implement something, you've got the basics already in place and don't need too much incremental learning. Also if you have some fun ideas for FPGA projects, implement your heart out - that sort of stuff looks great on a resume.
Good luck!
Don't buy your CFLs at Walmart, the grocery store, etc - the Sunbeam/Great Value/etc bulbs that you find at those kinds of stores are shit.
Buy professional CFL bulbs. Hit up the GE or Osram/Sylvania online product catalogs, write down some part numbers with the size/color temp/lifetime that you want, and call up a local industrial/lighting supplier - Harris & Roome is my go-to place here in Canada.
My house is full of GE "FLE10HT2/827" bulbs, 40W equivalents that pull 10W, have a warm color temperature (2700K) and have a 12000 hour lifetime. Which I can believe - I bought a case of these bulbs about 4-5 years ago when I swapped out every incandescent I could find, I still have plenty of them left, and I honestly can't remember the last time I changed a lightbulb in my house - it's been years.
Unless they're implanting this fucking chip in you, the big brother implication of this chip is pretty much bullshit.
It has one damn good application - reliable navigation, indoors and out. Suppose you've just arrived in Montreal and don't know a thing about the place, but you want to hit up Schwartz's for the sandwich and a pickle that everyone's told you to try. Now your phone can direct you to the nearest subway station, direct you to the correct platform so you don't take the train in the wrong direction, tell you when to get off the train, transfer you to a bus, and drop you off for some kosher deliciousness without having to ask anyone for directions. (Which in Montreal, will either get you told off in French, or you'll end up getting directed to the "club with the best girls" instead of where you want to go..)
100% agreement on what constitutes a good keyboard.
I'm rockin' a Das Keyboard "Model S Professional Silent" on my work PC, and I'm pretty happy with it.
I won't call it silent by any stretch; the switches don't click at the tactile point like an IBM Model M keyboard, but there's a pretty loud clack when you bottom out a key. The tactile "give" is still there - but it's a lot "softer" than the M. When you push down on a M key, there's an increasing resistance until BANG, the bottom pretty much falls out of the key. With the Das Silent, it's more of a 'buttery' tactile feel at lack of a better description - it feels like you "push through" the tactile point instead of "breaking" the tactile point. The Das Silent also requires a fair bit less force to type on than a M.
I've gone back and forth between the Das Silent and the M a few times for the sake of comparison, and I've concluded that I prefer the Das Silent.
The Das Silent uses "Cherry Brown MX" keyswitches - keep an eye open for a deal on a keyboard using the same switches, I think you'll like it.
The $15 card you're describing is a PCMCIA card form factor. Feel free to explain how to power it, plug a keyboard into it and hook it up to their TV without another $15 card with all the connectors you need for a practical application.
Also, the Broadcom on the Pi is about as obsolete as the ATMega parts used on the Arduino card. It gets the job done in the application it's put in. God forbid it's slower than a high end Cortex-A8 processor...
I think C-11 is an absolutely horrible bill.
- It allows warrantless monitoring of a person's internet activity, which is an inexcusable invasion of privacy.
- According to a law professor I've spoken to, it makes warrantless monitoring of my telephone legal since I'm using a VOIP telephone system provided by the cable company, and VOIP is just extra internet traffic. This is despite the fact that warrantless monitoring of a landline is highly illegal.
- Oh, and if I burn a copy of a DVD I own, I've broken a digital lock and I'm now I'm potentially liable for unlimited damages. They were nice enough to add a backup provision to C-11, except it's not applicable if you're breaking a digital lock.
Not to mention the fact it's being thrust on us by a majority government that got less than 50% of the overall vote at election time, may not actually have a legitimate majority thanks to the ongoing robocall scandal... and the whole fucking bill is being pushed on us with "support it or you're with the pedophiles" bullshit.
And here's what I've done about it:
- I've written my local (Conservative) MP about the bill, and got a template reply. Wrote back again, no response.
- I wrote a former (NDP) MP in my previous district, and got back a really great "100% with you on this, but unfortunately they've got a majority so all we can do is whine and complain in the house" e-mail.
What the frig else should I do?
Converting from one sample rate to another, provided it's done using a proper asynchronous sample rate conversion algorithm, will be just as acoustically transparent as converting between two rates that are multiples of each other.
Having the two sample rates you're converting between be multiples of each other, or rational, does help with the computational efficiency somewhat. But other than that, it's mathematically the same process.
The worst assumption you can make is that since one audio sampling rate is a multiple of the other, it's an easy process of just "adding and dropping samples". It's not; any rate conversion process has to be combined with a filtering process in order to prevent high frequencies aliasing to low frequencies (if lowering the sample rate) or low frequencies being 'duplicated' up into higher frequencies (if raising it).
(DSP engineer here, I've been writing audio processing code for almost 10 years..)
DSP is a *process*, not a *processor*.
You can run the same mathematical processes on a TMS320, DSP56K, FPGA, desktop PC, GPGPU video card, embedded ARM, PIC, AVR, even a 4-bit micro if you feel the need. Sure, a purpose-built DSP might be a bit faster at it than a microcontroller, but if the processor you choose is fast enough to run your DSP-like algorithm at the speed you need, what's the damn problem?
Case in point, there's really nothing better suited for decoding MP3/AAC audio than a DSP chip - a proper DSP can rip through filterbanks, MDCTs, etc. Yet almost every MP3 player on the market decodes these formats with an ARM core.
And back to the Arduino... AVRs have an 8x8->16 multiplier, post-increment/decrement addressing, and can run up to 20MHz so they're actually not that bad a choice for simple DSP algorithms that don't need a lot of dynamic range. An AVR has no trouble doing DTMF decoding on values captured from the ADC, or running a few DDS channels to generate sine waves and throw them out a PWM channel, despite those being classical DSP applications.
- Signed, someone who does DSP for a living.
Same thing happened here. I bought a "used" iPod Nano off eBay for a good price, which had normal looking pictures of a genuine model taken out of the box. When it came in the mail... it was a different color, quite a bit larger, and when I turned it on a horribly pixelated screen showed the Apple logo and the text "HELLO" below it. Right.
I disputed the payment and got my refund from Paypal, and they asked that I destroy the iPod clone and take pictures. I proceeded to clamp it in my bench vise and saw it in half with a hacksaw, while my girlfriend took pictures of the process. Well, I struck the battery with the hacksaw - smoke and fire ensued. Once it died down and I had aired the smoke out of my basement, I finished sawing the now burnt and discolored iPod clone in half.
I'm pretty sure those pictures are thumbtacked to someone's cubicle wall at Paypal now.
Actually, voltage matters substantially.
The gate of a FET is effectively a capacitor. Even with the FET in the on state, if you keep increasing the gate voltage it'll still keep taking electrons. And like a capacitor, energy stored in a FET gate = 1/2*C*V^2. You also have source/drain and gate/drain (miller) capacitance - source/drain has to be discharged (another 1/2CV^2 loss) and the miller capacitance has to be discharged and then charged at the opposite polarity (a CV^2 loss).
Overall, neglecting leakage current, power loss is proportional to frequency, but it's also proportional to voltage squared.
Power loss is also proportional to transistor count, which is why ARM is such a low power processor.
A GPU will spank a dedicated DSP chip at just about everything, even the highest end TI's and TigerSHARCs. Both DSPs and GPUs are designed to haul data out of memory and do vector multiplication on it, but the GPU has a heck of a lot more of both memory bandwidth and processing grunt.
A big FPGA card, or FPGA array system like a Copacobana, might be quicker assuming I/O limitations aren't a problem for the algorithm to be run. But FPGA hardware for HPC isn't really a commodity so it's awfully expensive - you're looking at $5K+ for a big Virtex on a PCIe card. Plus buying FPGA tools and IP blocks, and getting the VHDL/Verilog written, will eat up a budget really quick.
But if this is being done in an academic environment and there's no looming deadline for this project, the FPGA method might be something you can get a grant for and throw a computer engineering grad student or co-op student at.
Other than, "it's new and people don't fully understand it" ? Or, ?
If people had that same mindset/fear of the unknown that they did when penicillin and vaccines came out, I think we'd be seriously fucked as a human race.
I seem to remember the potato blight being a terrible thing that killed millions of people in the Irish/Scottish/European famines. And I personally know a family in Newfoundland who were farmers - several years back their potato crop contracted late blight, antifungals didn't help, they lost the crop and ended up bankrupt at the end of it. A blight resistant strain of potatoes seems like a pretty fantastic idea to me.
Besides, the more food that we grow that doesn't need antifungals, pesticides and other "of course they're toxic, they wouldn't work otherwise" chemicals sprayed on it for it to grow, the better. I'd eat a GM vegetable any day over that.
(Mind you, I'm personally against engineering salmon to be 10 times bigger and growing them in offshore fish farms. Grow that shit in an inland fish farm where it's guaranteed that they won't take over and fuck up an ecosystem.)
I still have the 486DX2/66 that I bought with my first summer job. 8 megs of RAM, SB16 and GUS side-by-side, 540 meg hard drive, and 2X Panasonic CDROM with a proprietary interface. It has a cheesy yellow 7-segment LED dispay on the front that displays the computer MHz and switches between 33 and 66 when you push the TURBO button. And I love the thing - the computer came from one of my favorite times in the computer scene, the demoscene was thriving as well as the MOD/S3M/IT community, games were a ton of fun, and the computer's loaded up with all of that good stuff.
Despite all of that, I haven't turned it on in over a year. It's a heck of a lot easier to start up DOSBox than drag the old computer out.