What I don't understand is why don't they don't propose using the traditional Lattitude/Longitude with minutes and seconds. Assuming three digits for Lattitude/Longitude and two each for minutes and seconds, you need 14 digits.
This translates to 7 bytes using BCD and should get comparable accuracy to what is being proposed. AND, it would be immediately compatible with GPS systems already available.
Wasn't the tracking/selling of viewer habits something that Tivo promised not to do when it first came out? I seem to remember this was an issue when Tivo first came out (along with its need for a phone line to download programming information, there were concerns it would phone back out and pass along personal information).
I love the last quote from Martin Yudkovitz: We want to embrace the opportunity to help advertisers and broadcasters better understand the current and future impact of DVR so they can adapt content and advertising strategies to this new medium. If they want to help advertisers and broadcasters, why are they selling the information?
Of course it might not be a bad thing; if they base the advertising that's passed to me on which commercials I don't fast forward through, it will be 24-7 of Victoria's Secret commercials - Party at my house, dudes!
It's funny reading this article because I can think of a number that came out a dozen years or so ago with exactly the same theme; games were all the same, there was nothing new, game developers were rehashing the same old concepts over and over again...
And then came Wolfenstein and suddenly games weren't boring any more and people couldn't get enough FPS. When things got boring again, out came "The Sims"...
I have no doubt that the pattern will repeat itself and somebody out there has a great new concept for a game and it will be "discovered" now that people consider the current ones boring and manufacturers are looking for new concepts to exploit.
myke
This is what got me banned...
on
Making Change
·
· Score: 1
at my work cafeteria. The manager of the cafeteria said that by doing this, "I made the cashiers feel stupid."
I tried pointing out that this was grade 3 math and it made it easier for the cashier because they would have to return less change. The advantage to me was there was less shrapnel in my pocket.
The upside was that I could go out whenever I wanted because the cafeteria refused to serve me. (See the previous article about dodge work).
myke
But it doesn't fix the big problem
on
DVRs for Cop Cars
·
· Score: 1
I'm sure that digital technology will help out with some of the current issues but what about the situation in which the cop and the offender are out of the frame? I've seen a number of tapes where the suspect has been taken out of the frame and has later sued the police for brutality and/or violating their rights. Ideally, the camera would follow the action, even if it was not directly in front of the car.
I've discussed this question with a hospital ethicist - the 8 year old would almost always get the liver. But, for practical purposes the question is mostly hypothetical.
The decision is usually made based on the simple reason that in the case of the alcoholic, there is generally a lot more wrong with them than just the liver, replacing the liver will extend their lives but not dramatically. You're talking two years or so at the most.
The little girl would get a lot more use out of the liver.
There is another question that comes into play and that is whether or not the alcholism will negate the success of the transplant. Heavy drinking will attenuate the effectiveness of the anti-rejection drugs as well as impair the patients ability to follow a medication schedule.
Right now, there is a push to make sure organs are passed to the "best" match, not the next "acceptable" match in line. The better the match, the higher the probability that the transplant will be successful.
In the case of livers, as I understand it, very few transplants are now done from unrelated donors. Partial liver transplants from relatives seems to be the preferred method with organs harvested from donors being used rarely.
I would use the term "satisfaction"; I remember being very pleased with myself for coming up with a sixteen bit by sixteen bit shifting multiplication algorithm in just 24 instructions or realizing that self-modifying code is needed to perform an indirect jump or data access anywhere in the 16 bit address space.
If you can competently program the 6502 in assembler, then you will be able to tackle any other 8 bit processor architecture with ease - a very satisfying accomplishment.
myke
Re:Next we tested IIS on both Linux and Windows
on
What's Microsoft Up To?
·
· Score: 4, Interesting
Actually, I would have liked to see a comparison on different hardware. I have a few questions regarding the hardware:
Why was HP servers only used in the comparison?
I'm not sure of the chipset, but I believe that these servers use HP proprietary Northbridge/Southbridges which could affect performance, I would have preferred to see a Dell with true Intel and/or RCI chipsets included in the test.
Along these lines, I'm suspicious as to why the DL380 servers were configured with 1.4 GHz PIIIs and not 2.8 GHz Xeons (and the DL760s had 900 MHz Xeons and not 2 GHz Xeons)
I would have also liked to see different size SDRAM configurations for the test.
As you said, a comparison of different network file systems would be nice as well.
Finally, an interesting measure would have been how often each OS crashed and had to be rebooted.
Now having said all this, I'm not surprised, I've been reading performance comparisons for 25 years and strangely enough, the sponsoring company's hardware/software/operating systems always seem to come out on top. This started with comparing the 8086 to the 68000 and has continued on to the present day.
The important/best thing about the review is that it states very clearly at the top that the test was sponsored by Microsoft.
Maybe I'm showing my age, but AV was the best dodge you could have in high school.
You asked, where is the fun carrying around a TV? It is infinitely more interesting than many things that you would have to do if you weren't carrying around a TV (ie sitting in a health class looking at pictures of cholesterol blocked arteries).
Want to miss a class? Come in five minutes before the final bell, apologizing and mutter under your breath about the splicing job you just had to do.
Splicing is great. In the '70s, it could mean film, (audio) tape, power cords, speaker cords or video co-ax.
AV means never being called on the carpet for the classes you cut AND they make allowances for the material you missed when its test time.
This is a follow up to my previous response asking who read over the letter (because anybody halfway competent could have figured out what was done even if the web mirrors were taken out).
Hiding behind DMCA is clearly not the answer. It destroys credibility for the company and ultimately for the law itself. If these cards are used for purchasing products/services than potentially millions of dollars are at risk - nobody can realistically expect a simple law to protect against theft of tangible items as well as theft of intellectual property.
But, how much does a letter from a lawyer cost? From the company's perspective, this is the most cost effective way to deal with the problem.
The second most cost effective way of dealing with the problem is to hire two guys to show Virgil and Acidus the errors in their ways (preferably from the perspective of the inside of a Cadillac's trunk).
I would suspect that this would be a bit more expensive than a lawyer's letter, but almost certainly more effective.
If either of the first two methods don't work, then Blackboard could, gasp, fix their product. I disagree that the encryption couldn't be built into the individual boxes - there are some very clever things that can be done that would make figuring out the communications very difficult.
This is the most expensive option, but the one that would have the least amount of liability for Blackboard and could establish them as the brand owner (get rid of all those pesky ATT labeled boxes).
Reading through the C&D letter, I have to wonder who approved it from Blackboard's perspective and if anybody technical thought through what may be the result of it is.
There sounds like there is enough information in the letter so that somebody that knows what a 75176 is (I would disagree with the assertions in the paper about RS-485's obscurity), can program a PIC or an 8051 and can use an oscilloscope can reproduce the work done by Messrs. Griffith and Hoffman. Along with this it sounds like the readers are connected to standard cabling via standard connectors.
So, the result I would expect from this letter is, 1) it will be put on the Internet for all to read, 2) boxes throughout the different colleges and universities that use the system will be pulled out of walls and vending machines with many of them stolen or vandalized to see what's actually inside them, next 3) The protocol and hardware will be distributed on a variety of web sites (probably ending with.ru or.iq) and finally 4) Blackboard's reps get innundated with phone calls, emails and letters complaining that their system is not secure.
This begs the question on what Blackboard should have done. (next reply).
This seems like a nice technology to remove the heat from the CPU, but what I'm always wondering about is, where will the heat actually be dissipated into the environment? At some point, there has to be a heat exchanger where all this heat collected in the tiny bubbles is passed outside the unit. This is going to take a fair amount of space - one of these days we're going to see ads for heat exchangers that take up less space than the "standard" box available from Intel.
I'm looking forward to a Beowolf cluster not only performing amazing calculations but also heating the building it's in.
Wouldn't the Linux/OSS advancement efforts be better served if maybe this team of people had worked for a week and released a complete "Duke Nukem 3D" for Linux? The problems listed above (including the need for an MS-DOS non-shareware version of the game) will probably tend to turn people off and not think that Linux is ready for the prime-time desktop.
I would think that this effort would be a better advertisement for Free Software/Linux if shareware data files could be processed, networking/modem/multiplayer modes worked along with correct audio and mouse/joystick support. I would also think that an RPM of the application should be available as well.
This should be much less effort than porting the package to Windows or getting demos working but would be impressive enough that casual computer users would consider Linux for their desktops.
As other posters have indicated, reflowing solder paste in a toaster oven has been done for years and simulates the standard process for SMT assembly.
Using Solder Paste, you can use a hot air gun to place components on a PCB as well. A woman I know at work (Celestica) made a video demonstrating the SMT solder process using a hot air gun - it came out quite nicely and her joke on customers was saying that she followed the board through the oven (it was vapor phase at the time). Many customers were impressed with her tolerance to extreme heat.
As somebody noted, most components will stop working at 140C and in the oven they will go over 200C - they will survive, but the PCB should not be powered up until the PCB has cooled to room temperature.
If you're going to try this at home, a few comments:
1. Solder paste will only stay reliably sticky for 30 minutes. Make sure that you have your components ready and the oven at the primary temperature before you break the seals on the syringe and start applying paste. Make sure that you don't have more components than you can place in 20 minutes.
2. Solder "paste" is made up of finely ground solder held together by flux. Both the solder and the flux are poisenous and during the solder process you will see a build up of flux on the inside of the oven. Along with this, the solder may form "balls" that can be thrown off the PCB. An oven used for SMT experiments cannot be used for food preparation afterwards.
3. Solder paste must be capped and refridgerated when not in use. If you are storing it in a fridge where food is stored, make sure that it is in something like a tupperwear container and well marked (especially if children are around). It looks like pate or liverwurst, but will sit in your stomach like a ball of lead (sorry, couldn't resist).
4. The PCB coming out of the oven is very hot and will take several minutes to cool down. I've heard of a number of people that have built SMT boards in a toaster oven, only to forget oven mitts and tongs to handle the hot PCBs and ended up dropping the PCB on the floor and burning themselves. One genius I heard about was sitting down when he pulled the board out of the oven without any mitts or tongs... Make sure you have something like a barbeque grill ready for the PCB to sit on when it comes out of the over.
5. The PCB should be as dry as possible. Before putting on paste/components, you might want to put it into the over for a day or so at the lowest setting to try and bake out any water that has gotten trapped in it. Let the PCB cool before applying paste.
6. The PCB pads should be "HASL" ("Hot Air Solder Leveled") for best results (do not try this on bare copper and you may have to experiment with gold finishes).
7. I would suggest using parts with leads on 0.050" (50 mil) centers, 0805 chip components and SOT-23 transistor and diodes. Anything smaller will make applying the solder very difficult. The article indicates the author used smaller spacing components, but not how many and how the PCB was laid out.
8. Do not use surface mount connectors. Unless you are very comfortable with doing your own soldering, you will find that it is difficult to get a uniformly strong joint on every pin.
9. If you are designing your own PCB, you can use Protel's "EasyTrax", which is an MS-DOS Command Line program that can be downloaded for free from a variety of sources (you should be able to find where on Google). I have added IPC standard pad layouts for the library components.
I've done it a couple of times with an old toaster oven and it works surprisingly well. Just make sure you plan out what you are going to do and if there are any terms that I have used above that you are unfamilia with, make sure that you investigate them before trying it out on your own.
Reading the memo, figuring out whether the policy change is for hardware or software is somewhat unclear.
Like many other people, I have had to move up to Win/XP and I have found a lot of my hardware/software no longer works, causing me to have to buy replacement products that are "designed for Win/XP". I'm sure for a retailer, this is a major headache as inventory comes back and customers gripe over having to buy replacements.
The question I have to ask is, why is Office Depot sending the memo out to a whole bunch of suppliers when they should only be sending out one? Doesn't it make a lot more sense to send a memo to Microsoft saying that they have found that their new operating system has caused a lot of distress for their customers and rather than requiring their customers and other vendors conform to Win/XP could Microsoft please fix the problems with the operating system so that legacy products will work? In the meantime, Office Depot will cancel all pending orders of Windows/XP until the bugs in it are fixed.
I wouldn't characterize them as being like the Seinfeld "Soup Nazi", it would be more accurate to say they aren't very patient with stupidity.
Personally, I think if you showed up at 12:15 PM, when there are a couple of dozen people at the single front counter and held up an ancient "Princess Phone" asking if they have the rotary dial version in purple and then ask how long the warranty they give on it is, then you deserve to be called a throwback (or worse) in a slavic language.
Common sense would indicate that the example above would be a dumb thing to do, but I can't tell you how many times I have seen people do things just like the phone example (and at other stores to be fair) - at Active Surplus you just don't get responses to these kind of questions with the kind, patient, medicated voice you would expect from The Gap.
myke
Ettiquette for Active Surplus in Toronto
on
Great Surplus Stores?
·
· Score: 5, Informative
Actually, they're pretty good for components - the problem is you have to saddle up to the front counter and ask. They're also very good about looking up equivalents for you and suggesting stores that will have what you are looking for.
I have been going there for over thirty years and I've noted a few things that people new to the store should be aware of to make their experience a positive one:
Go there often - at least once a month. Make sure the guys behind the counter know your face by asking not stupid questions. A "not stupid" question is "Where can I find geared motors?", not "what does a resistor do?". They don't expect their customers to know everything, but there is a certain basic level of knowledge and intelligence expected before somebody can walk in the doors.
As part of the previous point, accept that everything there is "ASIS" unless they tell you otherwise. Asking what "ASIS" means will get an explanation geared for somebody with an IQ less than 50 that is delivered at a LOUD volume with questions afterward to make sure you understand what you have been told. Nothing pisses them off more than somebody bringing back an old cassette drive transport, demanding their money back because a pinion gear is missing inside it. They'll generally let (and help) you test stuff if it's reasonable to do and there isn't a huge lineup at the counter.
Don't ask complex questions between 11:30 AM and 2:00 PM. A "complex" question is anything that requires an answer longer than "Aisle 2". This is their busiest time and if they're answering your questions they're not taking in money and that pisses them off.
If you're buying something and you have people behind you, have your money/cards ready. Watching you fumble with your wallet pisses them off.
If you are looking for something and they don't seem to have it, ask when they'll have it in. There's always a good chance that it's upstairs or they're expecting a shipment sometime in the future.
Despite what they tell you, they're always getting in new stock. If you see something you like, but the only ones they have are incomplete or apparently damaged DON'T BUY IT! - they'll probably have more in stock in a few weeks or they'll pull more down from upstairs when the current stock is sold.
Learn Polish, Ukrainian or Russian. At the very least it's entertaining to listen to what they think about their customers (I caught them saying that they wished a certain customer, who was no better than a misbegotten dog (literal Ukrainian translation), and all their descendents would be hit by a bright bolt of lighting. This is a mighty powerful curse.)
I've worked a bit with memory and memory tests and there are a couple of things that I can add here.
I've always heard of this practice as being called "Yield Management" and is very common. I don't know why the article doesn't use this term.
It is very prevalent - and not only in memory. There are a lot different devices that are shipped with known problems - chip designers often design in redundancies that can be activated when the chip is wired to the lead frame.
The best memory test there is is the operation of a (true) operating system. A deterministic test just does not provide the randomness required to bring out errors. This is why all manufacturers test PCs/Systems running the operating system before shipping them to customers.
One of the reasons why SDRAM was adopted so quickly and aggresively by PC manufacturers (along with processors with large caches) is the block reads and writes, which like deterministic tests, results in much lower failure rates than the random reads and writes you would get if the processor accessed the memory directly. SDRAM and processor caches created an improvement in system reliability without increased testing.
Your statement "If the chip is half-bad, there are a good chance that it has defects in the other half." is inaccurate. If it were true, then in the case where one chip had a defect you would have to assume that all the chips on the wafer are bad (because what is a defect in one half of a chip is the same distance to features on another chip, the only difference being is where the wafer was diced). Carrying this on to its logical conclusion, then if there was one defect in the silicon ingot, then no wafers or chips could be manufactured from it.
The company (Celestica) has invested literally tens of millions of dollars in SDRAM test including coming up with pseudo-random patterns that match the reads/writes of an operating system.
This is typical of all school computer programs that I have seen - some multi-media, some text processing and some instruction on using the internet for looking up information. This is based on what I see my kids doing with computers at school.
I wouldn't be surprised if the laptops became a tool for text messaging as well as downloading and sharing MP3s and pr0n. In the end, I wouldn't be surprised if the big winners turned out to be the RIAA's lawyers...
myke
Fans now and then
on
Ask Larry Niven
·
· Score: 2, Interesting
Hi Larry,
I've been a fan for a long time (I have a first printing of "Ringworld" that I'd like you to sign some day).
While being a fan, I don't consider myself a rabid fan (ie I don't dress up as a Kzinti or make scale models of Dream Park out of Lego and I have never considered a career as an organlegger), but over the years I've been amazed at the lengths fans will go to emulate and understand the characters and locations in your books. One of the classic examples this is the efforts to understand how strong scrith would have to be. It seems like your fan base, while somewhat smaller, is at least as hardcore as the "Trekkie" (or "Trekker", for purist) culture and has been so for thirty years or more.
I'm curious to find out if you think your fan base has changed over the years because of the Internet (ie getting larger, smaller, more or less extreme). Do you find that personal information (address, phone number, email) has become more difficult to keep private from (cyber) stalkers and other people that have a need to tell you how great your work is?
The interesting thing is when you go the other way down the airport - you get the little appreciated "ad-ood-eed-a-piz" that nearly destroyed Julie Andrews' career.
Slashdot Mirror
What I don't understand is why don't they don't propose using the traditional Lattitude/Longitude with minutes and seconds. Assuming three digits for Lattitude/Longitude and two each for minutes and seconds, you need 14 digits.
This translates to 7 bytes using BCD and should get comparable accuracy to what is being proposed. AND, it would be immediately compatible with GPS systems already available.
myke
Wasn't the tracking/selling of viewer habits something that Tivo promised not to do when it first came out? I seem to remember this was an issue when Tivo first came out (along with its need for a phone line to download programming information, there were concerns it would phone back out and pass along personal information).
I love the last quote from Martin Yudkovitz: We want to embrace the opportunity to help advertisers and broadcasters better understand the current and future impact of DVR so they can adapt content and advertising strategies to this new medium. If they want to help advertisers and broadcasters, why are they selling the information?
Of course it might not be a bad thing; if they base the advertising that's passed to me on which commercials I don't fast forward through, it will be 24-7 of Victoria's Secret commercials - Party at my house, dudes!
myke
It's funny reading this article because I can think of a number that came out a dozen years or so ago with exactly the same theme; games were all the same, there was nothing new, game developers were rehashing the same old concepts over and over again...
And then came Wolfenstein and suddenly games weren't boring any more and people couldn't get enough FPS. When things got boring again, out came "The Sims"...
I have no doubt that the pattern will repeat itself and somebody out there has a great new concept for a game and it will be "discovered" now that people consider the current ones boring and manufacturers are looking for new concepts to exploit.
myke
at my work cafeteria. The manager of the cafeteria said that by doing this, "I made the cashiers feel stupid."
I tried pointing out that this was grade 3 math and it made it easier for the cashier because they would have to return less change. The advantage to me was there was less shrapnel in my pocket.
The upside was that I could go out whenever I wanted because the cafeteria refused to serve me. (See the previous article about dodge work).
myke
I'm sure that digital technology will help out with some of the current issues but what about the situation in which the cop and the offender are out of the frame? I've seen a number of tapes where the suspect has been taken out of the frame and has later sued the police for brutality and/or violating their rights. Ideally, the camera would follow the action, even if it was not directly in front of the car.
myke
I've discussed this question with a hospital ethicist - the 8 year old would almost always get the liver. But, for practical purposes the question is mostly hypothetical.
The decision is usually made based on the simple reason that in the case of the alcoholic, there is generally a lot more wrong with them than just the liver, replacing the liver will extend their lives but not dramatically. You're talking two years or so at the most.
The little girl would get a lot more use out of the liver.
There is another question that comes into play and that is whether or not the alcholism will negate the success of the transplant. Heavy drinking will attenuate the effectiveness of the anti-rejection drugs as well as impair the patients ability to follow a medication schedule.
Right now, there is a push to make sure organs are passed to the "best" match, not the next "acceptable" match in line. The better the match, the higher the probability that the transplant will be successful.
In the case of livers, as I understand it, very few transplants are now done from unrelated donors. Partial liver transplants from relatives seems to be the preferred method with organs harvested from donors being used rarely.
myke
I don't know if I would call it "joy"...
I would use the term "satisfaction"; I remember being very pleased with myself for coming up with a sixteen bit by sixteen bit shifting multiplication algorithm in just 24 instructions or realizing that self-modifying code is needed to perform an indirect jump or data access anywhere in the 16 bit address space.
If you can competently program the 6502 in assembler, then you will be able to tackle any other 8 bit processor architecture with ease - a very satisfying accomplishment.
myke
Now having said all this, I'm not surprised, I've been reading performance comparisons for 25 years and strangely enough, the sponsoring company's hardware/software/operating systems always seem to come out on top. This started with comparing the 8086 to the 68000 and has continued on to the present day.
The important/best thing about the review is that it states very clearly at the top that the test was sponsored by Microsoft.
myke
Maybe I'm showing my age, but AV was the best dodge you could have in high school.
You asked, where is the fun carrying around a TV? It is infinitely more interesting than many things that you would have to do if you weren't carrying around a TV (ie sitting in a health class looking at pictures of cholesterol blocked arteries).
Want to miss a class? Come in five minutes before the final bell, apologizing and mutter under your breath about the splicing job you just had to do.
Splicing is great. In the '70s, it could mean film, (audio) tape, power cords, speaker cords or video co-ax.
AV means never being called on the carpet for the classes you cut AND they make allowances for the material you missed when its test time.
myke
Thanx for the pointer,
myke
This is a follow up to my previous response asking who read over the letter (because anybody halfway competent could have figured out what was done even if the web mirrors were taken out).
Hiding behind DMCA is clearly not the answer. It destroys credibility for the company and ultimately for the law itself. If these cards are used for purchasing products/services than potentially millions of dollars are at risk - nobody can realistically expect a simple law to protect against theft of tangible items as well as theft of intellectual property.
But, how much does a letter from a lawyer cost? From the company's perspective, this is the most cost effective way to deal with the problem.
The second most cost effective way of dealing with the problem is to hire two guys to show Virgil and Acidus the errors in their ways (preferably from the perspective of the inside of a Cadillac's trunk).
I would suspect that this would be a bit more expensive than a lawyer's letter, but almost certainly more effective.
If either of the first two methods don't work, then Blackboard could, gasp, fix their product. I disagree that the encryption couldn't be built into the individual boxes - there are some very clever things that can be done that would make figuring out the communications very difficult.
This is the most expensive option, but the one that would have the least amount of liability for Blackboard and could establish them as the brand owner (get rid of all those pesky ATT labeled boxes).
Comments?
myke
Reading through the C&D letter, I have to wonder who approved it from Blackboard's perspective and if anybody technical thought through what may be the result of it is.
.ru or .iq) and finally 4) Blackboard's reps get innundated with phone calls, emails and letters complaining that their system is not secure.
There sounds like there is enough information in the letter so that somebody that knows what a 75176 is (I would disagree with the assertions in the paper about RS-485's obscurity), can program a PIC or an 8051 and can use an oscilloscope can reproduce the work done by Messrs. Griffith and Hoffman. Along with this it sounds like the readers are connected to standard cabling via standard connectors.
So, the result I would expect from this letter is, 1) it will be put on the Internet for all to read, 2) boxes throughout the different colleges and universities that use the system will be pulled out of walls and vending machines with many of them stolen or vandalized to see what's actually inside them, next 3) The protocol and hardware will be distributed on a variety of web sites (probably ending with
This begs the question on what Blackboard should have done. (next reply).
myke
Where does the heat go?
This seems like a nice technology to remove the heat from the CPU, but what I'm always wondering about is, where will the heat actually be dissipated into the environment? At some point, there has to be a heat exchanger where all this heat collected in the tiny bubbles is passed outside the unit. This is going to take a fair amount of space - one of these days we're going to see ads for heat exchangers that take up less space than the "standard" box available from Intel.
I'm looking forward to a Beowolf cluster not only performing amazing calculations but also heating the building it's in.
myke
When did the English pound become "real money"?
myke
When I read this, I immediately thought of these two articles that were just posted: Too Much Free Software and Linux Audio Development.
Wouldn't the Linux/OSS advancement efforts be better served if maybe this team of people had worked for a week and released a complete "Duke Nukem 3D" for Linux? The problems listed above (including the need for an MS-DOS non-shareware version of the game) will probably tend to turn people off and not think that Linux is ready for the prime-time desktop.
I would think that this effort would be a better advertisement for Free Software/Linux if shareware data files could be processed, networking/modem/multiplayer modes worked along with correct audio and mouse/joystick support. I would also think that an RPM of the application should be available as well.
This should be much less effort than porting the package to Windows or getting demos working but would be impressive enough that casual computer users would consider Linux for their desktops.
myke
As other posters have indicated, reflowing solder paste in a toaster oven has been done for years and simulates the standard process for SMT assembly.
Using Solder Paste, you can use a hot air gun to place components on a PCB as well. A woman I know at work (Celestica) made a video demonstrating the SMT solder process using a hot air gun - it came out quite nicely and her joke on customers was saying that she followed the board through the oven (it was vapor phase at the time). Many customers were impressed with her tolerance to extreme heat.
As somebody noted, most components will stop working at 140C and in the oven they will go over 200C - they will survive, but the PCB should not be powered up until the PCB has cooled to room temperature.
If you're going to try this at home, a few comments:
1. Solder paste will only stay reliably sticky for 30 minutes. Make sure that you have your components ready and the oven at the primary temperature before you break the seals on the syringe and start applying paste. Make sure that you don't have more components than you can place in 20 minutes.
2. Solder "paste" is made up of finely ground solder held together by flux. Both the solder and the flux are poisenous and during the solder process you will see a build up of flux on the inside of the oven. Along with this, the solder may form "balls" that can be thrown off the PCB. An oven used for SMT experiments cannot be used for food preparation afterwards.
3. Solder paste must be capped and refridgerated when not in use. If you are storing it in a fridge where food is stored, make sure that it is in something like a tupperwear container and well marked (especially if children are around). It looks like pate or liverwurst, but will sit in your stomach like a ball of lead (sorry, couldn't resist).
4. The PCB coming out of the oven is very hot and will take several minutes to cool down. I've heard of a number of people that have built SMT boards in a toaster oven, only to forget oven mitts and tongs to handle the hot PCBs and ended up dropping the PCB on the floor and burning themselves. One genius I heard about was sitting down when he pulled the board out of the oven without any mitts or tongs... Make sure you have something like a barbeque grill ready for the PCB to sit on when it comes out of the over.
5. The PCB should be as dry as possible. Before putting on paste/components, you might want to put it into the over for a day or so at the lowest setting to try and bake out any water that has gotten trapped in it. Let the PCB cool before applying paste.
6. The PCB pads should be "HASL" ("Hot Air Solder Leveled") for best results (do not try this on bare copper and you may have to experiment with gold finishes).
7. I would suggest using parts with leads on 0.050" (50 mil) centers, 0805 chip components and SOT-23 transistor and diodes. Anything smaller will make applying the solder very difficult. The article indicates the author used smaller spacing components, but not how many and how the PCB was laid out.
8. Do not use surface mount connectors. Unless you are very comfortable with doing your own soldering, you will find that it is difficult to get a uniformly strong joint on every pin.
9. If you are designing your own PCB, you can use Protel's "EasyTrax", which is an MS-DOS Command Line program that can be downloaded for free from a variety of sources (you should be able to find where on Google). I have added IPC standard pad layouts for the library components.
I've done it a couple of times with an old toaster oven and it works surprisingly well. Just make sure you plan out what you are going to do and if there are any terms that I have used above that you are unfamilia with, make sure that you investigate them before trying it out on your own.
myke
I would have thought that the major problem next year is that Russia would be a bigger energy supplier than Saudi Arabia...
myke
Reading the memo, figuring out whether the policy change is for hardware or software is somewhat unclear.
Like many other people, I have had to move up to Win/XP and I have found a lot of my hardware/software no longer works, causing me to have to buy replacement products that are "designed for Win/XP". I'm sure for a retailer, this is a major headache as inventory comes back and customers gripe over having to buy replacements.
The question I have to ask is, why is Office Depot sending the memo out to a whole bunch of suppliers when they should only be sending out one? Doesn't it make a lot more sense to send a memo to Microsoft saying that they have found that their new operating system has caused a lot of distress for their customers and rather than requiring their customers and other vendors conform to Win/XP could Microsoft please fix the problems with the operating system so that legacy products will work? In the meantime, Office Depot will cancel all pending orders of Windows/XP until the bugs in it are fixed.
myke
I wouldn't characterize them as being like the Seinfeld "Soup Nazi", it would be more accurate to say they aren't very patient with stupidity.
Personally, I think if you showed up at 12:15 PM, when there are a couple of dozen people at the single front counter and held up an ancient "Princess Phone" asking if they have the rotary dial version in purple and then ask how long the warranty they give on it is, then you deserve to be called a throwback (or worse) in a slavic language.
Common sense would indicate that the example above would be a dumb thing to do, but I can't tell you how many times I have seen people do things just like the phone example (and at other stores to be fair) - at Active Surplus you just don't get responses to these kind of questions with the kind, patient, medicated voice you would expect from The Gap.
myke
I have been going there for over thirty years and I've noted a few things that people new to the store should be aware of to make their experience a positive one:
- Go there often - at least once a month. Make sure the guys behind the counter know your face by asking not stupid questions. A "not stupid" question is "Where can I find geared motors?", not "what does a resistor do?". They don't expect their customers to know everything, but there is a certain basic level of knowledge and intelligence expected before somebody can walk in the doors.
- As part of the previous point, accept that everything there is "ASIS" unless they tell you otherwise. Asking what "ASIS" means will get an explanation geared for somebody with an IQ less than 50 that is delivered at a LOUD volume with questions afterward to make sure you understand what you have been told. Nothing pisses them off more than somebody bringing back an old cassette drive transport, demanding their money back because a pinion gear is missing inside it. They'll generally let (and help) you test stuff if it's reasonable to do and there isn't a huge lineup at the counter.
- Don't ask complex questions between 11:30 AM and 2:00 PM. A "complex" question is anything that requires an answer longer than "Aisle 2". This is their busiest time and if they're answering your questions they're not taking in money and that pisses them off.
- If you're buying something and you have people behind you, have your money/cards ready. Watching you fumble with your wallet pisses them off.
- If you are looking for something and they don't seem to have it, ask when they'll have it in. There's always a good chance that it's upstairs or they're expecting a shipment sometime in the future.
- Despite what they tell you, they're always getting in new stock. If you see something you like, but the only ones they have are incomplete or apparently damaged DON'T BUY IT! - they'll probably have more in stock in a few weeks or they'll pull more down from upstairs when the current stock is sold.
- Learn Polish, Ukrainian or Russian. At the very least it's entertaining to listen to what they think about their customers (I caught them saying that they wished a certain customer, who was no better than a misbegotten dog (literal Ukrainian translation), and all their descendents would be hit by a bright bolt of lighting. This is a mighty powerful curse.)
The simple rule is, don't piss them off.myke
- I've always heard of this practice as being called "Yield Management" and is very common. I don't know why the article doesn't use this term.
- It is very prevalent - and not only in memory. There are a lot different devices that are shipped with known problems - chip designers often design in redundancies that can be activated when the chip is wired to the lead frame.
- The best memory test there is is the operation of a (true) operating system. A deterministic test just does not provide the randomness required to bring out errors. This is why all manufacturers test PCs/Systems running the operating system before shipping them to customers.
- One of the reasons why SDRAM was adopted so quickly and aggresively by PC manufacturers (along with processors with large caches) is the block reads and writes, which like deterministic tests, results in much lower failure rates than the random reads and writes you would get if the processor accessed the memory directly. SDRAM and processor caches created an improvement in system reliability without increased testing.
- Your statement "If the chip is half-bad, there are a good chance that it has defects in the other half." is inaccurate. If it were true, then in the case where one chip had a defect you would have to assume that all the chips on the wafer are bad (because what is a defect in one half of a chip is the same distance to features on another chip, the only difference being is where the wafer was diced). Carrying this on to its logical conclusion, then if there was one defect in the silicon ingot, then no wafers or chips could be manufactured from it.
The company (Celestica) has invested literally tens of millions of dollars in SDRAM test including coming up with pseudo-random patterns that match the reads/writes of an operating system.myke
This is typical of all school computer programs that I have seen - some multi-media, some text processing and some instruction on using the internet for looking up information. This is based on what I see my kids doing with computers at school.
I wouldn't be surprised if the laptops became a tool for text messaging as well as downloading and sharing MP3s and pr0n. In the end, I wouldn't be surprised if the big winners turned out to be the RIAA's lawyers...
myke
Hi Larry,
I've been a fan for a long time (I have a first printing of "Ringworld" that I'd like you to sign some day).
While being a fan, I don't consider myself a rabid fan (ie I don't dress up as a Kzinti or make scale models of Dream Park out of Lego and I have never considered a career as an organlegger), but over the years I've been amazed at the lengths fans will go to emulate and understand the characters and locations in your books. One of the classic examples this is the efforts to understand how strong scrith would have to be. It seems like your fan base, while somewhat smaller, is at least as hardcore as the "Trekkie" (or "Trekker", for purist) culture and has been so for thirty years or more.
I'm curious to find out if you think your fan base has changed over the years because of the Internet (ie getting larger, smaller, more or less extreme). Do you find that personal information (address, phone number, email) has become more difficult to keep private from (cyber) stalkers and other people that have a need to tell you how great your work is?
A loyal, but not weird, fan,
myke
Hi Larry,
When will the final part to "The Integral Trees" trilogy be out?
myke
The interesting thing is when you go the other way down the airport - you get the little appreciated "ad-ood-eed-a-piz" that nearly destroyed Julie Andrews' career.