I'd love a Diaspora update, too, but in the last three months I've seen precisely zero activity there. Could just be my friends... but there is an uneasy balance between attracting/retaining enough people to get networking effects, and still ensuring their privacy. One thing I like about Google+ is it starts out with limited posting and limited circles and you have to work to reduce your privacy. It appears to me that Facebook has worked in exactly the opposite manner: you have to work to block and filter groups out. (I may be wrong: I used facebook briefly about three years ago.) When Google launched Buzz, they did something like opt-out rather than opt-in: Buzz included everyone on Google. That seemed to piss off a *lot* of people. I could understand why they did it: it got lots of people involved. But the bad press it got seemed like it overwhelmed any possible benefits of network effects. This time Google looks like they're trying something closer to Diaspora and hoping that because it's Google it'll be used by people other than privacy-obsessed techies (who are, or at least claim to be, actively repulsed by networking effects.)
One other interesting effect I've seen is that Buzz, which was getting more interesting over time, has really dropped off since G+ launched, because everyone who was on Buzz is now spending their effort on G+, and there are *lots* of new people on G+ as well. It'd be kind of sad if that was the main competitor that G+ killed.
Would you rather be t-boned by an idiot driver who runs a stop sign or hit in a glancing blow by an idiot driver who can't navigate a roundabout? A good roundabout where the curbing forces tangential entry is safer.
As a cyclist I'd much rather deal with someone not yielding in a roundabout than someone running a stoplight. A friend got hit by someone running a red light last week and he just got out of the hospital yesterday. (Also people tend to pull to the inside of a turn, so roundabouts leave extra room for bicycles.)
We *have* come up with good UI's: you just pay through the nose for them.
My favorite nice intertool integration feature of Cadence is that you can open schematic in one window on one monitor, layout on the other monitor, put layout in 'move part' mode, then click on a part in schematic and mouse over to layout and that part is stuck to the cursor, ready to be placed. When you're doing huge boards with repetitive layouts that's an unbelievable time-saver. On Altium, my favorite intertool integration feature is that I can change the definition of a part (symbol or footprint) in the library files, then click on it and hit 'update' and the tool will find all occurrences of that part in open documents, and update them, so (for instance) increasing drilled hole and annular ring size on a DIP's legs to make it fit the drill I have rather than the one it was modelled with is a thirty second operation. Both Altium and OrCAD have spreadsheet-driven setups, which I also love, because I can write an open office macro to fill a pin grid array numerically and with x/y coordinates, then open the spreadsheet file in the layout/librarian tool and import a mostly-finished footprint. All I have to do is add the silkscreen/keepouts.
At work I use Altium, which is fairly easy to learn but has some weird bugs, draconian licensing, and breathtaking price, and Cadence, which is *not* easy to learn, has the most convoluted, cumbersome, and wretched new-parts-creation system I've ever seen, exostratospheric price, but the best PCB layout I've ever used. It's just dreamy. I also occasionally use OrCAD, which is quite reasonable if not outstanding on all fronts: schematic, layout, and parts creation.
At home I use gEDA. Schematic works fine, parts creation is pretty straightforward, layout is kludgy and not pleasant. (But newer revs have a gcode exporter built in, so it can go straight to a mill, which is nice.) I've used Eagle a little and it's pretty reasonable, although yeah I'm frustrated by the limitations. A friend and I are starting up a little company, and we use Eagle for small easy boards, and gEDA for larger more complex boards because it doesn't have those limits. I'm writing some scripts that I hope will allow me to import Eagle parts into gEDA formats (although I've written a bunch of scripts for gEDA already so it's so easy to make symbols it's not worth importing them, and it's fairly easy making footprints.)
I just downloaded KiCAD but haven't had a chance to try it.
I've also used PADS, which is okay but I feel like it's primitive for the price.
Sparkfun.com also has a cheap board fab service if you have long lead times, but it hasn't been competitive with myropcb.com for me, but that may be because I'm running production lots rather than one-at-a-time boards. (I cut those on the mill at work, and until recently was doing those via chemical etching at home, although now I have a mill at home.) I'll check out Seeed's prices/capabilities.
I used to but don't anymore, it doesn't appear. I read your other post in this article and your company will be seeing an RFQ next time I have a big ugly piece of test hardware to get fabricated. Specific questions: do you do conductive fill vias? do you do laser (or whatever technology) blind microvias on the 0.008" size range? and how small a drilled via can you put through a 0.187" thick board? I'm jbump at front range internet.
I design PCB's for a living, these days. Most board shops I work with have yearly tours/open houses: if you want to see an up-to-date shop see if they're throwing one. It's pretty cool to see. I'm mostly impressed by the electrical test machines: they look like a dozen mechatronic herons madly going after fish.
When I can wait a bit, I use myropcb because if I'm ordering 200 boards the size of postage stamps they're less than a dollar a piece including soldermask and silkscreen on both sides, if I can wait 10 days. (It'd be a lot faster but they tend to go slowly through US customs.) If I'm willing to pay a bit more, I use Circuits West, who will crank out up to 60 square inch boards for $31, and have had great quality.
However, the really great thing about milled boards is the turn time -- if you have a mill. I regularly go from hastily drawn schematic to finished, working board in under two hours, if it's a simple design. We can do three revisions of a board in a day, and *then* send it out to get a green board, once we're sure we have something working, and have a tested design ready for large-scale production in three days. We *love* having a PCB plotter in-house. It takes some thinking and experience to lay out good boards for it, but it sure helps productivity.
While I'm shilling companies that have saved me in crises, Vector Fabrication is not the cheapest place to get PCB's, but they'll produce a 30x30 cm 14 layer board with 3 ounce copper in two days.
I've got a passing interest in glasswork, and one of the things I learned is that it's more complicated than "melt into mold, let it cool". Glass has to go through a carefully controlled cool-down period so that the molecular structure will set up properly. Otherwise, the resulting object is far more brittle than it should be. If not done properly you can have cracks form during the cooling phase, ruining the object.
Does the incremental deposition solve the annealing problem? Being able to make glass objects without having to carefully control the cool-down would be very nice.
I was a glassblower and glass bead artist for a while. Careful cooling is pretty essential for lime glass, which is what we mostly use. It's less important for borosilicates like Pyrex, which is why glass casserole dishes can survive being put onto 200C metal racks in the oven, and it's even less important for fused quartz that's straight silicon dioxide. You can stick a pyrex rod that's less than a centimeter in diameter straight into an oxypropane flame without it splitting or snapping, and I believe you can do the same with a 3 or 4 cm quartz rod. Obviously this stuff isn't pure silicon dioxide, but it's closer to SiO2 than it is to lime glass.
Incremental deposition probably won't solve the annealing problem, but it'll change it: instead of having strain across big areas, you'll have little bits of strain distributed between each layer of glass that's put on so you're liable to get a lot of small cracks through the porous material, rather than one big catastrophic crack. However, all those little cracks generally tend to grow, but that may be somewhat helped by it being an amorphous, impure material: it's harder for cracks to run in long straight lines in crappy heterogenous stuff.
Given the whole square/cube thing about heat-producing volume vs heat-radiating surface area, it wouldn't surprise me if feathers developed as a way to keep juvenile dinos warm until they grew large enough that they didn't need them. Smaller dinos might have needed them all the time, but T. Rex "chicks" would have shed when they got larger.
You could be right. Many mammals -- cats and dogs, certainly -- are born essentially cold-blooded and will freeze to death if taken away from their mother and siblings, and don't start self-regulating their temperature for a day or so.
The thing with feathers being evolved to deal with rain was based on the observation by biologists studying hypothermia and different hibernation strategies, that it's pretty easy to stay warm if you're dry, and fairly easy to stay warm if it's snowing, but cold rain is extremely difficult to deal with. Huge feather-like structures can act like a raincoat. Look at the primary flight feathers on geese: massive close-knit waterproof feathers that when folded cover almost their whole backs, with little insulation value but great rain-shedding value. Likewise, their fine feathers on the rest of their bodies provide good insulation, with only some water resistance. When the bird is hot, it can stretch out its wings and expose its nearly unfeathered armpits to dump heat to the environment. Development of scales into insulating and water-shedding feathers could be an entire thermoregulation strategy, and flight an unexpected but very welcome by-product.
The point about body mass, if it's not clear, is that metabolism produces huge amounts of heat. When we digest food, about 50% of the energy that's released in converting complex molecules to carbon dioxide and water is released as heat. A huge animal can keep warm just from that. Mammals and birds maintain their temperature in a very narrow range. However, it's more complicated than just that would indicate: hibernating mammals (and one species of bird that hibernates) allow their temperature to fluctuate with the outside temperature. Likewise, there are reptiles that do some things to reduce their temperature variation, by seeking sunlight or shade, which is a type of active regulation. One current theory about dinosaurs and the evolution of feathers is that they showed up primarily as a thermoregulation system, providing insulation (particularly in rain) but allowing the animals to fluff their feathers to increase temperature losses to again actively thermoregulate.
It didn't have a radio. (That I saw: I was looking over his shoulder.) The main reason he was so into it was precisely because he could fix everything on it -- and he did, too. He had a lot to say about why people shouldn't own anything they couldn't fix
With OBDII, the electronics in a modern car aren't that bad to figure out.
Sure, I couldn't probably make the average car sensor from scratch, but I couldn't make a '69 VW Beetle's brake drum from a block of steel.
Not that this is particularly relevant but my car -- a 2002 Subaru -- has a whole raft of undocumented OBDII stuff. Sure, it has the standard data, but a bunch of information like shift data seems to be only accessible if you have a Subaru diagnostics system, which sort of sucks.
Brake drums are cast out of cast iron. Doing cast iron has been too exciting for me: I stick to cast aluminum (and sometimes, if there's a problem, bits of cast aluminum stick to me, hence my unwillingness to try iron) but machining a brake drum from a casting isn't all that difficult. Now, casting a vented brake rotor: that's a challenge. There are plenty of parts that are pretty infeasible to try to make at home, like a ground cam or a nitrided crankshaft, or even the excitement inherent in casting the flammable magnesium engine block on a VW, but those are all several orders of magnitude easier than making your own electronic engine control system from silicon.
One of my coworkers spent several hours chatting with Bob Pease about making fuel injection systems, last time Bob was visiting here, as it happens. My coworker had just finished adding one to his 1965 Jeep, that he'd machined most of the hardware for (although he'd purchased the fuel pump and injectors) and Bob was interested in some of the technical details of how it managed fuel/air ratios.
True in some places in Colorado things like rain barrels are illegal. Which doesn't make much sense to me at all but there you go.
It's not that they're illegal per se. It's that in arid states, water rights are more than 100% used, and first in time of use gets first priority, meaning the people who have water rights older than yours, get to use their entire allotment of water before you get to use any. Collecting rainwater means someone else doesn't get their full allotment. As it turns out, California is one of the major water rights holders, so it's not because the guy down the street is complaining that you can't retain the water that would otherwise run off your land into the gutter: it's because the likes of California and Arizona have won large numbers of SCOTUS rulings upholding their water rights. (I've been told that water rights cases are the largest category of cases to go to the Supreme Court, but don't have a citation.)
And not being able to collect water in a rainbarrel is by no means the weirdest consequence of this: people who live near irrigation ditches or rivers have been required to cut down trees, because the trees are sucking up groundwater that is being replenished by leakage into the water table from the canal or river.
Water rights is a huge facet of local and state governments in arid states.
The capitol building of Denver has the dome covered with 200 ounces of 24 carat gold, and it needs replacing about every 40 years. That means it's losing several ounces per year, and most all of it is coming off in rainwater that ends up dumping through the drainage spouts down the sides. People tried to collect it, like this guy. I've been told, but can't find an online reference, that collecting rainwater from the Capitol Building was made illegal to prevent mayhem, so now it goes down the sewer.
It didn't have a radio. (That I saw: I was looking over his shoulder.) The main reason he was so into it was precisely because he could fix everything on it -- and he did, too. He had a lot to say about why people shouldn't own anything they couldn't fix, and about how nice it was to be able to walk down to the corner auto shop and get most all the parts he needed to repair or replace anything on the Beetle.
And I want Apple to defend it with all the power it has... So that only Apple devices are blocked and all other devices are unaffected.
As if. Apple will get this, then they'll license it for an enormous sum to all the other cellphone companies when the Apple and police lobbies manage to get mandatory implementation of this enshrined in law.
For what it's worth, there are already several open standards for lighting communication, including 0-10V and DALI, and both have reasonably wide adoption. The consumer marketplace could well be controlled enough that a major manufacturer might be able to push a protected, patented, proprietary interface and give it some goofy name like Y11, but in the industrial lighting space -- and that includes outdoor, sign, and indoor in factories, warehouses, malls, government buildings, and sometimes even apartment complexes -- the market is fragmented and people would rather design things that work with existing control systems than try to talk customers into buying complete systems. I know a bit about this since I've been designing a lot of DALI transceivers lately. They're pretty cool: noise-resistant, multi-topology, and polarity-agnostic so you can string them up beside industrial 440VAC and hook the wires together any old way you want and they'll still work.
There's also a lot of work going on in mesh networking, but that's nowhere near as standardized, as far as I can tell. As for the 'forgot the phone' thing, if this is being aimed at consumers that's merely a happy side-effect: the main point is so building management companies can have fine-grained control over dimming, and lighting that reports back when it's dying or dead to minimize maintenance costs. It'll be centrally controlled so forgetting phones won't be a problem. Most of the stuff that's in the pipeline does ambient light sensing, motion sensing, heat sensing, and sound sensing, and likely RFID, so it can figure out if people need lights on even though they're still (in the bathroom) or if they need lights turned off (in the bedroom), or if lights should turn on well in advance of someone approaching (as they cross a parking lot or walk up to the front door) or if it can come on milliseconds before the person gets to the place where light's needed (hallways).
Read up on the Bonus Army riots of 1932 and the background/history of US compensation to armed forces members. Your experience is historically speaking an anomaly. I'm glad the government is treating veterans well these days, but it appears that such treatment is rare.
There was an interesting article in the New Yorker about two weeks ago, detailing the science behind the attempts to grow artificial meat. Unfortunately that link only gives an abstract and a small section, but it's an interesting read if you wander down to the library and check it out. So far, the state of the art is a piece of artificial meat about 14mm long, 3mm wide, and maybe a couple cells thick, and even then it's just muscle cells without any actin/myosin structure so it'd be like, I dunno, moosh. But they're making progress.
There's no shortage of psychos and crazies. What made Kaczinsky interesting was that he was a very smart psycho -- but even then, there are lots of those. All the interesting stuff he wrote is public record because it was used as evidence during the trial and published in the court proceedings. Ideally they'd sell all the stuff and use the proceeds to help repay medical costs/give disability benefits to the people he injured, but more likely they're going to use it to indirectly defray some of the cost of the trial and his ongoing incarceration.
If you want a cheap radiation source, you can buy 2% thoriated tungsten rods for TIG welding. Find a local store and buy a 1/16" rod: individually they only cost about $8. Other people have suggested Coleman lantern mantles but the ones you can buy these days don't seem to have thorium in them anymore, because the old ones were *seriously* radioactive. If you *have* a Geiger counter you can go to an Army/Navy Surplus store and check the ones they have since a lot of old radioactive ones are still in stock.
My homebuilt geiger counter, using a surplus Russian GM tube, can easily detect a single thoriated tungsten rod if held up close to the tube, as can my vintage Civil Defense CDV700. Both will also detect a smoke detector.
If you want to build your own geiger counter and have a tube, here are instructions for building a high voltage power supply from a hacked-up flash unit from a disposable camera and here is the detector circuit that translates that into audible clicks. If you optoisolate that detector circuit you can feed it into an Arduino and log/display counts per second on a laptop. (It needs optoisolation because the output of the audio click board is negative with respect to power and way more than 5V, so it'll cook an Arduino, as I found out. Although an Arduino analog input can withstand -200V and still function, amazingly enough.)
If you just want to detect ionizing radiation, you can build an ionization chamber. My company supplies DIY kits but we also have detailed instructions for making your own with a component list of like four transistors and a handful of resistors, and a tin can. They're more sensitive than a Geiger tube, although they're much slower to react, taking seconds to change their reading. One neat thing is you can build them as chambers so you can actually put a sample inside the chamber, if you want, and they detect alpha, beta, and gamma.
And as other people have said, any sample you buy that'll allow you to characterize your radiation detector will expose you to tens to thousands of times as much radiation as anything in Japan unless you're actually inside the grounds of the power plant, so this whole project might not do what you want.
None. Ever. Even the underpants bomber made it through.
Since the inception of the TSA, they have stopped *zero* hijacking/bombing attempts from the airport.
Biggest waste of money on security theater going.
-- BMO
They're not trying to stop bombing attempts. They're trying to stop their employers getting reamed by the media when a bombing attempt is successful. "But we x-rayed everyone and groped *babies* -- what more could we do?" is probably the best defense the likes of the FAA, FBI, and TSA have available to them the next time someone blows something up and the media is all "but look at all the WARNING SIGNS: how could you be so dense as to not notice this one dude out of 300 million other dudes who did four things that in hindsight make it clear he was building bombs but at the time looked like relatively innocuous actions?" The whole TSA is a giant exercise in ass-covering. Stopping actual terrorists is a slight side-effect.
What I mean is the human body has easily enough power to raise itself up a vertical ladder or rock face so presumably a huge amount of this power must be lost just moving air around when that power is used inside a human powered helicopter. But how much power is wasted , or to put it another way , how much power put into the system is actually used to raise the mass of the helicopter?
I'm not exactly sure what you're asking here, but think about it this way. Gravity is accelerating you downwards. When you stand on a ladder, the ladder's structure is resisting your weight, which is why you stay up. With a helicopter or airplane, you don't have a structure to hold you up, so instead you accelerate air downwards using wings. You push the wings through the air, they accelerate air downwards, you get supported.
So the easiest answer to your question is *all* the power put into the system actually raises the mass of the helicopter. A slightly more nuanced answer would be that the mass of the whole system (pilot + helicopter) minus the pilot's weight, is how much energy it takes to raise the mass of the helicopter, but that's sort of silly.
A side-note: there's something called ground effect that changes how helicopters (and airplanes and anything else relying on accelerating air downwards to maintain flight) work. When a wing is within about one wingspan of the ground, the air it forces downwards is somewhat constrained between the ground and the wing, giving the wing more lift than it would in free air. As a result, if you're within a wingspan of the ground it requires significantly less power to stay in the air. (If you watch airplanes land you'll see this effect as a change in their angle of approach just before they land.) The Sikorsky Challenge requires the helicopter to hover at a height that is within ground effect for most practical designs, so the idea of making a practical human-powered helicopter is even harder than making something that fulfills the Sikorsky Challenge.
[...]It sounds like they were supported by two Chinooks, that came in a bit later, and those *were* seen by the Pakistani air defense, but the first group in weren't seen. A lot of other countries are going to want to figure out how we did this.
I suspect that a phone call from Obama might make a few helicopters 100% invisible to radar for a reasonable period of time.
... and been followed shortly by another phone call from someone working at the air defense unit to someone who knows someone who lets the target know about it, because that's the underlying problem with this whole situation.
I'm guessing that in the last two years there have been a lot of US radar operators working with Pakistani radar operators to coordinate air traffic involving US drones and spyplanes with Pakistani traffic, and the US radar operators were spending a lot of time watching just where US aircraft could fly with good radar return signatures and where they were really hard to see. Coordination of overflights with the radar operators would be easy to manage, but much more difficult for the Pakistani operators to realize that the US was mapping the Pakistani radar defense system like wardrivers map wifi.
The first time I heard about this whole mission, I thought, whoah, American helicopters managed to fly 150 km into Pakistan without being noticed? Pakistan isn't a slouch when it comes to military equipment: they've fought several wars with India, and are used to trying to track some of the finest military hardware in the world. Yet two helicopters flew in, invisibly. It sounds like they were supported by two Chinooks, that came in a bit later, and those *were* seen by the Pakistani air defense, but the first group in weren't seen. A lot of other countries are going to want to figure out how we did this.
There have been a lot of US projects in making low-observable helicopters, from the modified Hughes OH6 Loach used to surreptitiously place wiretaps on lines during Vietnam, that also used increased numbers of blades, and the cancelled RAH-66 Comanche, that was supposed to be quiet and have a vastly reduced radar signature. The ones used Monday are probably Blackhawks modified based on the stuff learned from the Comanche, but they could be completely new aircraft: the descriptions of the amount of personnel and material taken in are at the very edge of what two stock Blackhawks could carry, and adding lots of stealth technology adds a *lot* of weight.
Among other interesting things I've read and observed: the stock Blackhawk is manufactured with sheets of aluminum riveted together along the edges, like most planes. The pictures show rivetless construction, and in one picture it looks like there's a long weld seam that appears to have been done by hand rather than machine, making me think there are a very small number of prototypes of this. I also saw a link somewhere, that I can't find now, to a press release by a company who was adding small servos into the collector linkages that added continuous slight variance to the blade angle, to minimize noise by distributing it across different frequencies, which seems pretty cool. I've even seen a few claims that the whole aircraft was covered in material that could emit low levels of light, to blend it visually against a lighted sky (a technique used back in WWII by putting headlights on the leading edges of aircraft wings so that they could dive-bomb submarines without being seen until it was too late for the sub to dive. This was distinct from the british Leigh lights, that were used in after-dark attacks along with radar.)
I'm betting a whole lot of people are bidding on the wreckage that was recovered -- which is, itself, surprising, at least to me, because it sounds like the commandoes were able to completely destroy the whole main fuselage, leaving just the tail. Under the hurried circumstances that's pretty surprising. (I wouldn't be surprised to find out they actually hooked it to one of the Chinooks and dragged it out along with them.)
Sorry to reply to myself but the sequence about finding the old Empire nuclear reactors and talking about them with their caretakers, who claimed that their repair was a pointless question because it was impossible that they'd ever fail, was actually published in 1944 as "The Big And The Little" and later became the chapter "The Merchant Princes" in the first Foundation novel (by publication date) as per wikipedia: http://en.wikipedia.org/wiki/Foundation_(novel)
Slashdot Mirror
I'd love a Diaspora update, too, but in the last three months I've seen precisely zero activity there. Could just be my friends... but there is an uneasy balance between attracting/retaining enough people to get networking effects, and still ensuring their privacy. One thing I like about Google+ is it starts out with limited posting and limited circles and you have to work to reduce your privacy. It appears to me that Facebook has worked in exactly the opposite manner: you have to work to block and filter groups out. (I may be wrong: I used facebook briefly about three years ago.) When Google launched Buzz, they did something like opt-out rather than opt-in: Buzz included everyone on Google. That seemed to piss off a *lot* of people. I could understand why they did it: it got lots of people involved. But the bad press it got seemed like it overwhelmed any possible benefits of network effects. This time Google looks like they're trying something closer to Diaspora and hoping that because it's Google it'll be used by people other than privacy-obsessed techies (who are, or at least claim to be, actively repulsed by networking effects.)
One other interesting effect I've seen is that Buzz, which was getting more interesting over time, has really dropped off since G+ launched, because everyone who was on Buzz is now spending their effort on G+, and there are *lots* of new people on G+ as well. It'd be kind of sad if that was the main competitor that G+ killed.
Would you rather be t-boned by an idiot driver who runs a stop sign or hit in a glancing blow by an idiot driver who can't navigate a roundabout? A good roundabout where the curbing forces tangential entry is safer.
As a cyclist I'd much rather deal with someone not yielding in a roundabout than someone running a stoplight. A friend got hit by someone running a red light last week and he just got out of the hospital yesterday. (Also people tend to pull to the inside of a turn, so roundabouts leave extra room for bicycles.)
We *have* come up with good UI's: you just pay through the nose for them.
My favorite nice intertool integration feature of Cadence is that you can open schematic in one window on one monitor, layout on the other monitor, put layout in 'move part' mode, then click on a part in schematic and mouse over to layout and that part is stuck to the cursor, ready to be placed. When you're doing huge boards with repetitive layouts that's an unbelievable time-saver. On Altium, my favorite intertool integration feature is that I can change the definition of a part (symbol or footprint) in the library files, then click on it and hit 'update' and the tool will find all occurrences of that part in open documents, and update them, so (for instance) increasing drilled hole and annular ring size on a DIP's legs to make it fit the drill I have rather than the one it was modelled with is a thirty second operation. Both Altium and OrCAD have spreadsheet-driven setups, which I also love, because I can write an open office macro to fill a pin grid array numerically and with x/y coordinates, then open the spreadsheet file in the layout/librarian tool and import a mostly-finished footprint. All I have to do is add the silkscreen/keepouts.
I'll give KiCAD a run this weekend.
At work I use Altium, which is fairly easy to learn but has some weird bugs, draconian licensing, and breathtaking price, and Cadence, which is *not* easy to learn, has the most convoluted, cumbersome, and wretched new-parts-creation system I've ever seen, exostratospheric price, but the best PCB layout I've ever used. It's just dreamy. I also occasionally use OrCAD, which is quite reasonable if not outstanding on all fronts: schematic, layout, and parts creation.
At home I use gEDA. Schematic works fine, parts creation is pretty straightforward, layout is kludgy and not pleasant. (But newer revs have a gcode exporter built in, so it can go straight to a mill, which is nice.) I've used Eagle a little and it's pretty reasonable, although yeah I'm frustrated by the limitations. A friend and I are starting up a little company, and we use Eagle for small easy boards, and gEDA for larger more complex boards because it doesn't have those limits. I'm writing some scripts that I hope will allow me to import Eagle parts into gEDA formats (although I've written a bunch of scripts for gEDA already so it's so easy to make symbols it's not worth importing them, and it's fairly easy making footprints.)
I just downloaded KiCAD but haven't had a chance to try it.
I've also used PADS, which is okay but I feel like it's primitive for the price.
Sparkfun.com also has a cheap board fab service if you have long lead times, but it hasn't been competitive with myropcb.com for me, but that may be because I'm running production lots rather than one-at-a-time boards. (I cut those on the mill at work, and until recently was doing those via chemical etching at home, although now I have a mill at home.) I'll check out Seeed's prices/capabilities.
I used to but don't anymore, it doesn't appear. I read your other post in this article and your company will be seeing an RFQ next time I have a big ugly piece of test hardware to get fabricated. Specific questions: do you do conductive fill vias? do you do laser (or whatever technology) blind microvias on the 0.008" size range? and how small a drilled via can you put through a 0.187" thick board? I'm jbump at front range internet.
I design PCB's for a living, these days. Most board shops I work with have yearly tours/open houses: if you want to see an up-to-date shop see if they're throwing one. It's pretty cool to see. I'm mostly impressed by the electrical test machines: they look like a dozen mechatronic herons madly going after fish.
When I can wait a bit, I use myropcb because if I'm ordering 200 boards the size of postage stamps they're less than a dollar a piece including soldermask and silkscreen on both sides, if I can wait 10 days. (It'd be a lot faster but they tend to go slowly through US customs.) If I'm willing to pay a bit more, I use Circuits West, who will crank out up to 60 square inch boards for $31, and have had great quality.
However, the really great thing about milled boards is the turn time -- if you have a mill. I regularly go from hastily drawn schematic to finished, working board in under two hours, if it's a simple design. We can do three revisions of a board in a day, and *then* send it out to get a green board, once we're sure we have something working, and have a tested design ready for large-scale production in three days. We *love* having a PCB plotter in-house. It takes some thinking and experience to lay out good boards for it, but it sure helps productivity.
While I'm shilling companies that have saved me in crises, Vector Fabrication is not the cheapest place to get PCB's, but they'll produce a 30x30 cm 14 layer board with 3 ounce copper in two days.
I've got a passing interest in glasswork, and one of the things I learned is that it's more complicated than "melt into mold, let it cool". Glass has to go through a carefully controlled cool-down period so that the molecular structure will set up properly. Otherwise, the resulting object is far more brittle than it should be. If not done properly you can have cracks form during the cooling phase, ruining the object.
Does the incremental deposition solve the annealing problem? Being able to make glass objects without having to carefully control the cool-down would be very nice.
I was a glassblower and glass bead artist for a while. Careful cooling is pretty essential for lime glass, which is what we mostly use. It's less important for borosilicates like Pyrex, which is why glass casserole dishes can survive being put onto 200C metal racks in the oven, and it's even less important for fused quartz that's straight silicon dioxide. You can stick a pyrex rod that's less than a centimeter in diameter straight into an oxypropane flame without it splitting or snapping, and I believe you can do the same with a 3 or 4 cm quartz rod. Obviously this stuff isn't pure silicon dioxide, but it's closer to SiO2 than it is to lime glass.
Incremental deposition probably won't solve the annealing problem, but it'll change it: instead of having strain across big areas, you'll have little bits of strain distributed between each layer of glass that's put on so you're liable to get a lot of small cracks through the porous material, rather than one big catastrophic crack. However, all those little cracks generally tend to grow, but that may be somewhat helped by it being an amorphous, impure material: it's harder for cracks to run in long straight lines in crappy heterogenous stuff.
Given the whole square/cube thing about heat-producing volume vs heat-radiating surface area, it wouldn't surprise me if feathers developed as a way to keep juvenile dinos warm until they grew large enough that they didn't need them. Smaller dinos might have needed them all the time, but T. Rex "chicks" would have shed when they got larger.
You could be right. Many mammals -- cats and dogs, certainly -- are born essentially cold-blooded and will freeze to death if taken away from their mother and siblings, and don't start self-regulating their temperature for a day or so.
The thing with feathers being evolved to deal with rain was based on the observation by biologists studying hypothermia and different hibernation strategies, that it's pretty easy to stay warm if you're dry, and fairly easy to stay warm if it's snowing, but cold rain is extremely difficult to deal with. Huge feather-like structures can act like a raincoat. Look at the primary flight feathers on geese: massive close-knit waterproof feathers that when folded cover almost their whole backs, with little insulation value but great rain-shedding value. Likewise, their fine feathers on the rest of their bodies provide good insulation, with only some water resistance. When the bird is hot, it can stretch out its wings and expose its nearly unfeathered armpits to dump heat to the environment. Development of scales into insulating and water-shedding feathers could be an entire thermoregulation strategy, and flight an unexpected but very welcome by-product.
When I was poking around reading about this I was surprised to find out that there are thermoregulating plants: the sacred lotus maintains 85F even when the air temperature is close to freezing. Nature never fails to come up with something weirder than I could have imagined.
The point about body mass, if it's not clear, is that metabolism produces huge amounts of heat. When we digest food, about 50% of the energy that's released in converting complex molecules to carbon dioxide and water is released as heat. A huge animal can keep warm just from that. Mammals and birds maintain their temperature in a very narrow range. However, it's more complicated than just that would indicate: hibernating mammals (and one species of bird that hibernates) allow their temperature to fluctuate with the outside temperature. Likewise, there are reptiles that do some things to reduce their temperature variation, by seeking sunlight or shade, which is a type of active regulation. One current theory about dinosaurs and the evolution of feathers is that they showed up primarily as a thermoregulation system, providing insulation (particularly in rain) but allowing the animals to fluff their feathers to increase temperature losses to again actively thermoregulate.
How did the presentation go?
It had a happy ending...
With OBDII, the electronics in a modern car aren't that bad to figure out.
Sure, I couldn't probably make the average car sensor from scratch, but I couldn't make a '69 VW Beetle's brake drum from a block of steel.
Not that this is particularly relevant but my car -- a 2002 Subaru -- has a whole raft of undocumented OBDII stuff. Sure, it has the standard data, but a bunch of information like shift data seems to be only accessible if you have a Subaru diagnostics system, which sort of sucks.
Brake drums are cast out of cast iron. Doing cast iron has been too exciting for me: I stick to cast aluminum (and sometimes, if there's a problem, bits of cast aluminum stick to me, hence my unwillingness to try iron) but machining a brake drum from a casting isn't all that difficult. Now, casting a vented brake rotor: that's a challenge. There are plenty of parts that are pretty infeasible to try to make at home, like a ground cam or a nitrided crankshaft, or even the excitement inherent in casting the flammable magnesium engine block on a VW, but those are all several orders of magnitude easier than making your own electronic engine control system from silicon.
One of my coworkers spent several hours chatting with Bob Pease about making fuel injection systems, last time Bob was visiting here, as it happens. My coworker had just finished adding one to his 1965 Jeep, that he'd machined most of the hardware for (although he'd purchased the fuel pump and injectors) and Bob was interested in some of the technical details of how it managed fuel/air ratios.
True in some places in Colorado things like rain barrels are illegal. Which doesn't make much sense to me at all but there you go.
It's not that they're illegal per se. It's that in arid states, water rights are more than 100% used, and first in time of use gets first priority, meaning the people who have water rights older than yours, get to use their entire allotment of water before you get to use any. Collecting rainwater means someone else doesn't get their full allotment. As it turns out, California is one of the major water rights holders, so it's not because the guy down the street is complaining that you can't retain the water that would otherwise run off your land into the gutter: it's because the likes of California and Arizona have won large numbers of SCOTUS rulings upholding their water rights. (I've been told that water rights cases are the largest category of cases to go to the Supreme Court, but don't have a citation.)
And not being able to collect water in a rainbarrel is by no means the weirdest consequence of this: people who live near irrigation ditches or rivers have been required to cut down trees, because the trees are sucking up groundwater that is being replenished by leakage into the water table from the canal or river.
Water rights is a huge facet of local and state governments in arid states.
The capitol building of Denver has the dome covered with 200 ounces of 24 carat gold, and it needs replacing about every 40 years. That means it's losing several ounces per year, and most all of it is coming off in rainwater that ends up dumping through the drainage spouts down the sides. People tried to collect it, like this guy. I've been told, but can't find an online reference, that collecting rainwater from the Capitol Building was made illegal to prevent mayhem, so now it goes down the sewer.
It didn't have a radio. (That I saw: I was looking over his shoulder.) The main reason he was so into it was precisely because he could fix everything on it -- and he did, too. He had a lot to say about why people shouldn't own anything they couldn't fix, and about how nice it was to be able to walk down to the corner auto shop and get most all the parts he needed to repair or replace anything on the Beetle.
And I want Apple to defend it with all the power it has... So that only Apple devices are blocked and all other devices are unaffected.
As if. Apple will get this, then they'll license it for an enormous sum to all the other cellphone companies when the Apple and police lobbies manage to get mandatory implementation of this enshrined in law.
For what it's worth, there are already several open standards for lighting communication, including 0-10V and DALI, and both have reasonably wide adoption. The consumer marketplace could well be controlled enough that a major manufacturer might be able to push a protected, patented, proprietary interface and give it some goofy name like Y11, but in the industrial lighting space -- and that includes outdoor, sign, and indoor in factories, warehouses, malls, government buildings, and sometimes even apartment complexes -- the market is fragmented and people would rather design things that work with existing control systems than try to talk customers into buying complete systems. I know a bit about this since I've been designing a lot of DALI transceivers lately. They're pretty cool: noise-resistant, multi-topology, and polarity-agnostic so you can string them up beside industrial 440VAC and hook the wires together any old way you want and they'll still work.
There's also a lot of work going on in mesh networking, but that's nowhere near as standardized, as far as I can tell.
As for the 'forgot the phone' thing, if this is being aimed at consumers that's merely a happy side-effect: the main point is so building management companies can have fine-grained control over dimming, and lighting that reports back when it's dying or dead to minimize maintenance costs. It'll be centrally controlled so forgetting phones won't be a problem. Most of the stuff that's in the pipeline does ambient light sensing, motion sensing, heat sensing, and sound sensing, and likely RFID, so it can figure out if people need lights on even though they're still (in the bathroom) or if they need lights turned off (in the bedroom), or if lights should turn on well in advance of someone approaching (as they cross a parking lot or walk up to the front door) or if it can come on milliseconds before the person gets to the place where light's needed (hallways).
Read up on the Bonus Army riots of 1932 and the background/history of US compensation to armed forces members. Your experience is historically speaking an anomaly. I'm glad the government is treating veterans well these days, but it appears that such treatment is rare.
There was an interesting article in the New Yorker about two weeks ago, detailing the science behind the attempts to grow artificial meat. Unfortunately that link only gives an abstract and a small section, but it's an interesting read if you wander down to the library and check it out. So far, the state of the art is a piece of artificial meat about 14mm long, 3mm wide, and maybe a couple cells thick, and even then it's just muscle cells without any actin/myosin structure so it'd be like, I dunno, moosh. But they're making progress.
There's no shortage of psychos and crazies. What made Kaczinsky interesting was that he was a very smart psycho -- but even then, there are lots of those. All the interesting stuff he wrote is public record because it was used as evidence during the trial and published in the court proceedings. Ideally they'd sell all the stuff and use the proceeds to help repay medical costs/give disability benefits to the people he injured, but more likely they're going to use it to indirectly defray some of the cost of the trial and his ongoing incarceration.
If you want a cheap radiation source, you can buy 2% thoriated tungsten rods for TIG welding. Find a local store and buy a 1/16" rod: individually they only cost about $8. Other people have suggested Coleman lantern mantles but the ones you can buy these days don't seem to have thorium in them anymore, because the old ones were *seriously* radioactive. If you *have* a Geiger counter you can go to an Army/Navy Surplus store and check the ones they have since a lot of old radioactive ones are still in stock.
My homebuilt geiger counter, using a surplus Russian GM tube, can easily detect a single thoriated tungsten rod if held up close to the tube, as can my vintage Civil Defense CDV700. Both will also detect a smoke detector.
If you want to build your own geiger counter and have a tube, here are instructions for building a high voltage power supply from a hacked-up flash unit from a disposable camera and here is the detector circuit that translates that into audible clicks. If you optoisolate that detector circuit you can feed it into an Arduino and log/display counts per second on a laptop. (It needs optoisolation because the output of the audio click board is negative with respect to power and way more than 5V, so it'll cook an Arduino, as I found out. Although an Arduino analog input can withstand -200V and still function, amazingly enough.)
If you just want to detect ionizing radiation, you can build an ionization chamber. My company supplies DIY kits but we also have detailed instructions for making your own with a component list of like four transistors and a handful of resistors, and a tin can. They're more sensitive than a Geiger tube, although they're much slower to react, taking seconds to change their reading. One neat thing is you can build them as chambers so you can actually put a sample inside the chamber, if you want, and they detect alpha, beta, and gamma.
And as other people have said, any sample you buy that'll allow you to characterize your radiation detector will expose you to tens to thousands of times as much radiation as anything in Japan unless you're actually inside the grounds of the power plant, so this whole project might not do what you want.
>2) How many times have explosives been found?
None. Ever. Even the underpants bomber made it through.
Since the inception of the TSA, they have stopped *zero* hijacking/bombing attempts from the airport.
Biggest waste of money on security theater going.
--
BMO
They're not trying to stop bombing attempts. They're trying to stop their employers getting reamed by the media when a bombing attempt is successful. "But we x-rayed everyone and groped *babies* -- what more could we do?" is probably the best defense the likes of the FAA, FBI, and TSA have available to them the next time someone blows something up and the media is all "but look at all the WARNING SIGNS: how could you be so dense as to not notice this one dude out of 300 million other dudes who did four things that in hindsight make it clear he was building bombs but at the time looked like relatively innocuous actions?" The whole TSA is a giant exercise in ass-covering. Stopping actual terrorists is a slight side-effect.
...compared to just climbing up a ladder?
What I mean is the human body has easily enough power to raise itself up a vertical ladder or rock face so presumably a huge amount of this power must be lost just moving air around when that power is used inside a human powered helicopter. But how much power is wasted , or to put it another way , how much power put into the system is actually used to raise the mass of the helicopter?
I'm not exactly sure what you're asking here, but think about it this way. Gravity is accelerating you downwards. When you stand on a ladder, the ladder's structure is resisting your weight, which is why you stay up. With a helicopter or airplane, you don't have a structure to hold you up, so instead you accelerate air downwards using wings. You push the wings through the air, they accelerate air downwards, you get supported.
So the easiest answer to your question is *all* the power put into the system actually raises the mass of the helicopter. A slightly more nuanced answer would be that the mass of the whole system (pilot + helicopter) minus the pilot's weight, is how much energy it takes to raise the mass of the helicopter, but that's sort of silly.
A side-note: there's something called ground effect that changes how helicopters (and airplanes and anything else relying on accelerating air downwards to maintain flight) work. When a wing is within about one wingspan of the ground, the air it forces downwards is somewhat constrained between the ground and the wing, giving the wing more lift than it would in free air. As a result, if you're within a wingspan of the ground it requires significantly less power to stay in the air. (If you watch airplanes land you'll see this effect as a change in their angle of approach just before they land.) The Sikorsky Challenge requires the helicopter to hover at a height that is within ground effect for most practical designs, so the idea of making a practical human-powered helicopter is even harder than making something that fulfills the Sikorsky Challenge.
[...]It sounds like they were supported by two Chinooks, that came in a bit later, and those *were* seen by the Pakistani air defense, but the first group in weren't seen. A lot of other countries are going to want to figure out how we did this.
I suspect that a phone call from Obama might make a few helicopters 100% invisible to radar for a reasonable period of time.
... and been followed shortly by another phone call from someone working at the air defense unit to someone who knows someone who lets the target know about it, because that's the underlying problem with this whole situation.
I'm guessing that in the last two years there have been a lot of US radar operators working with Pakistani radar operators to coordinate air traffic involving US drones and spyplanes with Pakistani traffic, and the US radar operators were spending a lot of time watching just where US aircraft could fly with good radar return signatures and where they were really hard to see. Coordination of overflights with the radar operators would be easy to manage, but much more difficult for the Pakistani operators to realize that the US was mapping the Pakistani radar defense system like wardrivers map wifi.
The first time I heard about this whole mission, I thought, whoah, American helicopters managed to fly 150 km into Pakistan without being noticed? Pakistan isn't a slouch when it comes to military equipment: they've fought several wars with India, and are used to trying to track some of the finest military hardware in the world. Yet two helicopters flew in, invisibly. It sounds like they were supported by two Chinooks, that came in a bit later, and those *were* seen by the Pakistani air defense, but the first group in weren't seen. A lot of other countries are going to want to figure out how we did this.
There have been a lot of US projects in making low-observable helicopters, from the modified Hughes OH6 Loach used to surreptitiously place wiretaps on lines during Vietnam, that also used increased numbers of blades, and the cancelled RAH-66 Comanche, that was supposed to be quiet and have a vastly reduced radar signature. The ones used Monday are probably Blackhawks modified based on the stuff learned from the Comanche, but they could be completely new aircraft: the descriptions of the amount of personnel and material taken in are at the very edge of what two stock Blackhawks could carry, and adding lots of stealth technology adds a *lot* of weight.
Among other interesting things I've read and observed: the stock Blackhawk is manufactured with sheets of aluminum riveted together along the edges, like most planes. The pictures show rivetless construction, and in one picture it looks like there's a long weld seam that appears to have been done by hand rather than machine, making me think there are a very small number of prototypes of this. I also saw a link somewhere, that I can't find now, to a press release by a company who was adding small servos into the collector linkages that added continuous slight variance to the blade angle, to minimize noise by distributing it across different frequencies, which seems pretty cool. I've even seen a few claims that the whole aircraft was covered in material that could emit low levels of light, to blend it visually against a lighted sky (a technique used back in WWII by putting headlights on the leading edges of aircraft wings so that they could dive-bomb submarines without being seen until it was too late for the sub to dive. This was distinct from the british Leigh lights, that were used in after-dark attacks along with radar.)
I'm betting a whole lot of people are bidding on the wreckage that was recovered -- which is, itself, surprising, at least to me, because it sounds like the commandoes were able to completely destroy the whole main fuselage, leaving just the tail. Under the hurried circumstances that's pretty surprising. (I wouldn't be surprised to find out they actually hooked it to one of the Chinooks and dragged it out along with them.)
Sorry to reply to myself but the sequence about finding the old Empire nuclear reactors and talking about them with their caretakers, who claimed that their repair was a pointless question because it was impossible that they'd ever fail, was actually published in 1944 as "The Big And The Little" and later became the chapter "The Merchant Princes" in the first Foundation novel (by publication date) as per wikipedia: http://en.wikipedia.org/wiki/Foundation_(novel)