Forgot your password?
typodupeerror

Comment: Re:The industry will screw you anyway... (Score 1) 182

by smellsofbikes (#48014873) Attached to: Breakthrough In LED Construction Increases Efficiency By 57 Percent

Couldn't they make the phosphor on the led's slower?

I'm not a phosphor chemist so I may not be right on this, but it's my understanding that despite the word 'phosphor' the coating that downconverts light in an LED is actually a fluorescent phenomenon, meaning the metastable states have lifetimes on the orders of tens of nanoseconds. Actual phosphorescent phenomena have lifetimes long enough to make a visual difference but because they stay in an excited state a lot longer they have a lot more time to engage in non-radiative relaxation, so their conversion efficiency is like 10x lower.

Comment: Re:The industry will screw you anyway... (Score 2) 182

by smellsofbikes (#48006291) Attached to: Breakthrough In LED Construction Increases Efficiency By 57 Percent

Buy Crees. I work in LED driver design, and Cree, who I don't work for but I work with, seem to do a good job of making sure their LED's don't get associated with junk. Philips similarly, to a lesser extent.

Weird. My experience has been the opposite. I've tried the Cree bulbs from Home Depot and they suck because they strobe at 120 Hz (verified on a scope). That's not usually noticeable except when you move your eyes quickly (like reading), or if something moves quickly like your kid swinging a baton. The strobe effect really bothers me. I also have 15 of the Philips L-prize bulbs that they discontinued after collecting their prize money, and those do not have any sort of strobe effect and they are more efficient than the Cree bulbs.

Strobing is a huge problem, and the easiest way to fix it is add big output caps to the switcher... which costs money.
It's sad to hear they did that. I will chat with someone who gets to make these decisions for them at the end of the month.
Strobing's even worse with car taillights because that's when people have the highest saccade rates and cars are moving quickly, so surprisingly high frequencies are clearly visible as strobe flashes.

Comment: Re:The industry will screw you anyway... (Score 2) 182

by smellsofbikes (#48006263) Attached to: Breakthrough In LED Construction Increases Efficiency By 57 Percent

But yeah, we really DO get what we pay for. So dear consumers who are reading this, please protest by not settling for the crappy stuff. Buy quality and prove to the world that's what we want! I've been 30+ years into electronics (many as a service tech). We've got a heck of a job in front of us, but I honestly believe the public will tire of the crappy products, hopefully NOT before it's too late.

The big open question for our time: how do we tell if stuff is quality?
Stuff that has the same manufacturer's SKU number, you open it up and it has all different guts than last year's because they've changed subcontractors.
They come out with a new version every four months, so by the time reviews are up on one you can't buy that model anymore.
Manufacturers have adopted influenza's tactic: change so fast that the system can't keep up with you and fight your badness.
Since us consumers need to buy stuff, we have no choice but to buy what's being offered, with only lemon metrics for judging.
Cree is an exception to this (in my experience), and I hope like mad that they stay that way.

Comment: Re:The industry will screw you anyway... (Score 4, Informative) 182

by smellsofbikes (#48005553) Attached to: Breakthrough In LED Construction Increases Efficiency By 57 Percent

...because it doesn't pay very well to sell you something that'll last forever, whether it's an Oled screen or LED bulb.

With LED's, it's a walk in the park for the industry to make them last less, all you need to do for your LED to last less than specified, is to OVERDRIVE them just a little, a little higher current and the LED's will die rapidly, they should be able to make the new LED lamps last just out the warranty period (that in most countries AFAIK is around 3-6 months), or cheap enough to avoid the warranty altogether.

There is nothing wrong with the LED's themselves, (we're talking the components...DIODES...not the whole circuit with drivers and all), I ordered strong RGB leds from China many MANY years ago, they're still glowing on my homemade alarm-systems so strong that I can use them as night-lights, yes...4 years later 24H day use...they still glow enough to lit up an entire room. And I just used Ohms law + 1% resistor values to calculate the right resistor value for my circuits. You can pretty much BET the manufacturers will "miscalculate" these values, or make the drivers for the stronger LED's last MUCH less in order to keep pumping out new ones for the consumers to waste and waste.

I'd rather pay a proper price for my LED lamps - and keep our environment safe from this mad overproduction that now has escalated totally out of hands. :(

Buy Crees. I work in LED driver design, and Cree, who I don't work for but I work with, seem to do a good job of making sure their LED's don't get associated with junk. Philips similarly, to a lesser extent.
So, from the inside, it's not that manufacturers generally scrimp on bulbs to make them fail faster so they can sell more. The economics of light bulbs don't support that business model. It's that people are crazy reluctant to pay $15 for a lightbulb when an incandescent costs under $1. So manufacturers engage in heavy-duty Muntzing until the bulb will just barely run, and they've cut the BOM by $1.45... and then it dies quickly. It's called value engineering, which as far as I'm concerned means removing all the value. They use cheap input filter caps, and scrimp on those, and they use cheap heatsinking which is poorly thermally coupled to the LED's, so the LED's operate at a high junction temperature and don't live very long.
Incandescents have visual inertia, for lack of a better term: if you pour a 30 hz square wave into one, it'll still look pretty good. LED's react in nanoseconds. Crappy dirty line power combined with dimming makes for a really demanding design, and designers and apps engineers have to work with a huge variation in dimmer designs. Consumers don't see any of that: all they see is "no way I'm paying $25 for a lightbulb" so they buy the crap ones and then get infuriated with them because they're visibly flickering and only last five times as long as an incandescent. I can't really blame them, either. There are really good lightbulbs out there. They're expensive. They should last 50,000 hours. But it's hard to tell what you're getting if you're not in on the design.

Comment: Unspoken and faulty premise (Score 1) 478

We largely agree that it sucks to be stuck drooling in a nursing home. But the reason people are keeping strict diets, exercising, and doing math puzzles is almost certainly not to live longer, but to live better during the time that they have. I want to die the moment living isn't fun anymore, but I want to delay that moment as long as possible. That's why I spend time and effort on keeping healthy: not because I simply want to live forever, but I want to feel like I'm able to have a really good time forever.

Comment: Re:Underspecced? (Score 1) 105

by smellsofbikes (#47961487) Attached to: Dremel Releases 3D Printer

A makerspace is definitely the best bet as regards hardware. If you think you're going to pursue this, start playing with modeling software now. It's at least as complicated. (Moreso if you get a 3d printer that already works and you don't have to assemble and tune it.) Sketchup, Autocad's new free 123d or whatever it's called, freecad, are all very usable for graphics-oriented, and openscad is good if you're a programmer. I find freecad the easiest combination of precision, adaptability, and ease of use, but other people have totally different opinions.

Comment: Re:Underspecced? (Score 1) 105

by smellsofbikes (#47949145) Attached to: Dremel Releases 3D Printer

What I tell people who are thinking about 3d printing is: if you have a specific project, that needs 3d printing, for which going through shapeways or something is either uneconomical (because you're going to need six tries to get your widget dimensioned correctly) or too slow (you're going to be making a ton of different prototype widgets) then a home 3d printer may be a good idea for you. Otherwise, you'll get it, print an octopus and a tardis, and then it'll gather dust and you'll kick yourself for having spent the money.
With that said, if you do have a specific project, and you use the printer for that, you will get enough time on it, and more specifically on using the software to make models, that you will have basically mastered the learning curve, and suddenly you'll be printing a lot of other things, that you didn't ever even think about making.
I'm co-owner of a plus-size mendelmax 2. We got it to print prototype circuit board adapters so we could stick x board on y piece of hardware. Once we'd gotten that hammered out, the other guy who owns it has printed a plug for his sewer drain, a rat trap for live-catch, buckets for a tiny pelton wheel generator, and I've printed lathe-holding tools, lcd bezels, automated printed circuit board test fixtures, and most of a fuel injection intake manifold for my car. We use it for everything.
But you need to have that first big complicated project that you have to get finished, to get to the point where it is a reliable tool, rather than a gadget.

With all THAT said, you'll always want a larger printer. But if the printer you have can cover 95% of your jobs, that's a whole lot better than none at all. Based on the stuff I've made, this printer could handle 95% of the demands I have, and there's always shapeways for the other 5%.

Comment: Re:Underspecced? (Score 3, Insightful) 105

by smellsofbikes (#47941117) Attached to: Dremel Releases 3D Printer

Is it me or does it sound a bit underwhelming for $1000? I don't mean the price is non-competitive, it just seems like I'd want something more capable if I was going to take the plunge. Burn $1000 and in a week won't you be hankering for a much more capable machine?

Yes. And spending two months debugging bed/head temperatures, print and extruder speed, and layer thickness, so your prints consistently stay solid and adhered to the bed rather than peeling, will be totally invisible to you because that $1K presumably means someone else already did that. There's a lot of value in getting something that's been debugged, and that's particularly the case for extrusion-based FDM 3d printers. It's okay to be hankering for a better machine, particularly if you're already printing. The best 3d printer is the one that's actually building parts for you.

Comment: Re:is it just me... (Score 1) 471

by smellsofbikes (#47874829) Attached to: Ask Slashdot: What Smartwatch Apps Could You See Yourself Using?

or is there a hidden strategy of increasing the phone sizes of new iphones to deliberately make them unwieldy, and create a problem which can be "solved" with a smart-watch? ie, more crap to sell.

Yes, and it started about five years ago. That's when my wife and several of my other female friends began complaining that phones no longer fit in the pockets of women's clothing that isn't made by Lululemon or something. They got things like the HP Veer precisely because it was still the size of an old flipphone and managed to fit in the roughly six cubic inches that manufacturers have decided is the amount of pocket space they will provide for women's clothes below size 2. Then all those went away as everyone raced to make phones that are really tablets, so now they're buying smartwatches and keeping the cellphone in the purse or in the desk. Sure, it's a big awkward batphone, especially on someone who's a size zero, but it's a lot less ginormous than any cellphone you can buy in the US, as far as we can tell. (If anyone can suggest a cellphone that's roughly the size of a Veer but is still supported, I'd love to know about it.)

Comment: Re:What I think would be most useful (Score 1) 471

by smellsofbikes (#47874753) Attached to: Ask Slashdot: What Smartwatch Apps Could You See Yourself Using?

>The only ant+ holdouts I can think of are cycling power meters. But those are very expensive bordering on niche sports training products.

I wonder about this. Out of the 20-some people I ride with regularly, only one person has an ANT+ accessory that isn't a power meter. Or, to put that another way, everyone who has ANT+ bits has multiple bits, of which one is a power meter. However, that is likely because of sample bias: they're all racers. But of the non-racers I know, none of them has cadence and if they have HRM they're using watch-based ones like Polar's.

>Speaking of exotic bike stuff. I know the Di2 electronic shifters support ant+.

DI2 supports a shimano-proprietary version of ANT+, apparently: http://bike.shimano.com/publis...
We have access to ANT+ hardware at work, since we make it, and we haven't managed to talk to my coworker's DA Di2 yet. It raises some interesting possibilities, of automatic shifting (and of the team manager/coach deciding when you should start your sprint, messaging you, and managing your shifting. Very Triplets-of-Belleville.)

Comment: Re:Me too (Score 1) 635

by smellsofbikes (#47792563) Attached to: Ask Slashdot: What Old Technology Can't You Give Up?

Out of curiosity, what model of Opel? I always wanted an Opel GT...

I've gone through three of those plastic box fans over the last ten years, and whenever one dies I drag out the big heavy steel box fan my parents had in the 1950's. Man does that thing move air, all day long, fairly quietly. It's just terrifying because if it fell out of the window it could kill someone, but whereas the plastic ones jump to their doom every couple of months I think it would take a volcanic eruption to shift the old fan.

Comment: Just about everything I own (Score 2) 635

by smellsofbikes (#47787907) Attached to: Ask Slashdot: What Old Technology Can't You Give Up?

Pushbutton hard-wired phones, world war two vintage drillpress, metal lathe, wood lathe, tablesaw, 1970 Triumph as my not-snowing car, 1990 bicycle for my non-race bike, MOO/MUD's that I've been hanging out on since 1992, Commordore Amiga 2000 (okay, I only fire that up about once every two months.) A lot of my wood chisels are from the 1890's. They all work just fine. My race bike is a brand-new marvel of carbon fiber and magnesium, but I bet it won't last another two seasons, whereas the old bike has over 150,000 kilometers on it. I do now design using switching power supplies, rather than LDO's, and I've moved from PIC to AVR, (and I've always programmed in C rather than assembly) but generally, there has to be a really clear advantage for me to change piles of experience and knowledge for something new.

Comment: Re:Its all in the gmail terms of use ... (Score 3, Interesting) 790

That means only the most incompetent pedos aren't already randomly tweaking their jpgs - the smart ones are doing it in the EXIF section so it won't even change the picture.

The smart implementations probably hash the image payload excluding EXIF, for exactly that reason - maybe downsample and reduce the colorspace too, so trivial tweaks won't have that effect any more.

This isn't definitive research, but in the early days of G+, some friends posted a lot of porn to see how quickly Google caught and deleted the pictures. What they found was that Google's algorithms, once trained with a picture, could find that picture if it had been resized, flipped along the vertical axis, and cropped. (one was cropped to the point where it was no longer technically porn, since it was just a person's face, and it still disappeared.)

Assembly language experience is [important] for the maturity and understanding of how computers work that it provides. -- D. Gries

Working...