Dan's Data letters #72Publication date: 12 November 2003.
Last modified 03-Dec-2011.
I recently bought a PC ATX case (midi-tower) and it has front USB ports that are 1.1 compliant. I am trying to find the best solution to connecting them and having them run with USB 2.0. What kind of device do I need to make these front ports useable?
The ports should Just Work. The cable spec for USB 1.1 and 2.0 is the same, as far as I know; USB is timing-limited rather than cable-quality-limited. You just need to be able to physically connect the front ports to your motherboard correctly, if they use pin-header connectors. If they use loopback cables that connect to rear panel USB ports, it's even easier.
I have seen companies that have BIOS updates, drivers, and the like that come as an executable and must extract an image to a floppy. This is a problem for me, as I do not have a floppy drive. While I know that sometimes you can open these executables with WinRAR to access the files inside, it does not always work. I was wondering if you knew of any program for Windows (XP/2K) that would be able to emulate a floppy drive. Something that worked similar to DAEMON Tools for CDs, but that was for 3.5 and/or 5.25 disk drives, and would support reading and writing.
Neither of these is any good if you want to boot from the faux-floppy to flash a BIOS, though. If you do, then that letters column also has an answer for you.
It appears my Gateway's CPU cooling fan quit. It runs for a while, then stops, heating up the CPU. I need a new cooling fan but am a little apprehensive about doing it myself. Are they difficult to replace on a Pentium 4?
The standard P4 cooler's fan is an integral part of the retention frame over the top of the heat sink. A new frame from a similar stock cooler ought to be a drop-in replacement (unhook the old frame, hook on the new one), and any decent computer store that assembles PCs will probably have some spare stock coolers sitting around that they'll sell cheaply.
Fans on more traditional CPU coolers are usually screwed in place, and are also easy to replace with another fan of the same dimensions, almost always without even taking the cooler off the CPU.
I bought a fancy digital camera a while back and love it, but I'd really like to be able to show off my best pictures without the need for a PC (i.e. at the pub). I've done some searches and have found products like the MemoryFrame line by Pacific Digital, but none of them seem to be particularly portable. I've thought about purchasing some sort of PDA for the purpose as well, but I don't think the resolution on most PDAs would be adequate for displaying photos.
It would seem to me that it wouldn't be that difficult to produce a little standard photo sized LCD device that accepts some format of memory card, a few AA batteries and displays my photos at 800x600 resolution, though I could be (and probably am) wrong about that.
Am I just not looking in the right place, or are there really no products out there like this?
Smart-arse answer: Buy a printer.
More helpful answer commences... now.
Digital picture frames that run from plugpacks could run from a battery pack as well, but it's not an elegant solution. There are also screen-equipped "digital wallet" gadgets, but they have pretty small screens and aren't cheap, since there's a laptop hard drive in there too.
Have you considered using a plain old second hand laptop? Overkill if all you need is a picture-displayer, but if you could use it for other things too, ye olde P-II ThinkPad with 128Mb of RAM and Win98 wouldn't cost you a fortune (an older Pentium machine would be even cheaper, but a bit marginal performance-wise), and would work very nicely as a picture viewer. Good-sized screen, PCMCIA slots for easy memory card access with cheap adapters, hard drive for storing a whole library of pics. You could also use it as an on-the-road image storage and review device.
With a zooming viewer, PDAs can be surprisingly useful for the jobbing digital photographer (see this software review, for instance), but as you say, they're not greatly more exciting than the back-of-camera screen for regular viewing.
I am working on a project, in which I will have a piece of plexiglass with a design etched on it. I wanted to make the design glow using LEDs, but the LEDs I bought were rather horrible. I have been doing most of my shopping at a local Radio Shack, since it is the easiest for me. The only ones that I found that are in stock in my area and look like they might work are these.
Now, I admit they are a bit big, but is 5000 MCD really going to make that panel glow? I am using a 5 volt power source, so if I buy two, and run them in parallel, do you think it will accomplish my task?
Radio Shack and disappointment. Now there's an unusual combination.
You don't need super-duper new-model ultra-LEDs for this job, so there are lots of decently priced options that don't involve going to the dreaded Rat Shack. Hosfelt Electronics are one of the best-known LED dealers online; there are also plenty of eBay dealers you could patronise; pick one with lots of positive feedback and you should be fine.
The candela (or millicandela, "mcd") isn't an overall brightness figure; the tighter an LED's beam, the higher its mcd rating will be. 10mm LEDs generally have pretty narrow beams thanks to their big lenses, so 5000mcd ones aren't actually likely to be very bright by current cheap-LED standards. I talk about this in more detail in the "Decoding light statistics" section of this old review.
When you say "safety switch" (in reply to Dawid, letters #71), I'm assuming you mean a Residual Current Device (RCD). I have heard electricians talk about these as earth leakage RCDs (as opposed to another type which I think just monitors the current flowing), so I assumed that they only offered shock protection if the connected device was earthed.
The house we are in at the moment has an RCD integrated into the bathroom power outlet, but most of the appliances that are connected to it are the two-pin double insulated variety. I wondered if these sort of appliances were still shock-protected, and an electrician I asked, after some umm-ing and ahh-ing, assured me that they would still be protected, despite not having an earth lead.
Is this the case, and how does the RCD detect a problem when there is no earth lead?
There are two basic kinds of small-current-fault-detecting "safety switch"; RCDs and Earth Leakage Circuit Breakers (ELCBs).
ELCBs detect current flowing in the circuit's earth wire, which does you no good if current's going from active to earth through a cold water pipe via your body, because the cold water pipe isn't the earth wire that the ELCB's monitoring.
RCDs are better, and more popular, because they monitor the current in the active and neutral leads. They trip if that current differs by more than a set amount. Maybe only a few milliamps, for a domestic "safety switch".
If current's going into the house on one lead, but not all of it's coming out on the other, the balance must be getting to earth, one way or another. It doesn't matter whether it's going to the actual mains earth connection or getting to earth down a cold water pipe or through the body of a luckless human standing in a puddle on the concrete garage floor; the RCD will still trip, because it's just looking at the difference between active and neutral current. So if your bathtub is earthed and your power circuit is RCD protected, throwing an un-earthed appliance into the bath probably will trip the switch.
The down side of both ELCBs and RCDs, besides their uselessness if current is going from active to neutral but passing through your heart on the way, is nuisance tripping. Many appliances, including PCs, have a milliamp or three of earth leakage. This doesn't necessarily mean they've got any loose wires sparking away inside; capacitive and inductive leakage can happen right through insulation, or as a result of deteriorating power supply components, motor starting capacitors and so on; PC power supplies commonly have a tiddly bit of earth leakage even when they're working perfectly.
If your appliances add up to more leakage than the safety switch will tolerate, it'll trip when nothing dangerous is happening.
In Letters #71, you observe that "most hair dryers, little plastic fan heaters and similar bathroom hazards" are double-insulated.
Why aren't they earthed?
The idea of earthing appliances, as far as safety goes at least, is to connect any metalwork the user could touch to ground. If a live wire gets loose inside and touches the chassis, earthing means the current will go to earth (and probably pop a fuse) rather than just energise the chassis and wait for some other earth-y thing, like a person, to come along.
You can defeat earthing if you really try, by going direct to the source with a conductive object that you're holding - the classic example is poking around inside your toaster with a butter knife.
Double insulated appliances have at least two layers of insulation between anything that could be energised - whether or not it ought to be in normal operation - and the user. They generally also have a plastic casing, which can serve as one of the insulation layers. You can't earth a plastic casing, but that doesn't matter; it can't become live, either.
Just a quickie reply in relation to "More bolt-on magic" Fuel Saver story.
I've never bought one of these devices, nor am I related to anyone that manufactures them, but I have read a bit about these gadgets that spin air in the intake before it enters the combustion chamber of the engine, supposedly increasing the airflow. Whether or not they work is beyond me, however try this experiment which may relate in some ways (even though it's using water, not air).
Fill a 2 litre soft drink bottle almost full with water. Hold your hand over the mouth of the bottle, tip it upside down over the sink, and release the water, taking note of the time it takes to empty the bottle (keep the bottle still, don't shake it).
Now try the same experiment again, except this time while it's upside down and your hand is still blocking the flow, proceed to swirl the bottle around its vertical axis so that the liquid is spinning around inside rapidly.
Then release the liquid again and measure the time it takes to empty. The liquid should come out spinning and the bottle should drain far quicker.
My guess is that the second method creates a clear flow for the air to rush into the bottle up the centre of the vortex of water, thus making it quicker for the water to release.
Do you think that a similar principal can be applied to air flowing into a combustion chamber whereby the airflow is more controlled by the fins of the fuel saver device?
You're quite right about the soft drink bottle trick, but it doesn't apply to the magic automotive air swirlers.
As you say, swirling the bottle means that instead of air bubbles going gloop-gloop-gloop chaotically up through the water, occluding the water flow completely from time to time, the air has a quite stable path up the middle and the bottle drains faster.
Any other easy way for air to enter will do just as well; cut a hole in the base of the bottle and it'll drain fast without any spinning.
The difference between this situation and the Fuel Saver Whatever one is that nothing is going back the other way in the car engine. There's no sealed vessel that air has to enter to replace some substance that's leaving, like water from a bottle; there's just an engine sucking air (or air/fuel mixture, in the case of carbureted engines) into itself from the outside world, and then blowing it out through its separate exhaust system.
You've written often about the fun and profit to be had while playing with magnets. Well, intellectual profit, maybe. But after reading your articles, and checking out Forcefield and such, a simple question comes to mind: What do you do with magnets when you're not playing with them? I mean, you can't turn them off, can you? How are these large, superstrong, rare-earth magnets shipped or stored?
I know they don't have much magnetic strength left even at short distances... but I imagine somebody putting them up on a shelf... a shelf on a wall shared with a neighbor... the neighbor who collects rare video tapes... hahaha... erm... ahem.
No, I didn't do that, I just like worst case scenarios.
Do your packages stick to the side of the UPS van?
Forcefield ship their magnets in boxes that're lined with mild steel U-sections, which mate together to produce a decent magnetic containment vessel to prevent your parcel attracting attention to itself by becoming attached to handling machinery, other parcels, or the delivery guy's belt buckle.
You can do the same thing for storage of big magnets at home, though it's not normally difficult to find some random shelf-end or overhead conduit onto which you can stick them to keep them out of mischief.
Little magnets can just be clicked together into a blob safely enough, or stuck on some random bit of iron. Really big magnets need to be wrapped in cardboard and tape for storage or shipping, to reduce the "pinch factor" they can achieve if they get near each other, and also to prevent them from slamming into each other at great speed and chipping, or just plain exploding. Cardboard padding protects them well.
The two giant cylinder magnets Forcefield sent me the other day came taped into a cardboard wrap, with plastic ring spacers on either end and copious newspaper stuffing inside their U-section box. Even wrapped, they're still not safe to handle near each other.
In Letters #69 you write "there are 16.8 million to the power of 307,200 possible images that can be displayed on a 640 by 480, 24 bit screen".
Ok, no problem with the basics, we're talking about all possible permutations of all possible colours in all possible positions, I can cope with that. But here's the tricky bit - if one were to theoretically actually do this, would all resultant images, excluding basic mathematical patterns, be effectively randomized pixel gibberish, or are we literally talking about every possible image in existence for that format, as in one of those myriad images would effectively replicate the image of this letter being written?
If the latter, wouldn't you effectively become the possessor of all knowledge conveyable by visual spectrum or intelligent analysis throughout the universe?
I scare myself.
Most of the images would be gibberish. Very large chunks of the remainder would be extremely similar - change one 24-bit pixel of a single image and the difference won't be very noticeable, but you can get 16.8 million variants right there. Allow differences on the level of "Frame 100,000 of The Wizard Of Oz, except Dorothy's blinking" and the number of effectively-identical variants screams out into the zillions.
So, by discarding noise-storm images and functionally-identical images, you could reduce the real number of images you'd need to store by many orders of magnitude. Much of this discarding would require intelligent supervision, though, which raises a whole other problem.
If you had the whole dataset, then yes, you would indeed by definition have all possible information that could be conveyed by a 640 by 480 image.
Since there wouldn't be anything like enough room in the universe to store all the data (even after pruning it), though, this achievement would be a lot less impressive than that of the people who built your data silo!