Dan's Data letters #63Publication date: 30 September 2003.
Last modified 08-Mar-2015.
somebody was talking here about upgrading their PC, and mentioned a case window. Then somebody mentioned radiation, and got flamed.
I was told once that computers (perhaps it was just the power supplies) emit EMF or something, so that's why cases are all-metal. That's also why companies couldn't sell cases with windows and a power supply installed. I assume the metal also helps to strengthening the case, but can you clear this whole thing up for us?
Laws differ in different countries, but yes, PCs with windows often break radiation regulations. It's not a health issue, but an interference one; a "nude" computer, or a computer in a case with a big window, will emit more radio frequency interference (RFI) than is permitted in many countries.
Note that modifying your own case so that it exceeds RFI limits - by, for instance, just taking the side panel off - is generally perfectly legal, unless you manage to screw up someone else's TV reception or dialysis machine or whatever and they (or their next of kin...) complain. Computer RFI isn't powerful enough to be a nuisance for pretty much anything that isn't sitting in the same room, though, so such problems are practically unknown. People who try selling gear that exceeds RFI limits are likely to encounter more problems.
I talk about this a bit in my Lian Li review here.
Having just bought a fancy new mobile phone (Nokia 6100), I was shocked to see the price of the genuine Nokia phone-to-PC cable (i.e. phone-to-USB) is $AU170. I figure Nokia do this simply because they can.
Anyway, a retailer told me he carries the "aftermarket" cable, but this cable will not work with the software supplied by Nokia; you must go to a website and download a (shareware) equivalent of the Nokia software.
Can software detect a cable as not being genuine and therefore not work? This sounds weird to me, assuming the pinouts are identical.
The two cables aren't the same thing. As far as I know there is no clone of the DKU-5 cable. This "cable" is actually an active device with a microcontroller in it; the phone itself does not have a USB interface, and the after-market cable is something different.
There's much chat on this subject here.
I read your article on the LED caselight (thoroughly enjoyed it, by the way!). Anyway, it got me to thinking that I should do something similar, given that I have 100 each of red, green, and blue LEDs rated at 3000mcd.
My plan is to pulse between the 3 colours - not necessary all that original, but still, I like it. Now to the crux of the problem - I want this circuit to be as small as possible. I don't need help setting up the circuit itself, all I need to know is the brightness of the LED array you made.
More specifically, given the results that you obtained, what would be your best guess as to the minimum number of LED's required to light up a case (remembering that they're rated at 3000mcd)?
Millicandela brightness ratings are not informative, unless you also know the beam width. The tighter the beam, the higher the millicandela rating, for the same amount of actual light. I talk about this in my old review here.
Anyway, the number of LEDs you'll need depends on how brightly you want the case lit, where you're lighting it from, the colour of the LEDs (green looks much brighter than red for a given power, with blue in between, and orange-red LEDs look brighter than full-red ones), where they're pointed, blah blah blah. As a ballpark figure, though, you'd probably be happy with something in the order of 20 reds, 16 blues and 12 greens. Double that if you want something really bright.
Do you know of a good small (fast and functional) keyboard that does away with the stupid Windows keys (which I usually accidentally press during a heated game of Counter-Strike) and the power/wakeup/screwup buttons?
I'm happily using my $10 keyboard but I wish the footprint was smaller -sometimes I feel like hacking off the number pad and arrow key section because I keep on hitting the mouse on it when I play games.
The Happy Hacking Keyboard ain't cheap, but would probably suit you.
There's a cheaper "Lite 2" variant that you might also be happy with.
I've reviewed a few small keyboards that use laptop keyboard modules (here and here), but they don't feel great to type on. The Scythe Ergo Diver has better keys, but it also has a seriously bizarre layout.
I was wondering how to make an LED flashlight that uses AA batteries that are partly used (i.e. have an output of about 1.0 volts). With three children, I seem to have a ton of these.
It seems that an LED light with a voltage regulator circuit to "drain them dry" would be a perfect thing to get the last bits of energy out of these batteries. I hate to throw anything out before it's completely dead. These batteries do produce light in my LED flashlights, but not as bright as I'd like. If a light could be designed hold more lower voltage AAs than it would using fresh AAs, these orphaned batteries could do useful work.
If it was designed around a Luxeon Star, I would be in heaven.
The simple way to do this would be to make a basic LED flashlight (batteries, LED, current limiting resistor, switch) that will run an LED moderately hard (35mA, say, for a single 5mm white LED) if you feed it fresh cells, and then just stick old cells in it. It'll still give worthwhile light down to quite miserable input voltages, if you make it use at least four cells.
The problem with this is that you'll need AAs that're in a similar state of discharge, and you'll also end up with a fairly chunky flashlight for its light output. If you used a Luxeon LED, it might be significantly brighter and more efficient, but you really wouldn't get much that you wouldn't get from a plain old 5mm LED.
Yes, there are step-up circuits you could use here, to let you use one AA at a time, but they'll kill the battery pretty quickly. These circuits are often used to allow people to substitute a nice beefy D cell or two where a 9V battery would otherwise go, giving far better battery life even after you take the inefficiency of the conversion into account. But you don't get something for nothing; higher voltage out means higher current in, and the current delivery capacity of worn out AAs is pretty miserable.
Here's an example of such a circuit.
The most elegant solution, I think, is just to buy a single-AA flashlight with a step-up circuit already in it. I've had a CMG Equipment Infinity Ultra sitting on my to-review pile for far too long now; it's a much brighter version of the original Infinity (reviewed here), and it's pretty cheap, for what it is.
Amazingly enough, this idea has occurred to others.
I recently bought a Hercules Muse DVD 5.1 sound card. It's good, but it is only connected to a pair of (quality, ear covering) headphones. Apparently the "XeaR" 3D positional audio only works with "open air" headphones, which I think means "not the type I'm using now".
So I was hoping to connect some old stereo-system speakers (bare wire) to the sound card. I don't think it's possible, but is there any way to do this? Or will I have to fork out on a proper setup? I have a lot of spare cables and 3.5mm jacks lying around to use if it is possible.
Different kinds of headphones have different response characteristics, but basically, closed-back headphones just block external noise better. Whichever kind you use, each ear can't hear any significant amount of the other ear's signal, so there's no reason why this difference should mean anything for positional audio.
As regards connecting speakers - in the olden days, sound cards all had weedy little amps on them, and could drive un-amplified speakers to better-than-nothing volume. Modern sound cards only have line outs, though; they can generally manage substantially higher than normal line voltage and thus drive headphones perfectly well, but you won't be able to get worthwhile speaker volume out of them. Fortunately, you won't need to; your current headphones should be fine.
Is this for real, or just something that the marketing pukes dreamt up? I suspect that the batteries will be killed after just a few charges.
Apparently, Rayovac's new 15 minute NiMH charger does work (when used with their special cells, anyway), but the fan ain't that quiet, and you have to wait for the cells to cool a bit before using them, so it's more like a 20 minute charger.
There's no data yet on the longevity of the batteries, but Rayovac aren't cheating by using more durable low-cap cells; the cells are have 2Ah capacity.
"Eight hours", of course, hasn't been the state of the art for NiMH charging for some time. Maha and Rayovac and probably someone else by now make one hour chargers for normal NiMH cells.
The Rayovac site doesn't seem to have much about the new products on it yet, but there's some bumf here.
I would like to make LED flashlights by modifying Maglites. I am a little concerned that Mag may try to burn me. Is there a legal way to make mods of their light and sell them?
Mag have, I'm told, been fairly heavy-handed with some retrofitters of their lights, taking the position that just using the word "Maglite" is a trademark infringement, even when there's clearly no real attempt to pass off a retrofitted light as a real Mag.
Obviously you deserve to get a legal whuppin' if you sell your products as "Luxeon Maglites", or something, but there shouldn't be anything wrong with saying "This is a Maglite(TM) which I have converted to use..." et cetera, with a disclaimer "This light is not a Mag Instrument product, is not subject to any Mag warranties or promises..." blah blah.
I Am Not A Lawyer, and I realise that practically speaking there's not a lot you can do if they actually follow through with a threat to sue, or just get eBay (or whoever) to pull your auctions. EBay is frequently very happy to do this for no very good reason; see here, for instance.
If you take the above precautions, though, you ought not to get a nastygram from Mag at all.
I enjoy photography very much, and of late find that I have been moving away from film and more into digital for the sake of convenience and other things like proper exposure control, indoor lighting, etc. I wouldn't have said this a year ago, but now it is obvious that digital is firmly entrenched in the pro segment too. I found your review while looking for a digital camera that would replace my Nikon SLR. Seems like the D60 is it.
What I really wanted to say is that I found your article extremely amusing and informative about technical aspects without being boring. I read it from start to finish in fascination and when I finished I learned much more about the "high end digital cameras" than I had from reading twenty digital photography magazines. I found out what I wanted to know and had several good laughs while doing so. Thank you!
Note that the EOS-D60 has been superseded by the EOS-10D since I wrote that review. The 10D is a better camera than the D60 in several respects, and has a lower list price; they're basically the same and it'd be madness to upgrade from a D60 to a 10D, but the 10D is the better choice unless you can find a mint D60 second hand for a good price.
Note also that Canon have also recently released the startlingly cheap EOS-300D, which lists for only $US999 with a lens That's cheaper than several integrated-lens "prosumer" cameras!
The lens the 300D comes with is a special one that can't be used on other EOS cameras, but it's a decent quality general purpose zoom with a good wide angle end. You can also get the 300D back by itself, even cheaper.
The downside: The 300D's in huge demand, and I don't think any have even come into Australia yet. The first few shipments will probably all be pre-ordered. Dirt Cheap Cameras are accepting $AU500 deposits for the 300D, but don't even have a price up for it yet. Expect it to be at least $AU1600 with lens, and probably $1700 or so.
In the meantime, they've got the EOS-10D back for $AU2769.
I noticed an interesting phenomena a few days ago when I was transferring data to a new hard drive I had bought to replace an aging IBM ATA100 unit.
While transferring the data, I had the new drive sitting on a box next to the PC with the side cover removed to make it easier to swap cables around, etc. For no good reason other than to fill in time while I waited for the transfer to take place (more than 12GB) I clicked on MBM5 to check the CPU temp while this was going on and noticed that it was up more than 8°C above any previous 100% CPU load test that I had carried out before. After the data transfer had completed and the new drive checked out ok, I reverted everything back to how I wanted it and replaced the side cover. As an exercise, I transferred a block of data of approx 5GB in size back and forward between the hard drives while monitoring the CPU temp again. After about half-a-dozen transfers, the CPU temp had not moved from where it typically resided with the case fully closed - about 8°C below where it was with the side cover removed.
I've got an Enermax 710-B case with two 80mm intake fans and rely solely on the PSU fan(s) for exhaust. This combination has served me well for quite a while now and is reasonably quiet. I use round IDE cables with all other cabling loomed up tidily to minimise clutter and internal air-flow resistance. This accidental experiment seems to belie the wisdom of running your PC with the covers removed, provided your internal case air flow is adequate. My case's internal air temp sits about 3-4°C above that of the CPU, since the CPU is water cooled and rarely rises more than 10°C above ambient regardless of load. Anyway, it proved to be an interesting accidental exercise.
Indeed; a case with good throughflow ventilation will, by and large, run cooler with all panels attached. Some bits will be cooler with the side off, but the important stuff will probably be happier if the box is complete. Intake and exhaust fans don't do much when one side of the case is missing.
If a case has crappy ventilation, though - no throughflow fan but the one in the PSU, unbundled cables all over the place, a more-steel-than-holes intake vent - then taking the side off will help. It's a quick fix for people who find their plain old beige-box computer goes flaky on the three super-hot days in the middle of summer. And my standard prescription for detecting ventilation inadequacies, mis-installed CPU coolers, dried thermal compound and other cooling shortcomings - remove the side panel and fire a desk fan at the works - remains valid.
Here's something so obviously dishonest I just had to send it to you for your enjoyment.
Recently a friend of mine was interested in replacing his aging 300W PSU in his main PC Tower with something a little stronger. As he has an account with a local distributor he contacted them, and was given a price on a 400W PSU at $AU99, or they could supply a midi-tower case complete with a 400W PSU for just $AU49!
Well we both thought $AU99 was pretty good for a 400W PSU (my Antec TruePower cost a LOT more) and really had our doubts about the quality of the case and PSU package.
Anyway - the only way to find out was to buy it and see!
Upon opening the PSU I took one look and said to him "This sure doesn't look like a 400W supply, there's bugger all in here". But the specification label says 400W, so it must be true, mustn't it?
Then I did the sums and according to the specs the maximum rating is just 284.7W!
Can you think of any legitimate way of including that line on the label which says 400W, because I can't (unless they rounded to the nearest 400). It looks like it isn't just quality that costs more, obviously honesty comes at a price too.
By the way, the case pretty much sucked too; a shame it couldn't suck as much air through it.
Your sums are correct. And that 284.7 watt total is including the negative and standby rails, which the most honourable manufacturers leave out of the power calculation.
This rating wouldn't be scandalous if the PSU were a cheap and nasty "300 watt" unit, and it'd be honest for a 275 watter. Calling it a 400 watter is simple contempt for the customer, though.
The rail ratings may or may not have anything to do with reality, of course. Maybe this PSU can't even manage what it says on the sticker, and maybe it can actually do better. Either way, I wouldn't trust it for anything vaguely important.
Jumper the green wire on the ATX connector to ground to power it up, and use it to run some CCFLs at your next party, or something.