Dan's Data letters #123
Publication date: 30 August 2004.Last modified 03-Dec-2011.
The poor man's SSD
What's faster, portable flash memory, or hard disks? Given the limited read/write life of flash memory, would it be feasible to use a 6Gb CompactFlash card as a hard disk replacement?
Jon
Answer:
Flash RAM has near-zero seek time, but lousy maximum transfer rates (a CompactFlash card with a pin adapter can plug
straight into a PATA cable, but will only support PIO transfers, not DMA of any flavour).
You can use flash memory as a hard disk replacement, as long as you don't put a swap file on it. It's perfectly feasible for no-swap systems, which means you can even run Windows if you use 98SE with virtual memory turned off, or something. Flash memory is a natural for lightweight Linux/BSD systems, of course; desktop Linux needs swap, but if you're just rolling your own network appliance, a near-free 8Mb CompactFlash card makes a great boot device.
I've written about this sort of thing on a couple of previous occasions.
Slowness, and PC-TV
I use iTunes for Windows for my music collection because I like it. But whenever I use it to rip a CD, my whole PC becomes molasses. Is this iTunes, or the way I've connected the drives? It didn't happen before I lost a hard drive and I re-jigged my PC. Both my CD burner and DVD player are on the same IDE cable; is this the problem? iTunes used to be able to rip and play the CD at the same time with no problem. Now it can't.
Also, when you review video cards, the talk generally revolves around games. But long before all this "Media PC" guff, I wanted to use the PC as a substitute video recorder. Is the video card the more important feature to getting good picture, or the TV tuner card?
Dave
Answer:
My first [and, as it turned out, correct!] guess would be that you don't have DMA turned on for the CD drive. You
don't mention which Windows flavour you're using. In Win95-series versions (95, 98, ME) you go to the Properties panel
for the drive to change DMA settings; in Win2000 and WinXP you go to the Properties panel for the appropriate
channel ("Primary IDE Channel", "Secondary IDE Channel") of the appropriate controller. Just fish around in
Device Manager until you find the appropriate one, and try clicking the DMA box or selecting "DMA If Available" (depending
on Windows version) for everything in sight. If it doesn't work, Windows should just restart with the drive back in
PIO mode - nothing, despite Windows' dire warnings, should actually stop working.
Different video cards with TV out have different image quality, but pretty much everything on the market at the moment has OK quality. Problems more commonly arise with image alignment; if your TV doesn't have full image size and positioning controls (which most don't), some cards won't let you get a nice full-screen image without black borders. I think all of the mainstream ATI and Nvidia cards these days are OK in this respect, though; the drivers from both companies have plenty of TV image tweakability.
The tuner card is important too, of course, but if you've got a good regular TV image then you should be OK. Note that the tuners in digital TV tuner cards may be rather fussier than an ordinary digital set-top box; don't assume that because an STB works fine for you that a digital tuner card will as well.
In my day, we used a crank
I've gotten myself into a few situations where my brilliant self tends to leave the headlights on when I get home. I was curious, would it be possible to charge a car battery, removed from the vehicle, with a computer PSU's 12V rail? Say I use the "wire trick" and bridge the green wire to ground to make an ATX PSU start up on its own, then touch the 12V to the battery positive and ground to the battery negative, how long would it take to charge the car battery to a reasonable amount enough to start a car? A few hours?
Gregory
Answer:
Yes, this'd probably work - and probably take less than an hour - but you couldn't charge the battery completely this
way.
A proper charger for 12V lead acid batteries needs to feed them, classically, 13.8 volts (it's more complex if you want to maximise battery life - I talk about that here).
A mere 12V supply will, however, be able to throw enough energy into the battery that you'll be able to start the car. I know this for a fact, since I once did it to my own car using a cheap 7Ah sealed lead acid "brick" battery. The trick there is to connect the freshly charged brick (a freshly charged lead acid battery will have a bit more than 12V terminal voltage) in parallel with your flat car battery; the battery can still be in the car, as long as the ignition is well and truly off. Leave it for half an hour, then disconnect the brick (which can't deliver anything like enough current to start the car, and will become rather unhappy if you ask it to), turn the key, and you'll probably be in business. After that, the car's own alternator can of course throw more charge into the battery.
You could do the same trick with a ten-cell NiCd or NiMH pack; with a really heroically enormous NiCd pack, like ten ten-cell model plane packs in parallel, you could probably even start the car directly without incident (you'd be wanting to unplug them smartly afterwards, though, as they wouldn't much like being charged by the alternator). Various other unlikely 12V sources would also do; in theory, a phalanx of carbon-zinc AAAs would be adequate, but wiring 'em up would be like harnessing 10,000 mice to a cart.
The main problem with using a computer PSU to get a car battery into this just-charged-enough state, though, is that the car battery may suck more amps than the computer PSU can deliver, early in the charge. This should only be a problem if the car battery's really flat, though; if the car's been sitting overnight with the headlights on then it will be, but if it's still got over ten volts then you'll probably be OK. A really high wattage computer PSU would also help, of course; old 150W AT supply, no, new 580W oversized ATX supply, yes.
Note that the flatter a car battery has been, and the longer it's stayed that way, the more likely it is to have been lethally sulfated and need either replacement or treatment with one of those magical desulfators which I still haven't gotten around to looking at in person. In a pinch, though, even a severely sulfated battery can usually be coaxed into starting the car a couple more times, which'll be enough to get you to the garage.
The tongue is always an option
Is there any way to use a multimeter to determine which hole is active and which is neutral, in an AC outlet?
Assuming the answer is what I think it is (connect each to earth and check the reading), is there any other way to do it, when there is no earth connection nearby?
How hard could it be for multimeters to include an "AC mains tester" mode?
M.R.
Answer:
Your guess is correct; you check each one to earth in AC volts mode with the outlet turned on, and the active one
will show X volts, while the neutral should show zero (but will probably show a bit of leakage voltage - if it shows
a lot of volts, you've probably got some kind of ground fault).
Here in Australia, absence-of-earth isn't an issue, because all normal outlets here are earthed. Without an earth, it's actually theoretically impossible to tell the difference between active and neutral just by touching both of them with probes, because all the probes can detect is the potential difference between the two, which in AC looks the same whichever way around you probe it.
Note, however, that you can provide a low-grade earth yourself just by holding the tip of the other probe in your hand, while poking the other probe into the socket. This is a pretty capital-D Dumb thing to do with a multimeter, because you'll zap yourself if your multimeter is in a low-resistance mode (like, one of its current-test modes), or if it's in voltage mode but has an (improbable) fault. You'll probably survive touching just the active terminal directly, mind you, as long as your body isn't actually connected to neutral or a good earth (bare feet on wet ground...).
To avoid the whoops-I'm-in-amps-mode-now-I'm-dead problem, you can buy little cheap pocket-clip AC-tester screwdrivers that contain a high-value resistor and a small neon bulb (or, nowadays, even a not-very-reliable LCD voltage display!). Poke the end onto a live wire with your finger touching the metal ring on the other end and the bulb lights up, without any chance of electrocution. There are also fancier mini-probe gadgets that combine this feature with various others.
Quieting a pizza box
As is the wont of all those with a lust for UNIX geekery, I've come to own a SparcStation 20. While the build quality and engineering of the SS20 is remarkable, the same can not be said for its acoustic engineering, which leaves both a great deal to be desired and a loud ringing sensation in the ears.
The primary culprit is most certainly the SCSI hard drive (Seagate Hawk ST32430WC) wedged into this machine, which performs the traditional imitation of a jet engine hitting its straps. This issue I hope to resolve with a new drive just as soon as Seagate tell me which model I should be looking to buy. Beyond this, though, there are three fans in the system which, while currently masked by the drive, are also far from whisper quiet. One is a dinky Nidec Beta V TA150DC (C33830-16) which cools the drives, while the other two are inside the PSU and are both Nidec Beta V TA225DC (M33515-55) fans (which should be here but aren't, an anomaly I've asked Nidec about but am yet to receive a response on).
I'm having a hard time finding information on quiet fans for this system. Most to all of the quiet computing information is highly PC-centric so fans that don't feature prominently in this world are simply not mentioned with regards their acoustic performance. Are you aware of any fans that could replace these three that could minimise noise?
I should also point out that the system is currently furnished with a dual HyperSparc CPU module running at 90MHz and the PSU fans are responsible for cooling that toasty fellow, which will hopefully be joined by another of it's brethren to make a quad CPU system (just as soon as I can find one on eBay that is...), so I can't really use a low power model lest I inadvertently poke the gremlins of thermal instability. Any help you could offer would allow me to retain full hearing for a markedly longer period of my life.
Andrew
Answer:
I don't think you're going to be able to make it a whole lot quieter if you just replace the existing fans with a
couple more 60mm units and a 40mm with similar specs to the stock fans. The only noise you can eliminate by doing
that is buzzing from old dying bearings.
However, if you don't mind taking to the thing with a jigsaw, you could hack a hole in the top of the pizza box case to suit a big old 120mm fan - which'll have to poke out of the top of the casing. 35 to 40 cubic feet per minute of airflow from a top-mounted fan should match the flow from the standard fans; even a slimline 120mm fan should manage 80cfm at full power. It won't be terribly quiet at full power, but it'll probably be less annoying that half as much airflow from small fans, because it'll make a lower pitched noise. If you wind down its speed (with any of a number of speed control solutions - a commercial speed controller, or an in-line resistor or Zener diode chain, or maybe just running the thing from the five volt rail I possibly foolishly imagine the Sparc PSU to have...) until it only moves 40cfm, you should have a quite acceptably quiet solution.
Alternatively, you could try ducting the air coming out of the box (which is probably where most of the noise is escaping) through a sound-absorbing muffler (cardboard lined with packing foam will do). Just squish a blanket up against the box vents for a moment to see the maximum damping you can expect from this strategy.
It's a Sony! Wait - no it isn't.
A friend of mine in Brazil tells me that I can get 3600mAh Sony AA NiMH rechargeables from stores and Web sites in Brazil, like this one. I ran quite a few searches on the Internet and the results always led me to Web sites in Brazil. No other countries have any ads for these batteries. I also looked for 3600s far and wide in electronic stores in my (US) State but came up empty-handed. Even Sony's battery site does not show these batteries. Their highest seem to be 2100mAh.
My friend has a good digital camera and swears that these Sony 3600mAh batteries are the thing to use and have been available for months in Brazil now. What do you know about this? Should I try them out?
I've also been reading good things about Maha external battery packs. since my camera is 2.4V, (2 AAs) can I use higher voltage external packs into the dc-in plug? If so, which voltage is ideal for taking lots of photos, good zoom, live view, interval recycling between shots, etc?
Arsenio
Answer:
My first guess was that these "Sony Cycle Energy" cells aren't a real Sony product at all, or if they are, then they've
been misleadingly relabelled for the Brazilian market. Perhaps there was some Brazilian legal loophole that allows
the capacity of two cells in parallel to be quoted, or something. I suspected they were just a knock-off, though.
Arsenio chased the story for himself, though, by calling the Sony support number...
...on the back of the package (which, a reader's now pointed out to me, contains that rarest of creatures - a misleading graph that makes a difference look smaller than it really is. The 72% figure's almost exactly right, though!).
As soon as he asked about 3.6Ah AAs, the Sony rep mentioned that they don't make them, and that yes, someone's selling fake batteries in Brazil.
Indeed, the only actual current loose Sony NiMH AA cells are plain 2100mAh cells, with the part number NH-AA-2DA. A reader's now pointed out to me that they also sell bundle packs of four 2300mAh AAs and a charger; the packs have part numbers BCG34HD4 and BCG34HRD4. They're included with their new Hi-MD MiniDisc players. If you're waiting for actual 3600mAh NiMH AAs from Sony or anyone else, though, you'll be waiting a while - possibly forever.
Regarding external battery packs - there'll probably be some input voltage latitude, but you shouldn't exceed the rated voltage of your AC adapter by much. Cameras that use a simple diode for reverse polarity protection on their DC input will lose some voltage there - so a 6V camera may have a plugpack that outputs about 7V, say - but it's generally OK to run a 6V device from 7.2V anyway. Not always, but generally. There's no way a 6V pack will be OK with a two-cell camera, though.
Most cameras get more volts from their plugpack than from their battery pack, which may give them slightly faster zoom and brighter displays, but the difference is seldom major.
The Maha external packs are aimed at cameras that have low capacity internal batteries - probably very small lithium ion packs - because they offer only 1.8Ah (for the NiMH pack) and 2.2Ah (for the Li-I pack) capacity. If your camera runs from four AA cells, then another set of four new 2Ah-plus AAs will give you about the same capacity as one of these external packs.
There are other, larger capacity external pack products that'll save you from having to make your own, but they're bulkier than the Maha packs.
Still waiting for the strapless version
I'm interested in a purported cordless anti-static wrist strap that uses the "Corona principle to harmlessly rid your body of static electricity before it can damage your data or computer equipment".
I've Google searched "Corona principle" and "Corona effect" but only found articles about electrical discharges around high voltage electrical lines. Do you know what the Corona principle is, and would it allow such an anti-static device to work?
Gerald
Answer:
If you're charged up enough that your body is exhibiting significant
corona discharge - even from a pointy probe held in your
hand, much less a smooth wrist strap gadget - then you're way, way beyond what's necessary to cause
electrostatic discharge damage to computer hardware.
So, as you might expect, that cordless anti-static strap worketh not.