Dan's Data letters #145
Publication date: 5 May 2005.Last modified 03-Dec-2011.
Can DSLRs be "classics" yet?
I wonder if I may ask a few questions about the RD-175 camera? I noticed you had written about it. I own one and like it very much, however the HDD tanked on it and I have not been able to do get it working. I am looking into using a CompactFlash adapter and a 128Mb CF card. Can you offer me any advice on this?
Matthew
Answer:
I don't think it'll work (not that I've tried it :-). However, PCMCIA Type 3 hard drives like the one the (1996
vintage) RD-175 uses can be had
very
cheaply, these days.
It's open to question how much life they've got left in 'em, of course, but it's not as if you have to pay $1000 to find out!
Raven probably doesn't have this problem
I work for the Alaska Volcano Observatory (AVO). We run a network of seismic instruments that extends throughout the Aleutian Islands. These instruments provide us with data that allow us to forecast volcanic eruptions. You might wonder why anyone cares if volcanoes erupt on (mostly) uninhabited islands. Well, it turns out that many trans-Pacific air routes go right over the Aleutians. Jet engines don't like volcanic ash.
Many of our seismic instruments operate from solar panels and batteries. Needless to say, the weather isn't always sunny in the Aleutians. Also, solar panels don't work well under three meters of snow. Anyway, the point of this letter is to ask your advice about lower power networking hardware, in particular switches or hubs. We'd love to migrate to IP based data telemetry (WiFi, etc), but running a hub or switch is just too costly in terms of power consumption. For a single device, we could use a crossover and connect directly to the IP radio, but using multiple devices seems to require a hub or switch. Any ideas about workarounds or leads on very lower power switches would be appreciated.
How low is low? Well, we engineer so that stations can run for 4000 hours (~6 months) without charge. We fret over milliamps.
Peter
Answer:
Hmm. Tricky.
After running through a series of dud ideas, I think a carefully chosen small switch and PCMCIA networking hardware - which you can use with laptops, of course, or with regular PC gear via a PCMCIA adapter, though that could create power consumption problems of its own - looks like being your best bet.
Small regular desktop switches can run from little batteries, but I don't know for how long. Running from a 9V battery at all, though, means the power draw can't be terribly high; 9V alkalines start being seriously unhappy from about 100mA up.
Switches exist, however, that can be powered from a USB or PS/2 port. PS/2 power is pretty darn weedy, though not clearly defined. Probably half a watt, tops, and those things run through a passthrough cable that runs a keyboard as well, so you could be onto a winner, there. It certainly wouldn't cost you much to pick one up and see.
I don't know whether the CompactFlash Ethernet adapters made for use in PDAs consume less power than the bigger PCMCIA versions (you can probably convert the former into the latter with a pin-adapter). CompactFlash versions generally carry ratings of something like "500mW max", but their real draw in normal use is likely to be way lower, well below 100mW. Apparently, though, it's normal for PCMCIA wireless adapters with power management features to consume an average power down around 10mW, if they're only lightly used and can stay in sleep mode most of the time. If your application sends one puff of data every few minutes, rather than a constant stream, you may well see numbers that good from PCMCIA 10/100BaseT adapters, but I'm not sure. There are PCMCIA Ethernet cards out there that're rated at as much as a third of a watt, even in sleep mode.
(Note: For your application, you could save a teeny bit of power by snipping off the power and activity LEDs!)
I think the lowest-power regular Ethernet switches still want more than half a watt per port - this is supposed to be some kind of record (though it's a big enterprise switch; the above 9V-battery info suggests that basic consumer chipsets, not at all useful to the enterprise market, suck less juice per port).
The 24 port SMC switch on my desk (no, I'm not using all of the ports; someone owed me money and I ended up with an unnecessarily large switch whose principal purpose today is keeping cats warm) is rated at 230V, 0.4A peak. It doesn't run nearly warm enough to be drawing all that all the time, but 15W or so would be perfectly believable.
I considered wireless networking for a brief moment, because you could run that in ad-hoc mode and not need an access point at all, but the cards draw more power and a blanket of snow would eat all of your signal. While I was hunting up info on that, though, I found a great example of the broad input power tolerance of well-designed equipment; the popular Linksys WAP54G is apparently happy from four to 15 volts.
Alternatively, you could forget about star-topology Ethernet altogether and go for bad old 10Base2 (I presume, perhaps incorrectly, that 10 megabit bandwidth will be plenty for you). But then you'd be using chipsets from years ago, that may well draw more power between them than more modern adapters do, even with a switch as well.
I also considered the idea of a high-capacity, reasonably portable battery pack that's cold-resistant and doesn't cost a million dollars; such an item would obviously solve a lot of problems.
For this application, none of the cheap or rechargeable battery chemistries are any good, as you no doubt already know. They all crap out at or mildly below freezing, and often have other shortcomings (self-discharge for NiMH, astounding weight for lead-acid...).
In situations where you can use the waste heat of the electronics to keep the power source warm you can use ordinary batteries, but in long-standby applications, or just really, really cold places, that doesn't work too well.
What's left is non-rechargeable lithium batteries. Lightweight, decent energy density, happy as clams well below zero (a commonly quoted temperature limit is minus 40 degrees, where the two scales meet up...).
Lithium-thionyl chloride D cells are the big lithiums I've encountered most often (which is to say, not very often at all); they give you 17-plus amp-hours at 3.6 volts nominal into a small drain.
So, 60-odd watt-hours, but they're about $US25 for a single cell, retail. No doubt a big purchase would bring that down a lot, and lithium primary cells all have shelf lives of many years, so you'd be OK buying five years worth at a time, and/or old surplus cells. But it'd still be a pretty big outlay if you need to get thousands of watt-hours worth of the things.
There are bigger models; a "long fat DD" lithium-thionyl chloride cell will give you well over 30Ah at about 3V and one amp, meaning around 90Wh, and probably work out cheaper per unit of energy, for non-trivial load current, than the smaller cells. But it still won't be cheap.
Military batteries could be an idea. The BA-5390/U lithium battery is the most popular military battery, apparently; it's a 330 watt-hour item (wirable as 15V with more amp-hours or 30V with fewer), bulk priced at $US100 per unit. "Bulk", in this case, means "hundreds of thousands of units per year", which is how many the US Army buy.
Startlingly enough, you can pay about the same price per watt-hour by buying lots of little batteries, though the idea's a bit silly. SureFire charge $US90 for a pack of 72 CR123A batteries, as used in various ferocious flashlights, cameras and so on; that's something in excess of 300Wh (123s are rated at about 1.5 amp-hours at 3V). SureFire also used to sell a nifty thing called the 12B battery pack light (mentioned here), which was 12 123s wired up as a convenient 6V pack. They've discontinued it for some reason, though, and obviously you don't want to be clicking zillions of little batteries into holders out of which they may well later pop as a result of thermal stresses, or something. Still, it's a thought, for small-to-medium energy needs.
Or you could go for an atomic battery. One the size of a refrigerator would run the station for decades, at no more than the price of a medium-sized icebreaker!
MTBF: half an hour
We have several Acer F1 Pentium 4 2.6GHz towers that must have inferior cooling fans on their CPU heat sinks. We have had three die so far, giving up the ghost with strange screeching sounds. Acer has been pretty good to send a new one each time we call, since they are still under warranty. But I would like to find a source for these fans where I could purchase a small bundle of them and have them on hand. I'm sure they aren't too expensive to the OEMs.
Do you know anything about a source like that for consumers? It is a FoxConn 2.75 by 5/8 inch square black plastic fan, attached with a screw down through each corner and into the aluminum heatsink. What I'd really like is a compatible fan from a different manufacturer, of course!
Russell
Answer:
"70mm" fans like this one weren't common in the PC world when the P4 first came out, but practically every standard
P4 cooler has one on it, so there are plenty of places that sell them now. Yours is a medium height version; there
are slimline and full-height 70mm fans as well, which you can attach to the same heat sink (provided there's enough
room above it in the case, of course) by just using different length screws (or any number of more gimcrack
methods).
Your fans have probably been dying because they have lousy bearings - cheap sleeve bearings, probably. I talk about extending the life of a fan bearing here, but by the time the fan's making a nasty racket, it's generally too late. What you want are fans with about the same power rating (the sticker on the fan probably tells you its rated current; current times voltage, in this case 12V, equals power), but a better bearing - which probably means a "ball" bearing, since high quality sleeve bearings are hard to find in the cheap-12V-fan market.
Electronics stores often only stock the more common fan sizes (40mm, 60mm, 80mm, 120mm), but plenty of specialist computer bits dealers have suitable 70mm models.
It slices, it dices
I am writing regarding your review of the Batterylife Activator. It seems like you have yourself convinced that in your all knowing nature that this product is not what it claims to be.
I purchased an Activator for my cell phone and since then several others. You didn't get results because you didn't use it properly, as the Batterylife people pointed out. I put it on my phone and noticed a HUGE difference. It performed even better than the ads said, I got an increase of battery life from 7 days to 14 days in just 2 charges. The phone speaker was also noticeably louder and clearer, and the processor seemed to be running faster like it's overclocked.
I immediately purchased several more. I have one on my cordless phone which does all the above plus enhances the range significantly. I put 2 on my car battery, and I don't know how the ions work, but I get significantly more horsepower and fuel economy. On the coldest day, my car will start like it's summer. Mind you, I installed the third one the casing of the car's computer, so that may account for the horsepower boost. I have also noticed that any product that is plugged into the cigarette lighter while the Batterylife Activator is installed also has significantly better performance. For example, my portable CD player will have better imaging, more acoustical accuracy and depth, better bandwidth and transient response, and can go up a lot louder while still having a reduction in subharmonic distortion. But it doesn't harm the car at all, the voltage is still 13.8VDC.
Either you fail to understand the nature of this awesome product and thus reject it, OR you are just paid off by the battery companies and who knows who else. Either way, you shouldn't deny your readers a chance to own something that is obviously a huge scientific breakthrough and is simply being repressed by the mass corporations.
Assaf
Answer:
I'm about 80% sure you're serious about this. Let's assume you are.
No, I don't think I'm all-knowing. I just tested the Activator, and it didn't work. That, plus the fact that its alleged means of operation could only be more ridiculous if it involved unicorns, and the further fact that the people who sent me the thing for review haven't yet been able to come up with any evidence that the alleged scientific support for the product actually exists - despite promising to get it for me - leave me confident in saying that the Activator is useless, until further evidence comes along.
Oh, and when someone pins the Batterylife people down, they get weaselly.
When Tom's Hardware, who have more energy than me, tried to track down the organisations that'd determined that the Activator worked, they had no luck.
So your quarrel, it would seem, is not only with me.
(Another voter has since been heard from, as well. Note yet another excuse from Batterylife for the product's non-performance; the casing of my battery was too thick, Tom's Hardware were told the product was a prototype {and then sent a press release about it being on sale...}, and IT Reviews were told that IBM batteries were a problem because they're "double sealed". Batterylife need to get their stories straight; they're getting pretty close to kettle logic.)
Batterylife said to me that the Activator wasn't made to work on batteries with a thick casing, despite the fact that the laptop version allegedly does exactly that. The Activator still didn't work when I put it right on the bare cells, with no casing at all, though.
People selling dodgy products love testimonials like yours. No matter how unlikely a product is, someone out there will not only be sure that it works, but be sure that it works in ways that the manufacturer didn't even promise. A gadget that's meant to "purify the blood" turns out to cure arthritis! A device that's meant to "block cellphone EMF" makes someone's stereo sound better! A gadget that's supposed to do something to lithium-ion battery chemistry turns out to work great on lead-acid, too!
The great thing about testimonials is that the manufacturer's pretty much in the clear, legally, if they publish them. They're not making any claims, they're just reporting what someone else said.
Have you run a benchmark to verify your faster processor speed, for instance? Of course not. You just say it "seems" to be running faster.
I remember, back in the day, when my colleagues at Australian Amiga Review got hold of our first Amiga 600, about which we didn't know much (this was in the days before Internet access was common, and well before the Web existed). We suspected it might have a higher CPU clock speed.
It didn't - it ran at the same 7.14MHz as the other PAL 68000 Amigas - but Amiga Review's editor was quite sure the thing was faster when he first played with it.
Then we ran benchmarks, of course. I suggest you try that.
As Richard Feynman (probably) said, science is a way of trying not to fool yourself. The first principle is that you must not fool yourself, and you are the easiest person to fool.
I know I'm about as foolable as the next schmoe, so I do my best to find out what's actually happening, rather than what I think, at first, is happening.
Regarding my being "paid off by the battery companies": Hmm, let's see, that brings your Credo total to Rules 9 and 12, possibly 22, and 28, after a fashion.
You've got a way to go, though. Good luck!
(My hopes of finally receiving the evidence that Batterylife so faithfully promised to send have been dashed, because the company has since gone out of business.)
Further constructive criticism
I wanted to take the time to tell you your research on Homeopothy is wrong. I've been using Homeopothy on myself and my family for the past 12 years and I am always amazed at how fast and effective it is. I've cured ear infections, respiratory infections, bladdar infections, sore throats, and many colds just to name a few using Homeopothy. Fortunately for my family and I, my Pediatrician who is a MD. researched and tried homeopothy on her children and now perscribes it daily in her practice. My neighbor who is from France said her Doctor and all Doctors in Europe perscribe homeopothy more than traditional medicine because it works! The medical establishment here in the US is very threatened by homeopothy because it is so effective and so cheap! Anyone like myself who has worked with Homeopothy will know how false and misleading your report is.
Sincerely,
Christina
Answer:
I'm not sure exactly what "report" Christina's talking about. Maybe this page.
Maybe this one.
Anyway, Christina: You (like, apparently, roughly one out of 4131 other people who try) can't even spell homeopathy.
(Note: Yes, maybe Christina's just dyslexic. But her mailer's AOL {astounding!} 8.0, which has a spelling checker. So my sympathy, even in that case, is limited.)
