life span durable ground based 'satellite' device.

[[email protected]]

The "time capsule" idea has some problems; for example, you wouldn't

know if the device stopped working a few days after you put it up,
and as others have mentioned, could lose all the data at any time.

True. I was thinking of some kind of a malfunctioning alert by a radio
transmitter or a Led light signal.
There always will be a change that some part will break down or a case
of severe/all data loss.
That's reality, but also the challenge.
To reduce the risk of malfunctioning there will be some parallel
connected spare/back-up parts. (instead of one hd, 3 hd's).
If hd_1 breaks down the next recorded data will be stored on hd_2. So a
simple piece of error detection software is also needed.

As far as lifetime goes, proper electronics could last several
decades, maybe a century, without any heroic effort. There are
scads of 1940's tube radios, 1960's transistor radios, and
1970's computers that still work just fine in spite of being
badly mistreated over the years.

I agree. As long as the solid state circuits aren't broken, split,
disconnected or oxidiced, electronic machinery, like your mentioned
examples, can run for a real long time. I still have the feeling that
the main issue with the time capsule concept is, power/energy.

However, unexpected problems could shut your thing down at any
time and it seems to me that you should expect your first effort
to work less than a month before failing, your second one to work
less than a year, etc. What climate would your device be in?
Will it be sheltered, or out in the dust, rain, snow, wind, and
freeze/thaw cycle?

Moderate climate, and probably no shelter.

For the long run, I think you would need to build it with no
moving parts, which in my opinion rules out wind power, disk
drives, motors, pumps, flywheels, automatic cleaning or active
aiming of your solar panel, etc. Except for Edison cells, most
batteries aren't worth considering. Super-capacitors don't have
a long-term proven track record yet, but probably are worth
considering. Anyway, all this said, I suggest you use an RTG
(radioisotope thermoelectric generator), or a solar-powered TG.

Who's got the e-mail adress of Jay Leno?
"Jay Leno's 1909 Baker Electric still operates on its original Edison
cells." (!)
(http://en.wikipedia.org/wiki/Battery_electric_vehicle)

Also Supercapacitors are an interesting option, though hard to obtain.

Thanks for your suggestions

 

[[email protected]]

Is this building abandoned? I'd think you could schedule at least a yearly

walk-around - I know if I owned a building, I'd try to maintain it.

The Lord owns the building.

It's a (former) church eh..eh..eh :Q

 

[[email protected]]

Use a GSM phone. The have a packet radio mode which sends the bytes
fairly cheap. You can also have a two way communication this way.

What if there's no GSM network after let's say 10 years?

 

[[email protected]]

Not wanting to sound like a "Free Energy" kook,

Don't worry, I'll be the kook here. ;-D

but you might want to look
into how much receive power you could siphon from local AM radio stations.
From the numbers you've provided, I came up with you needing less than 300
microwatts continuous (obviously you'd need some charge/collection
circuitry). How big an antenna can you tolerate?

Huh? Is this for real? If so, "Free Energy" indeed.
Like that Lucille Ball story?
http://www.healthyhearing.com/library/ate_content.asp?question_id=102

Do you have some more specific information about the workings of this
idea?

A big antenna wouldn't be a problem.

 

[[email protected]]

From my past in aerospace electronics (including mechanics) mechanical
components have much better defined reliability than electronic parts.
Just notice the work being done by them in cars.

Maybe so, but moveable parts need lubrication right?
Imagine a bicycle non-stop moving around, outside in a open environment
and weather, for let's say 25 years without lubrication. Theoratical,
with no accidents etc. If it still moves after these years it must feel
like a very, very high gear I think. However a telephone cable well
isolated in the ground (maybe not such a good comparison though) will
still work fine.

So a solution, when working on a "satelite" is to have experts in every
possible technology involved, and hearing pros and cons.

True.

 

[[email protected]]

Well you've got a heck of a problem here. Typical electronic stuff

these days isnt all that well designed to last. They're chock-full of
unreliable components-- switches, connectors, LCD displays, elastomeric
connectors, zoom lenses, motors, rubber parts, plastic gears,
lubricants, electrolytic capacitors, eproms, flash roms, and probably
more. There's no guarantee the camera will last for any length of
time.

I'll be using a stripted down camera or making self a primitive version
out of very basic/reliable components: Fixed wide angle lens, rugged &
good quality CCD/CMOS, I/O connections etc.

No big LCD's, plastic parts and rubber gears

Even if you buy an expensive, perhaps mil-spec "hi-rel" camera, how do
you test it and the rest of the equipment to verify it will last for
many years? You probably don't have ten years to waste testing the
stuff in real-time, plus at the end of that time, none of the parts
will be available, so you'll have to start over qualifying the
components.

True. That's why I prefer to make one (digi cam) myself.

I suspect hard disks are out of the question-- all those moving parts
are unlikely to stay well lubed for more than a few years.

I agree. But there are some very, very reliable Hard Disc's. A few
seagate's have an incredible high (1,400,000 hours) MTBF (mean time
between failures).

 

[[email protected]]

Solar cell reliability varies widely depending on type. At the high

end are single crystal cells with acrylic encapsulation on the back of
glass. Good luck finding one. At the low end are first generation
polycrystaline cells silicone encapsulated on the back of glass. This
type is readily available, but deteriorates rapidly (several percent
per year IIRC) due mainly to water diffusion into the grain boundaries
creating shorts and also slightly due to light induced yellowing of
the silicone. Prime new second generation polycrystaline cell panels
have passivated grain boundaries and have switched back to the more
difficult acrylic encapsulation; these are supposedly almost as good
as the old single crystal versions, but they are not generally
available in small quantities yet since production is not yet able to
meet the demand from the few system installers with buying agreements
in place. You might be able to persuade a system installer to sell
you a second generation panel but you will not find any for sale from
distributors that sell panels yet and they are only made in relatively
large sizes. So for any small panel you can get easily you should
plan on several percent per year reduction in capacity.

Glen

Thank you for your extensive information.

There's still a lot of development in the solar cell technology. Just
received this article:
http://www.physorg.com/news79796665.html

 

[mark]

Stanislaw,
I think that mechanical assemblies have their place. However, with
electronics at least you have the possibility of elimination of wear-out
due to mechanical motion (yes, I know there are inherent wear-out
mechanisms in electroncs, such as metal migration etc)

We all know what the solutions are: that there is no easy answer. Each
situation must be examined: as engineers we should be able to identify the
options, and come up with a least a ballpart ans. Are mechanical motion and
wear limiting reliability? will inherent electonic wear-out limit it? Can I
package the electronics in a hermetic chamber (and what is the leak rate?)
to prevent moisture migration?
TANSTAAFL. you have to do u'r homework
(But the rules of thumb we use sure are handy at times)

M Walter

 

[Christopher Ott]

I remember building a crystal radio once as a kid, and how it was able to
power a small earphone with no battery or apparent power source. The power
transmitted by the station powers the radio. Considering the small amount of
energy you'd need, it seems like you could adapt this principle. It does fit
your main criteria of a nearly indefinite power source. I don't expect AM
radio to be going away anytime soon. Of course this was also the principle
behind Nikola Tesla's Magnifying Transmitter (
http://en.wikipedia.org/wiki/Magnifying_Transmitter )

You only need 5 Watts for 5 Seconds (25 wattseconds) once a day. So the
other 86,395 seconds of the day this thing is just sitting there. It seems
if it could be drawing just a trickle ( approx 300 microwatts continuously)
from somewhere you could get your 25 wattseconds of energy for the next days
readings.

Perhaps some of the folks in the radio newsgroups could give a quick thumbs
up/down on the concept, as I don't know how much receive power would be
typical off the AM band.

Also, you didn't mention it, but you will have a small constant current draw
from the real time clock which will be needed to tell the microcontroller
when to come out of sleep mode.

Chris

 

[Ancient_Hacker]

But there are some very, very reliable Hard Disc's. A few
seagate's have an incredible high (1,400,000 hours) MTBF (mean time
between failures).

If you believe that, I have a whole lot of laptop batteries to sell
you.

1,4000,000 hours is 159.162341 years, give or take a leap day. They
obviously didnt run a batch of disks for that long. They probably ran
10 disks at elevated temps for xxx hours, had one failure, and
extrapolated the MTBF from that. Not exactly something I'd rely on.

In typical use, hard disks seem to last about 5 years before the
bearings start to dry out and get noisy.

Nobody knows how long the roms or electrolytics on the disks are going
to last.

Plenty of older EROMs only hold a charge for 10 years or so. Nobody
know how long surface-mount electrolytics are going to last. The
manufacturer's specs are not too encouraging-- the specs only guarantee
a few thousand hours (see www.digikey.com).

SMD electrolytics are already prematurely failing on hot boards, like
PC video cards.

 

[Rich Grise]

The Lord owns the building.

It's a (former) church eh..eh..eh :Q

I know, the Lord owns everything (actually, it's his Yin, The Mother of
Everything, who owns everything), but who has the title deed?

Thanks,
Rich