h a l f b a k e r yThe leaning tower of Piezo
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Having dealt with communications gear, when things go
broken, things change electrically, and thus thermally.
Having a device which has been thermally signatured can
allow for quick fault localization, not having to probe in
possible locations.
A thermal camera would take a signature of a
device
within its functional operating condition, as well as
signatures for various settings within the device (AM, FM,
CW, etc.)
Being able to see which components heat up and which
components do not can allow you to see which signals are
making it through your device, where they stop, and which
settings correctly adjust signal properties/destinations.
Allowing you to troubleshoot high voltage/high
temperature devices without reaching your hands inside
will make it a safer troubleshooting experience as well.
This system would be useful for critical systems such as
cellular communications towers, ship-board
communications systems, and space operations, where
time is definitely critical.
For Example
http://www.flir.com...graphy/americas/us/
Just look at some of the uses shown in the ads on this page. [DIYMatt, Dec 25 2012]
[link]
|
| |
This is pretty widely used already. Most of the advertisements for thermal cameras show them being used for diagnosing home heating and cooling problems and sometimes mapping heat sources in computers. |
|
| |
If not widely known to exist, it should be. Sometimes it's very easy - just look for the scorched place on the circuit board. |
|
| |
I was musing about the difficulty with looking at a
card 'end
on', so to speak, when I thought: if the card was
cunningly
painted with different thermochromic paint in different
places so that at running temperature the whole thing
showed the same colour, then it would be possible to
see
an abnormality by eye...by mirror as well. |
|
| |
Of course, the magic would be using a trainable paint, so
you simply said 'at this temperature: be white!' or
whatever. |
|
| |
When transistors fail, well, at least for me, they
rarely blow up. This would allow me to see which
ones are cold, which should be hot. This is not just
FLIR, it also incorporates a database of "normal"
thermal signatures, which can easily be cross-
referenced. |
|
| |
The problem with trying to record 'signatures' of devices is that
that is going to vary a lot under different load conditions, e.g the
output stages of an audio amp. And what if devices share a
heatsink? |
|
| |
For most devices it would be possible to run a test
program that would set up a steady thermal state (or
at most a couple of alternating ones as some things
cycled on and off). It wouldn't be perfect, as no two
components are exactly the same, but it would give
an indicator under many cases. |
|
| |
That being said, the difference between a circuit
that is normally idle, and one that is burned out open
would be largely undetectable. |
|
| |
I'm just glad to see that people are still interested in
repairing things at the component level. I was under
the sad impression that "repair" these days just
meant replacing the whole board, if not the entire
item. |
|
| |
well when dealing with mil spec, we replace down to
the component level. And what [8th of 7] said
about loading, take different images when in
different settings. Allowing you to see what turns
on and doesn't, according to the setting, will give
you much to go on. |
|
| |