Restart after power loss

Johan Stäck wrote:

The problem is that all cameras these days have USB ouput.
It will be diffucult to have the PC and the camera more than a couple of
meters apart, due to USB cable length restrictions.
So, I plan on having them in the same enclosure.

The specifications for USB cable length are for powering
a remote device at the maximum load current allowed.
That is a relatively short distance. Of course if you
are not powering the device from the USB port, the
distance over which data can be sent is significantly
longer!

I wouldn't doubt that it would work at some significant
distances, but for a unmanned remote it probably should
be limited to a few meters. Hence, if the camera is on
a tower adjacent to a building, go for it. If it is in
a building 30-40 meters away... that's likely too far.

However, I will certainly consider having a UPS standing at some
distance from the camera/PC enclosure.

A sharp power-down won't though. If the power fails
while the computer is in middle of a disk access, it's
possible (not likely, but possible) to have file
corrupted or worse, an unstartable system... or
ultimate worst-case, damaged hardware. A power SURGE
or brownout is even more likely to do damage.

True, but when running NTFS, I think that the risk is rather small.

But, if possible I will try and install a UPS at ground level.
The cameras will mostly be placed high.

For remote operation there are two design points that
need to be given high priority. One is KISS, or Keep It
Simple Stupid. The more stuff you add, the more likely
it will break. Another point is that spending extra on
the highest quality equipment will pay off in the long
run. (I'm used to dealing with situations where it
might cost $2000 to get a technician on site; hence
spending a few extra thousands initially pays off very
well over a period of years.)

One piece of equipment that has not been mentioned,
which can be put to very good use, is a modem controlled
AC power distribution unit. Say 10 AC outlets, each of
which can be turned on and off remotely via the modem.
That allows you to "reset", individually, each piece of
equipment, and do it in the order you choose. It could
also be arranged to have one outlet trigger a mechanical
device to physically push a button on a camera too!

That device will correct about 99% of all "problems"!

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)

(Floyd L. Davidson) wrote:
The specifications for USB cable length are for powering
a remote device at the maximum load current allowed.
That is a relatively short distance. Of course if you
are not powering the device from the USB port, the
distance over which data can be sent is significantly
longer!

Actually the signal round-trip time from host to device and back is a
limiting factor, too. Using the maximum number of chained hubs this comes
down to about 5m maximum cable length.
If you are using fewer hubs, then yes, you can increase the cable length
further, but at the cost of degrading the signal quality. At some point it
just won't work any longer.

There are better options which transfer the USB signal on CAT5 cables (max
length 50m) or fibre optic (up to 10km).

jue


I wouldn't doubt that it would work at some significant
distances, but for a unmanned remote it probably should
be limited to a few meters. Hence, if the camera is on
a tower adjacent to a building, go for it. If it is in
a building 30-40 meters away... that's likely too far.

However, I will certainly consider having a UPS standing at some
distance from the camera/PC enclosure.

A sharp power-down won't though. If the power fails
while the computer is in middle of a disk access, it's
possible (not likely, but possible) to have file
corrupted or worse, an unstartable system... or
ultimate worst-case, damaged hardware. A power SURGE
or brownout is even more likely to do damage.

True, but when running NTFS, I think that the risk is rather small.

But, if possible I will try and install a UPS at ground level.
The cameras will mostly be placed high.

For remote operation there are two design points that
need to be given high priority. One is KISS, or Keep It
Simple Stupid. The more stuff you add, the more likely
it will break. Another point is that spending extra on
the highest quality equipment will pay off in the long
run. (I'm used to dealing with situations where it
might cost $2000 to get a technician on site; hence
spending a few extra thousands initially pays off very
well over a period of years.)

One piece of equipment that has not been mentioned,
which can be put to very good use, is a modem controlled
AC power distribution unit. Say 10 AC outlets, each of
which can be turned on and off remotely via the modem.
That allows you to "reset", individually, each piece of
equipment, and do it in the order you choose. It could
also be arranged to have one outlet trigger a mechanical
device to physically push a button on a camera too!

That device will correct about 99% of all "problems"!

Jürgen Exner wrote:
(Floyd L. Davidson) wrote:
The specifications for USB cable length are for powering
a remote device at the maximum load current allowed.
That is a relatively short distance. Of course if you
are not powering the device from the USB port, the
distance over which data can be sent is significantly
longer!

Actually the signal round-trip time from host to device and back is a
limiting factor, too. Using the maximum number of chained hubs this comes
down to about 5m maximum cable length.

Hubs are powered from the cable, and *that* is precisely
what I excluded in the discussion above.

Regardless, the latency due to length of a 5m of cable
isn't significant at the speeds USB 2.0 works at.

If you are using fewer hubs, then yes, you can increase the cable length
further, but at the cost of degrading the signal quality. At some point it
just won't work any longer.

The point is that there are *no* hubs. And it was never
represented as a long haul mechanism.

There are better options which transfer the USB signal on CAT5 cables (max
length 50m) or fibre optic (up to 10km).

Off hand I'd say that over 10 meters, maybe a little
more, that would definitely be the appropriate solution.
That of course is just about exactly what I said
previously, as you've quoted below.

Regardless, the point is that the restriction the OP
mentioned is for power, not signal, distance over a
normal USB connection.

I wouldn't doubt that it would work at some significant
distances, but for a unmanned remote it probably should
be limited to a few meters. Hence, if the camera is on
a tower adjacent to a building, go for it. If it is in
a building 30-40 meters away... that's likely too far.

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)