From: gccradioscience on

The computer has an issue with a 5VSB LED blinking on and off. I was
trying to build a USB power supply adapter to connect to the USB
connector. The problem occurred when I applied a 6 volt 500 mA load to
the USB port, but the (DX-398) 6 volt 300 mA worked, but did not provide
enough power. The 6 volt 500 mA load was a shortwave radio portable
receiver, I did not know this would happen to the computer USB port, I
hope I did not fry out the motherboard and it's just the power supply
that needs to be replaced. Now I need to find a way to fix it cause
there is old files on the HDD I need to retrieve. Can someone help me
out with this? :confused:

http://www.youtube.com/watch?v=JvX_CVCpbXY


From: davy on

Anything could have happened...!

The first step try unplugging the main power at the wall sockets for
about three minutes and try a reboot.... I doubt that this would work
though.

The USB ports only supply 5 Volts, 500mA does seem to be rather a lot
for a single USB port... the best bet would be to check there IS 5 volts
dc there at one, well two counting the negative connection 'see this
article,' (http://www.girr.org/mac_stuff/usb_stuff.html) the article
should be self explanatory, obvious you don't need to go to this length,
a simple test meter will do the trick... or a USB device that lights up,
the USB lead of an old or defunct device, will make the connections
easy if you strip the sleeving off the 'free' end.

Are you sure you didn't short any other pins out accidentally? We need
to know if the 5V is still present for starters. Another idea is to tell
us the mother board.

davy


From: Paul on
gccradioscience wrote:
> The computer has an issue with a 5VSB LED blinking on and off. I was
> trying to build a USB power supply adapter to connect to the USB
> connector. The problem occurred when I applied a 6 volt 500 mA load to
> the USB port, but the (DX-398) 6 volt 300 mA worked, but did not provide
> enough power. The 6 volt 500 mA load was a shortwave radio portable
> receiver, I did not know this would happen to the computer USB port, I
> hope I did not fry out the motherboard and it's just the power supply
> that needs to be replaced. Now I need to find a way to fix it cause
> there is old files on the HDD I need to retrieve. Can someone help me
> out with this? :confused:
>
> http://www.youtube.com/watch?v=JvX_CVCpbXY
>
>

I would suspect the power supply has a problem. Normally, even if the
motherboard was a bit messed up, the 5VSB LED should stay on steady and
not blink. If you see it blinking, that probably means the supply now
has a problem. (At the very least, the supply currently thinks 5VSB
is being overloaded.)

There is a schematic here, of an ATX supply. This is a rather old design,
but the principles illustrated in here, still apply. You'll notice
the 5VSB comes from a linear regulator. If the linear regulator is forced
to provide too much current (above its rating), it gets hot. A 78L05 probably
turns off the output, when the casing gets hot enough to burn you. That could
cause a slow oscillation, as the thermal overload occurs and then cools off.

http://www.pavouk.org/hw/en_atxps.html

*******

If I needed to power a 6V load like that, I would look for a digital camera
wall adapter. I have one I got from RadioShack years ago, which has a
switchable output (it has settings up to 7V). RadioShack doesn't sell it
(273-8400) any more. But you may still be able to find a fixed voltage unit.
Example here:

http://www.pqbattery.com/battery.cfm/Digital-Camera-Batteries/Olympus/Camedia-C-2100-Ultra-Zoom/AC-Adapter-fits-Olympus-Nikon-6V-65V-2100mA-DC-Output-132835-3637

You want one which is definitely a regulated output, and not just a
"transformer and diodes" type unit. Switching regulator type wall
adapters, are compact and lightweight, while the wrong kind are
larger and heavier (as there is a large transformer inside).

The reason I recommend a separate adapter, is the sensitivity of the shortwave
will be degraded if it is powered from a computer. The computer has a lot
of electrical noise inside it, which will "notch" reception at certain
frequencies you're trying to listen to. Using a separate regulated
supply may give a cleaner power source. Using rechargeable batteries
to run the shortwave radio, is another option. If you keep a couple
sets of batteries, you can be charging one set, while using another
set.

When a hobbyist needs a regulated voltage, they start with a DC source
of somewhat higher voltage, then pass that voltage through a linear
regulator. These devices are available in all sorts of standard voltages.
For example, the LM7806 will produce a regulated 6V output at 1 amp.
You need to bolt a heatsink to the tab of the regulator, if powering
loads approaching 1 amp. Your computer has +12V (yellow) and GND (black)
on a Molex four pin disk drive connector, which would serve as a power source.
So hobby projects can be powered by the computer. If doing this,
you would put a fuse in series, coming from the computer, to protect
against accidents. Since the 12V can provide a lot of amps under
fault conditions, the fuse it there to protect the computer and
protect you. But in the case of your need to power a shortwave radio,
this would conduct too much wideband noise, to be good for shortwave
reception. Which is why I would not recommend this approach in this
case. (You could place a ferrite choke on the feed wires, but it
might not be enough.) But in future, if you need various values of
regulated voltages, this is how hobbyists used to do it. I have one
of these near me right now - it produced 12V @ 1A to power computer
fans for testing. It is powered by an unregulated wall transformer.
So I use these regulators for hobby projects. RadioShack doesn't
make enough different models of wall adapters, to power all my
hobby project needs. Sometimes, I feel the need to "roll my own".

http://www.fairchildsemi.com/ds/LM/LM7805.pdf

Paul
From: kony on
On Thu, 22 Apr 2010 17:47:35 -0400, Paul <nospam(a)needed.com>
wrote:


>There is a schematic here, of an ATX supply. This is a rather old design,
>but the principles illustrated in here, still apply. You'll notice
>the 5VSB comes from a linear regulator. If the linear regulator is forced
>to provide too much current (above its rating), it gets hot. A 78L05 probably
>turns off the output, when the casing gets hot enough to burn you. That could
>cause a slow oscillation, as the thermal overload occurs and then cools off.
>
>http://www.pavouk.org/hw/en_atxps.html
>

Not so sure the principle still applies since even the low
end PSU today aren't usually using a linear IC which would
overheat to the point of an integral thermal shutdown.
While a PSU may have thermal shutdown in general and 5VSB
subcircuit can be damaged over the long term from being only
passively cooled, it would more likely be the filter
capacitor in it going out or damage to the motherboard
itself causing the problem.