[PLUG] powering relays with usb-serial adaptor

[Larry Brigman]

On 6/30/07, Eric Wilhelm <scratchcomputing at gmail.com> wrote:
> The meaning of "usb-serial" in this case is the regular meaning :-)
>
> I've been playing with a diode+relay setup which needs 5V for the relay,
> works great on the old dell laptop serial port (the little reed relays,
> plus an led and resistor (light-emitting optional.))
>
> I'm going to hookup my external usb backup drive to only be on when
> backups are running (duh! turn off the lights when you leave.  Why
> don't more devices have power control?.)
>
> I got some $9 usb-serial adaptors with the Prolific PL2303 chipset.  To
> my dismay they don't show any voltage between the ground pin and the
> signal pins.  WTF?

It could be it doesn't turn the signals on until it sees (senses) the
other end of the
RS-232 signal lines.

>
> I jumper the ground pin to the usb ground (outer shield only) and viola!
> 5V  (Seems like those would have been wired together to start with
> (btw, the ground pin5 is also not wired to the outer serial shield.  I
> thought that was supposed to be the case.  Further, the serial shield
> and usb shield are not connected, though the cable has the shielded
> jacket which I would assume is the usb shield.  Will learn more soon
> through dissection.))
>
> What's the deal?  If I hook it to a modem it dials (though only at
> 9600baud for some reason that I haven't thoroughly investigated.)  Why
> does the modem see a signal when the voltage meter doesn't?  Why are
> the usb shield and the ground pin at different levels?
>
> If I'm only using it as a relay controller, is there anything wrong with
> soldering pin5 to the usb ground?
It could be totally isolated (opto-coupler) such that it doesn't create ground
loops.  Connecting the shield ground to the 232 ground could break the
isolation.