RE: Electronics question

["Paul Gordon" <paul_gordon@xxxxxxxxxxx>]

Hi Ant,

I suspect the same cause, - insufficient power available from the DTR pin
to
drive the circuit... - for instance, there's only 0.99Vdc between the VDD
&
GND pins for the DS1820 (as measured at the screw terminal on the 3145
board). The quasar page suggests there should be at least 4.4V between
these
lines *at the remote end* - so I think that seems fairly conclusive...

I'd thought about hacking a PSU in instead, rather than trying to take
power
from the DTR line. - I'm concerned about the isolation between any such PSU
& the Avocent ports though, - I'd hate to damage a port (or worse the
whole
unit) by feeding back in a voltage where it didn't like it!

I'm by no means any kind of electronics expert, (not even close!) so I'd
appreciate some sage advice about what (or what NOT) to attempt in any such
experimentation... for instance, the circuit clearly only requires 3 pins
from the serial interface: DTR, GND, RXD... I assume an external PSU would
need to be connected across DTR & GND... - but should I leave the
existing
DTR & GND connections to the Avocent in place or remove them? - I would
have
thought remove them, otherwise might there be a risk of either feeding back
a voltage into the Avocent (which I guess would happen if I connected the
PSU to the "wrong" side of diode 1N4003?) or might I cause
grounding
problems by having two isolated GND's connected to the circuit? But I'm not
sure if (or how) serial data will flow if *only* the RXD pin is connected
to
the Avocent? - wouldn't this require a reference to the RS232 port ground
from the Avocent to determine high & low states?

So would the answer be to place an external PSU between the DTR & GND
lines
of the circuit, disconnect the DTR line between the 3145 & the Avocent
(easy
enough by removing the pin form the modular DB9-RJ45 adapter), but to leave
the GND line connected to provide a reference?

Or am I just talking nonsense? (wouldn't be the first time!)

TIA

Paul G.


-----Original Message-----
From: ukha_d@xxxxxxx [mailto:ukha_d@xxxxxxx] On Behalf Of
ant
Sent: 25 June 2009 11:09
To: ukha_d@xxxxxxx
Subject: Re: [ukha_d] Electronics question

Check the output of the LM7805 regulator, are you getting a stable 5v? My
guess is your new thing can't drive enough current to power the board. I'd
hack in a wall wart and save yourself the pain of scavenged power. Or make
a
new USB one ;)



2009/6/25 Paul Gordon <paul_gordon@xxxxxxx>

> One for the electronics guru's here perhaps?
>
> I run a couple of the Quasar temp monitoring boards into my HA server.
I've
> just been trying to move them over to the new Avocent serial-over-IP
box
> that has replaced my old 8-port PCI RS232 card. - these are among the
last
> couple of devices to reconnect, I've already successfully configured 2
> Comfort UCM's and a CurrentCost...
>
> I'm not getting any data from the 3145's though, and I'm wondering if
its
> related to their use of power from the DTR line of the connected
serial
> port? Here's the circuit diagram:
> http://www.quasarelectronics.com/images/electronic-kit/3145schematic.gif
>
> I've put a meter across the DTR & GND pins on the RJ45 outlet
jacks on the
> Avocent box, and I'm seeing 3.5VDC. I've also put the meter across a
couple
> of points in the 3145 board itself, and I'm seeing voltages at various
> points, - for instance, there is 0.99VDC at the DS1820 signal &
GND
> connection points, so there appears to be power to the circuit, - but
my
> question is, is it enough? - I've not yet tried to test the output
from
the
> PC serial port it was previously attached to, - it's not easy to do
without
> butchering a serial cable, so I can't compare them. However, the
Quasar
> page
> for these circuits does mention the possibility of data errors due to
> unstable power from the DTR pin...
>
> I'm emailing the question directly to Quasar as well, but while I'm
waiting
> for their reply, I'd appreciate any thoughts anyone else might have...
>
> TIA
>
> Paul G...
>
>
>
>
>
> ------------------------------------
>
> Yahoo! Groups Links
>
>
>
>


[Non-text portions of this message have been removed]



------------------------------------