[arm-allstar] GPIO vs Nano

[Roselito de los Reyes]

Thank you for your beautiful work Barry


Lito KI9H


________________________________
From: arm-allstar <arm-allstar-bounces at hamvoip.org> on behalf of "Barry Buelow via arm-allstar" <arm-allstar at hamvoip.org>
Sent: Tuesday, December 6, 2016 1:56 PM
To: arm-allstar at hamvoip.org
Cc: Barry Buelow
Subject: Re: [arm-allstar] GPIO vs Nano

All,

I originally built a pcb that plugs into the GPIO of the RPi. It had
room for 2x FOBs (minus plastic cases) and buffers for PTT-COS. I have
experience with complex board design (avionics) and expected some
coupling from the RPi to the board, but it was significantly higher than
I expected. The audio always had "worms" and it was quite difficult to
knock down the spikes on the +5 and +3.3. The short answer is that the
IO board was too close and there was no simple way to get more
isolation. It was also clear that if there were some undesirable event
on the radio, the path to Gnd was via the RPi. An aside comment for now:
building a system with several floating ps is a really bad idea.

Over the past few months, I've shown Dave and Doug my work and received
comments. The NanoDB is not a bullet proof interface. Doug always argues
for cheap :-), but he wants the features of a more expensive unit.  The
NanoDB is rather minimalist in design. I had a 74HC14 buffer between the
inputs (COS, CTC) and the Nano, but took it out. The amount of
protection it offered was inadequate if there were a big event (as was
re-enforced at an IEEE EMI class a few months ago). The 74HC14, with
jumpers, also allowed the IO to invert signals to have LEDs indicate the
correct state regardless of COS=1 or COS=0 from the radio. I still like
that idea, but it can be accomplished in the Nano Sw.

What drives my designs is the desire to have no external ugly blight of
wires and components necessary in normal installs.  Why would anyone
want a design that requires external 2N2222 + bias resistors hanging in
midair off of a DB25?  My goal was to have the NanoDB work for 90% of
the simple nodes. It allows use of the potted FOB to build a cheap
system.  I'd be very interested to hear about cases where it did NOT
work as a future rev might address some issues.

I have also contemplated a more elaborate mechanical design which
accommodates IO for 2 radios, 2 FOBs, and integrates into a nice
extruded metal case with room for a pair of DB-9s on the rear. Think
bigger $.  Dave suggested an integrated USB HUB. Those are fine pitch
parts which exclude home bench soldering and small quantity builds are
more expensive.  Maybe someday.

I think the NanoDB will be a contribution to Allstar, which was my
original intent.

Barry w0iy
------------------------------


_______________________________________________

arm-allstar mailing list
arm-allstar at hamvoip.org
http://lists.hamvoip.org/cgi-bin/mailman/listinfo/arm-allstar
arm-allstar Info Page<http://lists.hamvoip.org/cgi-bin/mailman/listinfo/arm-allstar>
lists.hamvoip.org
This list is for support, announcements, and general information about ARM Allstar projects at hamvoip.org including but not limited to the BeagleBone Black and ...




Visit the BBB and RPi2/3 web page - http://hamvoip.org