[aprssig] Balloon - lessons learned (AVMAP-GPS) - GreenLight Labs GPS?

[Stephen H. Smith]

On 4/7/2014 12:24 PM, Robert Bruninga wrote:
> For the chase, we added an AVMAP GPS for the map display and plugged it into
> the “pass-through” socket on the Green light labs GPS.  Everything seemed to be
> working.
>
> 20/20 HINDSIGHT HINTS:  Occasionally we notice the D710 was not updating and so
> we cycled power and continued.  Some times the AVMAP was not updating.
> Sometimes when the AVMAP was plugged into the Green Light Labs GPS, then the
> D710 did not get GPS updates. But generally, a power cycle fixed it.  Did it
> maybe 3 times during the all day chase.
>

What kind of a kludge was this????       What does the "pass-through" port on 
the Greenlight actually do?  Is is routed through some sort of uController that 
can buffer and arbitrate data flow through the main port? (The way GPS 
passthrough is handled on the Kenwood radios?)

Or is it just a second jack physically wired in parallel with the main output, 
intended to strictly provide GPS data to a second receiver-ONLY such as a 
laptop running a mapping program, hooked up with a one-way two-wire connection?

The AVmap talks as well as listens, since it acts as THE GPS source, as well as 
a data receiver for mapping, when connected to a Kenwood radio's GPS port.

With TWO RS-232 devices,you have a simple cross-over "null modem" with TXD on 
each going to the RXD side on the other.   But how do you connect *THREE* 
RS-232 devices, each with an active output to each other simultaneously?

Presumably the Greenlight's TXD (active TTL or CMOS output) was going to the 
D710's RXD pin.  But so was the AVmap's TXD (GPS data out).  So now you have 
the AVmap -AND- the Greenlight trying randomly to talk to the Kenwood at the 
same time.

With THREE RS-232 talkers and TWO RS-232 listeners sharing a single set of 
wires, with no polling or time-slotting, it's a virtual certainty that RS-232 
transmissions from more than one device at a time would overlap, and further, 
create "tri-level" RS-232 (high, low and somewhere in-between) when a bit was 
HI on one sending device and LO on another.

In turn, it's no surprise that one or more of the receiving CPUs involved would 
choke and lockup when this kind of unparse-able trash was fed into them!