info at meshnetworks.co.nz
Tue Jul 5 20:07:49 EDT 2016
On 06/07/16 11:44, Kevin Custer wrote:
> Soldering a 48 pin LQFP and having it be fake is nothing short of a
> bad experience. Like Scott, I recently purchased a fair quantity of
> CM119A's from a top CM distributor. They told us the CM119 and CM119A
> are now legacy products and the CM119B is current. Unfortunately, the
> B model doesn't use a crystal and several developers agree that they
> may not be suitable for digital and analog radio adapter
> applications. There were still a few hundred CM119A's in the
> distribution chain as of a few months ago.
If the CM119 is dead, then we should scrap it. I can't really see why
we are focussing on that devices' built-in I/O, when we have computing
devices such as RPI3 with scads of easily accessible I/O.
Linux sound is really well developed - "just works everywhere" I'd say.
Why not build or reuse a DSP-based driver for that, that will handle
voting - now we are in business with all the toys, for the foreseeable
future. Likely a simple CLI config tool could be developed too - choose
RX audio device, TX audio device, signalling, save settings, done.
Rinse and repeat for three devices.
Hardware VOTERs are fine for a single repeater with a few outboard
receivers, but way too much money for non-millionaire clubs to convert
"everything" to asterisk - they will stick with their IRLP node for
sure, and that is a tragedy.. I see a few members here and elsewhere
make it their business to put tens of thousands of $$ into communication
systems, and that's enviable and laudable, but 90% of constructors don't
exist on that rarefied plane.
More information about the App_rpt-users