Originally Posted by Gypsy23
It turns out the "NMEA" output from the GME is not RS422, not even RS232
. It gets loaded down by the E120W NMEA input impedance (abt 600ohms) such that the signal is only -0.25V to +2.5V (viewed on a scope). All other "NMEA" signal sources into the E120W are at least -4V to +5V when plugged in to the E120W.
When I send captured data $CVDSC/DSE msgs back to the E120W from my PC with a USB-Serial cable which provides signals of -5 to +5, it responds correctly, the same as it did to $CDDSC/DSE. Unfortunately, it still doesn't work with OCPN.
I will have to take this further with GME or build a small interface circuit which presents a higher impedance to the GME radio
Since this and other posts I had a reply from GME Australia
, the manufacturer of my VHF
GX600D. They agreed that the talker ID $CV could be changed and also offered to reduce the output Z of the NMEA cct to better drive the typical 600ohm differential NMEA inputs found on Raymarine
and other NMEA devices. The GME NMEA o/p will still be 0 to +5V single
ended but not get ‘killed’ by driving a low impedance. Top marks to GME for their response to this issue.
I took my radio in for updating last week and the mods were done and worked perfectly. It seems that when they developed the radio the $CV talker ID seemed appropriate and it worked OK with the Garmin
plotter they had as a test bed
. The Garmin
ended NMEA i/o hi/med impedance ccts too so the GX600D was able to drive them correctly. However, GME now see $CD is a more widely accepted talked ID for DSC and will use it for future production.
This brings us back to OCPN. As noted in my earlier posts, I have logged a request in Flyspray for OCPN to support a wider range of DSC talker IDs. I still think this is a good idea even though my particular problem has been resolved by GME. It is clear that Raymarine and Garmin accept at least $CD and $CV, but I expect they would also accept $Cx as has been proposed in the change request. It is also clear that this whole DSC subject has been implemented differently by many VHF and/or Plotter designers so OCPN accepting a wider range of talker IDs will likely resolve other problems in the future as more users hook their systems up to take benefit of these features.