开云体育

On Feb 12, 2023, at 15:19, WA8LMF via groups.io <wa8lmf@...> wrote:

?On 2/11/2023 5:20 PM, wa5luy wrote:

Stephen,

I am betting on Vara. The previous test you made with Vara was a real accomplishment.
The first time I ran Vara on Winlink I was amazed that it could decode stations that I could not physically hear.


73 and I will be looking forward to your test results.

Wayne
WA5LUY

The test run was a complete fiasco with numerous hardware and software problems.

For example, I discovered the FLdigi program's audio detector on receive is so sensitive that transmitting on one mode would cause the other instance of FLdigi (on receive) to hear the RSID code and auto-switch to to the same mode! So I would beacon on MFSK16 -- then the next beacon (from the other copy of FLdigi) would also beacon on MFKS16 instead of OFDM.

I think some sort of sneak audio path in the hardware of the computer or interface, or perhaps possibly stray detection of transmitted RF.

I did days of testing in my carport before heading out on the trip. For the at-home tests, I had a 10dB 50-watt attenuator in the line between the transceiver and the mobile antenna. Partly to keep the nearby base station receiver from overloading so badly, and partly to hide the severe de-tuning of the mobile whip when parked under the carport canopy, from the transceiver. (In the interest of keeping the transceiver PA as linear as possible.) Under these conditions there were no problems. On the road, without out the attenuator with the antenna matched and radiating full power, the problems began.

Also, my experiment to have VSPE (Virtual Serial Ports Emulator) allow three applications to share the same serial port for PTT keying worked fine on the desktop pre-test, but came unglued in the real world. It would work for a while, then randomly lockup and stop doing PTT. I suspect this had something to do with the DigiRig interface I used having NO isolation between the computer side and the radio side; i.e. common ground connections, aggravated by automotive electrical noise circulation on the vehicle ground systems. (Again, the home desktop pre-test worked perfectly for days when the the transceiver (Yaesu FT-891) and computer were being powered by clean DC power from a 12 VDC 100AH battery bank being float-charged by a good 20A DC power supply.

At the same times,. a second VSPE serial splitter setup sharing a single GPS receiver with 5 destinations (MS MapPoint for navigation, the main two-meters APRS instance of UIview, and three copies of UIview being used for the experiment on 60 meters) would also quit. The normal setup (MapPoint, two UIviews, a GPS monitoring tool "GPS View" and sometimes Delorme Topo NA) has been absolutely bullet-proof with this setup in the past for over 100,000 miles of mobile operating.

I lost a lot of tine stopping at highway rest areas rebooting the Toughbook and restarting applications in a precise sequence. On the return trip from the hamfest, I began to suspect the interface and noise/ground loops as the problem.

I stopped by the side of route Michigan 115 and switched the interface to my own homebrew totally passive tone-activated self-powered and TOTALLY ISOLATED interface (that I had been demoing at the hamfest). I lost an hour re-configuring and tweaking multiple applications on the Toughbook; then resumed the drive. The GPS splitting and PTT was now bulletproof!

But I hadn't noticed that in the meantime the audio tone freqs in both copies of FLdigi had shifted downward 100 Hz, ruining the remainder of the test!

I plan to run a proof-of-concept test on a short drive from East Lansing to Grand Rapids (about 70 mi away) sometime next week. Before running a "real" test farther afield enroute to a ham fest in a month or two.


All this comes with the territory of mobile radio systems engineering which I have been doing for decades!