My Jan 2025 VHF Contest Outing—Another lesson “How NOT to do it…”
Background
My previous September 2024 VHF rover contest outing was excellent.
At that time, I successfully traveled up the Texas midriff while operating continuously on 6m and 2m FT8 while in motion. At 14 different stops within different grids, I activated SSB, CW, FT8 across 50/144/432/222 MHz bands in a frenetic but effective manner bouncing across bands making QSO’s wherever I could find a signal. I made an encouraging preliminary score. However, the score is not what matters, it’s the execution of incremental progress — and that’s intoxicating. But progress is only earned at the expense of a few dopes slap backwards before moving forward. … and that’s what happened last weekend.
For this outing, I got greedy. I wanted to take one giant step forward with a huge number of new additions and new capabilities:
– NEW power amps: a 300W/2m TE-1442g and a 300W/6m TE-5052g
– NEW 1296 MHz band capability:
– NEW 3 antennas
– a 1296 MHz 25 el loop Yagi
– a pair of 144 MHz Stacked PAR OA-144 Omni loops
– a 50 MHz PAR OA-50 Omni loop
– NEW 10′ mast pole on the newly installed trailer hitch
– NEW audio switchbox to share one headset and PTT across 3 radios
– NEW super lightweight W2ENY headset and lightweight PTT switch
– NEW Icom IC-7300 dedicated to 6m
– NEW 222 MHz XVTR configuration using a dedicated TS590s
– NEW WSJT-X FT8 capability added to the 222 MHz SSB mode
– NEW antenna cable routing path into car for more cable slack
– NEW paralleled 100A LiFePO4 batteries for the 6m amp
– NEW travel route across rare WTexas Grid sites (incl 2 Oklahoma sites)
– NEW 2 SWR meters to monitor 2m/6m output
– NEW 2 coax switches to flip between station and mobile antennas
– NEW external T/R switchbox for the IC9700
– NEW rack shelving to accommodate the extra radio and amps.
– NEW heatsink blower fans for each amp
– NEW version of WSJT-X SW (V 2.7.0-rc8 Beta) with message center feature
– NEW resized rooftop tower cut shorter to hold more antennas
The executive summary:
This was my route
Here’s a diagram of my setup
This was a headache learning adventure. All the new “improvements” created extra data and coax cabling paths which allowed EMI to creep into the radios and PC. I suspect this likely caused various corruption of the USB serial ports and audio CODEC drivers The USB connection integrity savagely torn at the attempts to maintain a stable FT8 operation. Little by little, my efforts to recover and diagnose the problem introduced even more issues by likely mismanaging the radio setup configurations in desperation. By Sunday, all 3 radios were almost inoperable using WSJTX. Note: 90% of contacts in this type of contest are on FT8. Afterall, WSJTX/FT8 is a weak signal mode, and that’s it’s so superior to SSB or CW- and while roving in the distant locations on VHF/UHF you need that extra advantage.
Here’s a short version of the breakdown:
DAY 1
- 2m power amplifier refused to switch on
- 6m power amplifier RFI wiped out WSJTX and N1MM USB connections between PC and radios
- Decision: Bypass the amps and operate in low power mode
- IC-7300 6m WSJT-x randomly began losing its USB connections (UART and Audio). Reboot/Recovery was difficult as USB COM/Audio channels reshuffled after reboots and during debug efforts. Orphaned WSJTX and N1MM processes often required the whole system reboot before recover could be attempted. The dedicated 6m station permanently refused to transmit a 6m FT8 audio carrier by mid-day.
- Decision: Continue the contest without 6m FT8 capability
DAY 2
- IC-9700 WSJT-x refused to transmit a 2m FT8 audio carrier but worked fine when using 70cm. Oddly, 2m worked fine at close of day 1, but didn’t work when initially powered on in the morning on Day 2.
- Decision: Abort operations and head home. Accumulated troubleshooting already exceeded 3 hours.
Dallas
- Noticed in the rear-view mirror: hitch mounted 10′ mast swaying more than usual. My imagination? Not sure.
- Decision: Pull into gas station to check. Discovered one of two roof mounted holding struts had detached from the mast: a high risk of losing the mast and 3 mounted antennas. Reattached strut and returned on road
La Porte
- Glancing at shadow of antenna stack beside car, I see lots of swaying. My paranoia? Normal?
- Decision: Pulled off in La Porte side street to find all 4 rooftop suction cup (Seasucker) guy wire anchors had slipped, loosening all the guywires holding down the mini tower stack. I restored the guying tension by repositioning the suction cups back to their original positions. Onward home.
Home
- Upon unmounting the antenna rack, I observed one of the U-bolts that secures the antenna tower to the roof rack had almost come completely undone. The nuts had backed off to the last bolt thread. In this state, the loose ubolt was likely contributing to the whole stack to shimmy and sway. Each road bump and wind shear forces at up to 90 MPH had clearly aggravated the loosening. The loosened guy wires (above) arguably magnified the vibrating energy … or maybe they saved the ubolt from detaching completely? Disaster avoided.
Operation Comments
It’s wildly unfun when struggling to troubleshoot ghosts, during a contest. My aggressive route plan had very strict time limits at key grid locations, so the unplanned time to debug was quite impactful. Every time, I lost a USB connection, I would have to either reboot or physically reconnect broken com ports and audio CODECs on a radio. This was a nightmare, since the USB port order often changed after a disruption and recovery attempt. That’s an artifact of Windows- they no way to hard code a Com or Audio line to a specific device. This took a huge amount of wasted time just identifying the correct ports to get the CAT/PTT/audio flowing correctly. To identify the correct COM port and associated Audio CODEC for a radio, one must physically disconnect the USB cable while observing the disappearance of the device under Device Manager.
It wasn’t (and still isn’t) clear if RFI was the single culprit. It seemed the SW was unstable and doing it by itself. I considered reinstalling N1MM and WSJT, but didn’t have the time nor patience to reconfigure all the switches. RFI is bad, and stronger RFI is worse. You’re also at the mercy of how well your device is resistant to EMI.
I started questioning if I had all the correct USB Data baud rate, parity, word lengths setup on each machine. I usually could regain CAT control but struggled trying to figure out why WSJT audio wasn’t transmitting?! I erroneously played with USB and CIV baud rates, word bits, and many configurations of Windows audio device settings with no joy. I reduced power and kept getting no meaningful results. When the 2m radio broke down with the same issue on Sunday, I threw in the hat. Just prior to shut down on Saturday, 2m WSJT worked fine. Early Sunday morning 5:30AM, when powered back up, the missing audio happened. What changed over the last 5 hours?!!
On the optimistic side, I made a few QSOs to validate that this new mid-West Texas route could propagate into San Anton and even into Louisiana on 2m and 70cm.
In hindsight, all of this was with testing and adjustments ahead of time. I simply ran out of time setting up all the new modifications/upgrades. I spent too much time before the contest repairing broken equipment (the 6m amp), recabling new devices, and implementing changes that required sawing, screwing, and assembling (rooftop tower, equipment shelving).
An almost disastrous loss of the rooftop tower was largely due to trimming down the tower leg sizes by 7″. This shortening of the legs was intended to accommodate a new 1296 MHz antenna along the mast stack. However, this shortening of the legs resulted in an uneven warp of its seating on the roof rack adding to instability during driving. AND it decreased the angle of attack between the guy wire and car roof, weakening the orthogonal (vertical) force needed to impart an adequate tensile force component on the guy wire. This explains why the Sea Sucker suction cup holders were found slipped and guy wires loosened when I stopped to inspect in La Porte.
Postscript:
With the newly added complexity and lack of testing, an RF rich environment radiated onto many new cables and coax, and wreaked havoc to the sensitive system. Ferrite beads were used, but obviously not as effectively as needed. In the past, I saw a similar WSJTx failure signatures (Com port and/or Audio device “not found”) only to be corrected with Ferrite Beads.
I had to constantly reset the PC multiple times to reconnect the drivers which often resulted in mysterious reordering of COM ports and MIC/Output CODEC drivers which caused lots of wasted time trying to recover by guessing which reordered drivers were to be reassigned to which radio. Each radio has an internal USB hub and associated USB UART Bridge and USB Audio CODECs, subject being susceptible to corruption. Likewise, the little mini-PC had processing complexity subject to EMI susceptibility, too.
Practically no 6m WSJTX operations were conducted since that subsystem was affected first and most prominently. Soon, 222 MHz WSJTx stopped working and finally the 2m WSJTX app also stopped responding (except when I returned home, and it worked again. At one point the 432 MHz which was mostly unaffected, also started to misbehave.
I spent several hours trying to troubleshoot the issue vs operating in the contest while continuing to travel up the state. Resetting the radios, changing application SW configurations, replacing ferrite beads on different cables, adjusting RF power levels, swapping antennas from Yagi to Omnis, removing and adjusting placement of USB and cable nests all didn’t make much difference.
A likely other element to try but didn’t have the ability, was to reinstall the N1MM and WSJT SW. Note: I was using a Beta version of WSJTx, but it may (or not) been a factor in the issues. There were other nuisance things that didn’t work out very well:
- The GPS auto grid didn’t update, so I had to manually change grid square into N1MM and the WSJTX configurations each time I entered a new grid.
- An echo of feedback was observed in the headset when the IC9700 was keyed in SSB
- The 222 MHz transverter barked loudly when transmitting on 144 Mhz.
- The newly reconfigured shelves held additional equipment, but the monitor screen (using the cupholder mount), obstructed a clear view and access to the IC-9700.
The Future (Corrective Actions)
I’m ready for the next contest in June. I’ve decided to simplify the setup to only use 2 radios vs 3.
- I will isolate USB radio cables to a direct connection to the PC vs a hub.
- I will clean up the cable runs.
- I will add more RF ferrite beads.
- I will replace my generic Beeline mini-PC with a more well shielded and rugged Intel NUC gen 8.
- I will remove the audio sharing switchbox to preclude excessive cabling and ground loops.
- I will strap all equipment to a common ground strap anchored to the car chassis.
- I’ll add a bandpass filter to the 2m radio.
- Test, Test, Test
Incredible journey and great details on all the roadblocks you were faced with! Tremendous effort and commitment! Thank you for sharing this experience and here’s looking forward to the next chapter! Wonderful presentation and great photos! There are few of us who could have accomplished this much!
73
KG5URA
Great write-up as always! You’re doing great Dave and I’m sure you’ll redesign and work out the bugs. Can’t wait to see the setup for the June contest.
Amazing post – I appreciate your patience and willingness to learn and always keep
Smiling when things do t go completely rihhy