|Method: HP Z3805A 10MHz GPSDO, Marconi 2019 HF synth, HP 3325A LF synth, HP 54601A o'scope, FT-847 in AM/narrow mode, 5BTV antenna, Win10 laptop running SpecLab, DigiPan for spotting, EZGPIB controlling Prologix USB-GPIB and USB-Serial to control and interrogate the instruments and radio, GPSCon for monitoring Z3805A; the 2019 was set to inject a reference beat signal to produce a ~400Hz beat tone, and the 3325A was manually tuned for minimum phase slip between it and the beat tone. Each run was also recorded to a .wav file through SpecLab, and at a later time replayed to export data to spreadsheet to refine submission numbers.|
Soapbox: Thanks Connie for runnung the event. Generally weak signals with the usual multipath spreading, plus 80M line noise supplied by PG&E for no extra charge. 80M late run wanted to be ~0.30 Hz higher, so had to guess a number. 73, L
|Method: Method: 98' long W5GI dipole oriented E-W connected to Icom IC-7000 (CW mode). SignaLink interface to laptop running Spectrum Lab. During call up, adjusted a GPSDO (Trimble Thunderbolt) referenced, HP 3335A frequency synthesizer to about the same frequency. After key-down, switched the radio input to the synthesizer (through an adjustable attenuator) and adjusted the synthesizer to the same frequency (used SL peak detect function). Read each frequency directly off the synthesizer. No radio or computer sound card calibration is needed with this method, but warmed both up for several hours so they were stable during the FMT.|
Soapbox: Both 40m & 80m signals strong (both runs) in FN30. Less doppler visible on the spectrum lab display for both signals at 0500 UTC, BUT the HP 3335A synthesizer was not working properly for the later run. Numbers are from the 0200 run only. Thanks to K5CM for his FMT efforts!
|Method: Using FlexRadio 6600 with WB0OEW GPS-disciplined signal generator, built from Sept/Oct 2018 QEX, with SpectrumLab software; combined with zero-beat using experience as a piano tuner.|
Soapbox: My first FMT attempt. I found it most useful to find frequencies while K5CM was sending CW, as I can more easily center the signal generator on the FMT frequency between dits, then later making final tweak by observing phasing between FMT signal and signal generator. We'll see if this method did any good!
|Method: SDR PlayRSP2 / Motorola 1200 Service Monitor / Various Internet SDR's|
Soapbox: Calibrated SDR Play with 10mz WWV and synced with service monitor.
One week warmup for all equipment. Checked various on-line SDR's for
propagation. Poor 40 M reception at this QTH, both runs. MUCH FUN !!
|Method: Icom 706 MKII and lots of audio interpolation.|
Soapbox: 1st timer here. Hope to be better prepared next time.
|Method: OpenHPSDR Mercury receiver,10MHz OCXO reference calibrated against CHU & WWV. WSJT-X Freq mode for measurements.|
Soapbox: Good signals on both bands in the early test. Didn't stay up for the late one. Thanks Connie!
|Method: Ensemble average of 65,536 point FFT in overlap-save mode after equipment calibration|
Soapbox: This was fun & much easier that previous attempts.
|Method: Used a Elecraft K3s and zero-beat the signal. Later I adjusted the
measured frequency to correct for errors in my tuner.|
Soapbox: This is my first try with this type of adjustment. The fractional part of the Hz is based on length of time between beats. I guessed at this
|Method: Equipment is a Yaesu FT-1000MP Mark V, 60-Hz CW bandwidth, Rockwell/Collins 180S-1 tuner, 40-meter dipole, Fluke 8845A multimeter. Calibrate receiver on WWV 2.5 and 5.0 MHz. Apply analysis to remove obvious fades, use statistical process control to discard outliers, correct for receiver drift.|
Soapbox: Good signals on WWV 2.5 and 5.0 MHz plus 40 and 80 meters. WWV 5.0 was remarkably consistent. The highest frequency variation was on 80 meters. Thanks to Connie K5CM for his effort and continued support of this interesting event.
|Method: Heterodyne with known reference|
Soapbox: Thanks to K5CM for running the fmt.
|Method: K3 with Buddipole antenna about 35 feet high. MacBook Pro with Fldigi software. Took WWV readings immediately before and after each run and averaged them to get receiver and doppler corrections.|
Soapbox: Somewhat poor propagation at my QTH. Heard the 40 meter signal at about S6, but never heard the callup for the 80 meter signal. While tuning around the band, stumbled on a very weak carrier (about S0.1) and measured that. The signal faded out at about the time K5CM would have stopped transmitting the test, so I'll submit it and see what happens. Even WWV was weak.
|Method: Flex-6300, FLdigi, DG8SAQ VNWA.|
|Method: Homebrew Receiver, modified sound adapter that accepts external clock and Rubidium oscillator. Local oscillator for receiver is PTS160.|
Soapbox: This is Bill WA4LAV, I have been doing the FMT at our club station for sometime, except for the past year or two. Had to find all my pieces and put them back together and had problems with windows 10 while configuring sound adapter. So I am a bit concerned about the final results.
K5CM was quit strong and stable during the both runs. No QRM and a bit of QSB.
|Method: ISOTEMP RESEaRCH INC. OCXO 134-12 Oven (Hamfest Special) $10 Reference connected to calibrated 1 week earlier with Rubidium Standard. Keysight Field Fox N9912A in Marker counter mode .1 Hz resolution. 1 hr. warm up before the measurement. OXCO reference on for 1 week. 80Meter Horiz.loop antenna up 30 ft., open wire feed to Palstar AT1500BAL tuner.|
Soapbox: No averaging, just made several measurements and picked the one with the largest signal amplitude that repeated itself several times. Next time I will do aging estimations after characterizing the oven reference, drift per day, and do some averaging on the measurement to improve resolution and average out the noise contribution, should be able to get close to .005 ppm next time.
|Method: Icom IC-9100 Receiver, 20-6 m Cushcraft MA6V vertical antenna, WSJT-X in Frequency Calibrate Mode.|
Soapbox: Good signals here for both late and early runs despite my off frequency vertical.
|Method: I put T connector in RG 8 going to receiver. Then I injected local signal from HP3336B. The HP3336B used 10Mhz reference signal coming from Thunderbolt GPS. Used homebrew software with waterfall to display both off-the-air signal and injected signal. On 40 meters antenna is Inverted V with apex about 40 feet. On 80 meters antenna is sloping wire hanging off side of tower with feed point up about 38 feet.|
Soapbox: 40 meters [0200UTC] --> Signal was about S4 and noise level also about S4. This is crude reading by just looking at S meter of ICOM 7600 receiver. 40 meters [0500UTC} --> Signal about 1 S unit stronger with noise level same as early FMT run. 80 meters [both 0200UTC and 0500UTC] --> signals S9 or better. Lots of dispersion on individual 40m frequency measurements -- each measurement lasting about 8 seconds. Signal was "smeared" across about a dozen FMT bins where bin size is about 0.12 Hz. On 80m individual 8 second readings were fairly tightly clustered. One both 40m and 80m I tossed out a few 8 second readings that were outliers and then averaged the rest to provide reported frequencies.
|Method: Receiver characterized to WWV, CW mode audio output to audio hybrid combiner with digital audio as second input, zero beat by ear.|
Soapbox: Noisy both bands, both early and late transmissions. Signals not strong but QSB tolerable. Thanks to Connie, et al., for another fine event.
|Method: My method added a digital audio recorder to the usual signal generator-bfo and audio frequency measurement approach. Having the recording gave me more than one shot at measuring the frequency of the beat note, which I cleaned up with a homemade 20 Hz audio filter. Lastly, using a scope, I subjectively matched its pitch and phase to that of the function generator. The FMT frequencies reported were of course the sums of the Signal Generator (rubidium-stabilized HP 8647A) frequency and the Function Generator (HP 33120A) frequency.|
Soapbox: Interesting exercise. Wish doppler didn't relegate milli-Hz measurements to a guessing a game. 80M measurement had to be done by ear since the signal was in the noise.
|Method: Apache 10E with GPS into trap vertical|
Soapbox: Barely out of noise here on 40m and no copy on 80m
|Method: Elecraft K3 with EXTREF option
Homebrew GPS disciplined 10 MHz oscillator
Soapbox: Signals were good for both times, but much more interference on 40m at 05Z.
|Method: FTdx3000 Compared against CHU on 3330 kHz. Recorded wav file of audio tone at 3598 and 7064 using 64000 kHz samling rate. Used Cool Edit 2000 analyze tool to measure freq of tone.
Measurement of FMT conducted at 05z|
Soapbox: Had better signal in Michigan for 05z test.
|Method: Receiver Yaesu FT-101EE, long wire antenna. HP3325A with GPSDO hetrodyned with unknown, beat note measured on android tablet running keuwlsoft Spectrum analyzer app.|
Soapbox: 0200 run was very poor, both signals right at noise level. 0500 run was better, 0500 run measurements submitted. 40 meters had very nice signal with a bit of QSB, 80 meters was a fight with the noise, difficult to get a good measurement.
Thank you Connie.
|Method: Icom 756 Pro III. 1000 Cycle tone from I-Pad to get audio beats against signal tuned 1 Khz higher on LSB. Calibrated against WWV same way. Counted beats above and below to find center freq.|
|Method: K3s calibrated against WWV and tuning forks|
Soapbox: Good signals
|Method: Flex-3000 using WSPR. Used the K1JT method for measurement|
Soapbox: 40 meter signal was quite weak here in KC. 80 meter signal 599.
|Method: Receiver used was Harris RF-590, audio was recorded and processed with Mathematica signal processing software.|
Soapbox: Conditions on 40m were poor. Low signal strength and some noise. Enough of a signal to process, however. 80m was better. Frequency wandered around some, but was stable for a short time.
|Method: Elecraft K3 REF CAL calibrated against WWV at 10 MHz before test. This method is described in the Elecraft Owners guide as method 2.
Antenna is inverted vee dipole cut for 40 meters with 20 foot apex.|
Soapbox: First attempt at FMT on 40 or 80 meters. I had measurable signal on 40 meters (7.064 MHz) but no discernable signal on 80 meters at 0215 UTC or 0515 UTC. Therefore, no measured signal 80 meters
|Method: IC7300, G5RV, and Fldigi on Linux. I was in CW mode...I hope I compensated correctly.|
|Method: Yaesu FT-991A plus Spectrum Lab software, calibrated using WWV at 5 and 2.5 MHz.|
Soapbox: I could not hear the early run on 40 meters, no signal at all. But the early run on 80 meters was strong and easy to measure. Thanks for all of your work on the FMT, always fun!
|Method: Yaesu FT-1000MP referenced to WWV 5Mhz, and Android app - Science Journal to measure the audio freq. Yaesu set to .625hz steps.|
Soapbox: Picked up both stations very well on the early test using a LOG (loop on ground antenna). I couldn't pick up WWV at 10Mhz to check calibration there so just used WWV at 5Mhz.
|Method: All homebrew equipment including a carrier phase tracking receiver, a WWVB referenced frequency standard, and a frequency counter. Method involved phase locking the receiver's LO to the frequency of the unknown carrier minus a 1 kHz offset, measuring the frequency of the receiver's LO over several 10 second integration periods, and taking a "selective average" of the readings after factoring in the 1 kHz offset. Details available at: http://www.qsl.net/kd2bd/fmt-methodology.html|
Soapbox: While measurements were made for both runs, results from the second run were submitted since they were more stable and higher in frequency than the first. Specifically, the late 40M run was 0.239 Hz higher than the first, and the late 80M run was 0.402 Hz higher than the first. The Kp index was low (1), and signals were very strong. The first 80 meter run suffered from some rapid phase reversals after signals emerged from deep signal fades. It was nice to hear Connie recognize the passing of fellow FMTer WA1ABI after the first run, and make some CW contacts after the second. Hope to be able to continue using WWVB as a frequency reference in the future!
|Method: FLEX 6300 calibrated to WWV & CHU
Spectrum Lab software to average the signal.|
Soapbox: 40 mtrs, only S5 and lots of QSB
80 mtrs, S9+10 and still some QSB
|Method: Flex 6400 (no GPS or external freq standard), SmartSDR, fldigi|
Soapbox: First time entering. Had a good time and hope to improve over time.
|Method: Turn on equipment. Receiver is Winradio G313. Check calibration with CHU and a 1MHz standard. Antenna is a 136 feet OCF dipole, used on 40M and 80M.|
Soapbox: 40M was weak, with slow fade. 80M signal was over S-9. Casey KG0HY
|Method: ts-990 on cw with pitch set to 1000Hz. read error from 1000hz on counter|
|Method: RSP2 calibrated with WWV.|
|Method: HP3586C Spectrum Lab|
Soapbox: Weak signals -80 & -83 DBm
|Method: GPSDO with Fluke 6060A/AN signal generator beating with rf signal using Icom IC-7300 in A.M. mode. Beat frequency measured using SpectrumLab.|
Soapbox: First entry in FMT. 40m signal at 0200Z was unreadable; others copied with some QSB.
|Method: Homebrew GPSDO & uncalibrated Icom R75 Receiver. Use 10MHz from GPSDO to generate Receiver calibration.|
|Method: Receiver is Flex-1500 locked to GPSDO, in CW mode. Analyzed ~600 Hz tone with Spectrum Lab. Selected most probable spectrum peak then corrected for Flex's DDS VFO offset.|
Soapbox: Local sunset was 01:00Z, one hour before the "Early" transmissions.
Both "Late" signals had slightly higher 'best' frequency peaks.
|Method: IC 7200 with end-fed antenna.Used WWV and CHU to acquire correction data for radio near each band. Using Spectrum Lab, averaged 100 samples on each band and applied corrections.|
Soapbox: Thanks as always to K5CM for hosting this fun event. K5CM 599 and in the clear on both bands. Almost missed the FMT altogether. only stumbled upon the announcement Wednesday night. Did not have the usual time to prepare and stabilize the computer and radio. We'll see.
|Method: Too far to be precise, doppler_, who knows! just for fun, SDRplay calibrated with a GPSDO, HDSDR/ARGO.|
|Method: Equipment: Elecraft K3 with P3 on AV640 vertical. AMD CPU PC, Spectrum Lab. Method: 120 min equipment warm up. For 40 meter test, tune WWV at 5 MHz; 400 Hz CW filter. Zero beat calibration of LO. Mark 600 Hz on Spectrum Lab. Zero beat test signals and compare/zero with Spectrum Lab. Zero beat to 5MHz WWV for 40 meters and then zero beat 2.5MHz WWV for 80 meters while checking 600 Hz zero align with Spectrum Lab.|
Soapbox: It's been a while since I have run this and as always a lot of fun. Strong signals from K5CM here in Hill Country Texas at new QTH.
|Method: Kenwood TS-2000 referenced to GPSDO 15.6 MHz clock, SpectrumLab for audio frequency counter. Turned on radio at 0159z. Reference check after each rx.|
Soapbox: Thanks for doing this Connie! I look forward to this every time!
|Method: HP 3586A signal level meter locked to HP Z3816A GPS reference.|
Soapbox: I definitely need more practice. 80 meter reading is the CW callup. Thanks for the test.
|Method: DigiPan V2. FT3000, Dipoles.|
Soapbox: Signals weak in STX.
|Method: IC746pro, zero beat on LSB mode.|
Soapbox: Thanks for putting on this FMT. 73
|Method: Yaesu FT920 zero beat to 5Mhz WWV. Determined correction factor of .9999972 . Direct reading from radio display x cf.|
Soapbox: No fancy equipment here... just using what I have.
|Method: IC-7300, Spectrum Lab software and WWV|
Soapbox: It should have been eazy. The signal was strong and had little doppler shifts.
|Method: Kenwood TS-480 radio and a computer to count cycles above a U3S ref signal.|
Soapbox: I received strong signals on both bands for the early test.
|Method: Icom IC-7200 and dipole; used K1JT FMT program calibrated with WWV.|
Soapbox: Good signal strength at 0200/0215Z on both bands. I didn't try the later run. Thanks!
|Method: Flex 6600M with GPS lock. wsjt-x 1.9.1 frequency calibration software.
get_error result from Flex API to correct dial frequency (-4 mHz on 80, -19 mHz on 40) Late test results only|
Soapbox: Interference on the early 40m test from commercial AM broadcast tower on O'ahu with two stations that mix to form many harmonic combinations with one right on 7.064. Also lots of static crashes from nearby thunderstorms. No 80m signal for the early test. Better signals for the late test, no O'ahu interference on 40 and good copy of the 80m signal. Still lots of static crashes taking out about 25% of the data . My first try. 73 and Aloha, Tom, NH6Y
|Method: SDR Play RSP 1 with 100 ft. Longwire antenna. Checked CHU at 7.85 MHz and WWV at 5 MHz before runs. Calibrate phase and tune to K5CM. Software is HDSDR. Used ECSS Mode and AFC and logged Lock Frequency. Also checked Sub Receiver on Ten Tec Orion and zero beat after using WWV for offset correction.|
Soapbox: Poor copy on 40 M first run, slightly better on second. 80 M signal solid in Minnesota on both runs. Thank You for your hard work on this project. Fun! 73, Mike Gorniak, NM7X
|Method: dipole -> GPS-clocked XCVR -> sound card -> FFT Hann window with A&R interpolation. Both 40M sigs very close in freq; later 80M higher freq then earlier but submitting later run values (better SNR).|
Soapbox: Another FB FMT, thanks Connie, Bruce. Not sure why later 80M appeared higher in freq while 40M did not. No Doppler "corrections" applied.
|Method: Ten MHz ref from GPS to lock sig generator, set to about 500 Hz below target signal. Signal detected in AM mode and audio beat note measured with SpectrumLab software. GPS has been running continuously for almost two months.|
Soapbox: Wow! Lots of scatter in my results. Observed only the early runs (02Z). Signal on 40 m was barely detectable in power line noise here. The 80 m signal was strong but with much QSB. Hope my results are at least in the ball park.
|Method: TS-870 with GPSDO and Fldigi.|
Soapbox: Just did the early run. 40 m was S0 and hidden under LSB. Still, Fldigi seemed to find the carrier. 80 m was S9 and in the clear. 80 m showed about 0.6 Hz doppler while 40 only showed about 0.1 Hz - which made me wonder if Fldigi was making up the numbers when the signal was hidden :-)
|Method: Yaesu FT817ND with TCXO upgrade, inverted V linked dipole antenna.
Home-brew audio interface to USB sound-card,
Laptop computer (Windows 10) running WSPR 2.12|
Soapbox: First time trying the FMT.
Equipment is basically what I use for SOTA activations.
I don't operate in town very often, Boy, is it noisy!
|Method: Receiver was IC-7410, CW mode 200 Hz BW. Used a Trimble 57963 GPS 10MHz clock and Rigol DG1032Z signal generator to give reference signals at frequencies very near the on-air signals. Spectrum Lab on a laptop was used to measure audio tone frequencies for on-air and reference signals. FFT bin size was 0.061 Hz and about 15 measurements were averaged for each tone. RF Freq = sig gen freq + (RF audio freq - sig gen audio freq).|
Soapbox: Thanks again for the challenge. Signals peaked S7 on 80m with rapid QSB. On 40m signals were inaudible most of the time, but still measurable with Spectrum Lab.
|Method: SDRplay - RSP1 and the new SDRplay Spectrum Analyzer software.|
Soapbox: Good signals to the east coast on both freqs.
Downloaded the analyzer software about 1/2 hour before the start of the Test. Once again extreme care and thoughtful planning failed to enter the equation at this QTH.
Tnx Connie for doing this. I like the new (or old) format.
|Method: Yaesu FT DX 3000 Direct Frequency Measurement Method|
Soapbox: Thanks again Connie! Bands were very noisy in Cedar Rapids, IA
|Method: E-MU 0204 soundcard and RS-HFIQ SDR with Fox924 TCXO for LO. Spectrum lab. 192 KHz sampling rate decimated by 12, 262144 input FFT. Over-the-air offsets calculated from CHU3330, WWV5000, and CHU7850 then interpolated linearly.|
Soapbox: Thanks for running the FMT again.
|Method: 80/40/20 fan dipole + NetSDR + custom Gnuradio stack + Digital RF (https://github.com/MITHaystack/digital_rf) + Python/Numpy/Scipy custom scripts|
Soapbox: Second time around. Used only the 02 UTC one as I was collecting data while away from my QTH, but both carriers seemed to come in well here. CHU at 3.33 MHz was not cooperating before the first call-up but came in strong afterwards, so hopefully the frequency cal is right. It remains the case that I/Q recording and ability to refine analysis are key. Hopefully the answers are somewhere within the ballpark. Thanks once again to K5CM. 73s to all.
|Method: IC7100 Transceiver--600 Hz. Audio Filter--Audio Generator--Oscilloscope--Frequency Counter/Use audio output~600 Hz. from IC7100 for CW and R-CW modes and view sinewave pattern on oscilloscope with audio generator. Audio generator adjusted to give stationary pattern and frequency of audio generator and receiver output measured with frequency counter. Results of near 600 Hz. CW and R-CW measurements used to obtain 0.1 Hz. resolution for IC7100 dial display. Dial calibration was obtained from WWV 15 MHz. calibration measurements.|
Soapbox: The signal phases for the early period were quite erratic for both 40 and 80 meters even though the 80 meter signal strength was very strong (S-9 +10 dB.). Both late period signals were S-9 + and the signal phases did not behave as badly as the early period phases. I was able to get the one Hertz resolution from both sets of measurements but the second set of tests was necessary to make reasonable measurements for 0.1 Hz. resolution. Thanks to Connie, Bruce and the staff at ARRL involved in the FMT program for their efforts.
|Method: I use a KWM-380(mod) and a sound card running Spectrum lab SW. Both are locked to a GPSDO. The antenna this time was a 41 ft. vertical matched on 80M. I use the '380 in USB with a narrow filter, and just add the SL reading to the dial freq.|
Soapbox: My measurements of the early and late transmissions on 40M agreed to within a few mHz. However, the early and late 80M readings differed by nearly 0.2 Hz. The Spectrum Lab plots showed the early 80M received energy to be spread over nearly 1/2 Hz, so I chose to report the measurement of the later 80M transmission.
|Method: OCXO,TCXO K3, Spectrum. Measured offset using LSB method. Calculations via Excel spreadsheet. Vertical antenna. Early run|
Soapbox: Good signals on early run with minimum Doppler. Some small differences measuring WWV and CHU before run and after. Used the after measurements for correction. Thanks to K5CM.
|Method: Elecraft K3 w/TCXO option; 30 Ft. vertical; Spectran audio analysis software. Prior to test, calibrated K3 against WWV (5 & 10MHz). During test, recorded K3 frequency on both bands where 600Hz sidetone frequency measured the same in Spectran while alternating between CW and CW-R modes.|
Soapbox: Measurements taken on the late run. Signals loud and solid here in Maryland. Always enjoy this event. Thanks and 73! Denis
|Method: Flex 6700 (with GPS locked ref. osc.) into Spectrum Lab|
Soapbox: Early 40 M sig was S8 with severe QSB/ Early 80 M sig was S9+15/ Late 40 M sig was S9+10 and the Late 80 M sig was S9+10. Doppler shift and multipath were not "totally impossible" but merely "highly improbable"! Tnx to all who make the test possible; especially Connie----
|Method: Kenwood TS2000, Bodnar GPSDO, Spectrum Lab. Calibrated TS2000 frequency error, measured transmitting station tone, applied correction factor and hoped to get by the doppler shift.|
Soapbox: Thanks to K5CM for the time and effort put into the activity. Always a learning experience for me. Good luck to all. Mack W4ENN
|Method: Synthesizer injection into receiver. Measure difference with audio-card spectrum analyzer.|
|Method: HP3586B with 102 ft doublet antenna at 30 ft. Used external Trimble $20 eBay 10MHz oscillator calibrated to WWV. Read freq off screen for 20 and 40M - good signal lock.|
Soapbox: Thanks Connie! Good -60 dBm signal on 40M and -56 dBm on 80M into NC.
|Method: GPS: Lucent/Symmetricom Z3810AS
RECEIVER: GPS Locked RACAL 6790/GM
Tunable in steps of 1000Hz, 30Hz or 1 Hz.
SYNTHESIZER: GPS Locked HP 3336B
ZERO BEAT INDICATOR: Tektronix 2213A
ANTENNAS: GAP Voyager, GAP Titan DX, NVIS Dipoles 40 & 80m
CONFIGURATION: The post-filtered 455 KHz IF from the RACAL is fed to CH A
of the Tek. A precise 455,000.000Hz carrier from the HP3336B is fed to the EXT input channel of the Tek. The scope is then set to display 2 cycles of the 455KHz IF waveform, instead of a traditional Lissajous pattern.
PROCEDURE: Frequency readings to 1Hz resolution are quickly made by tuning the RACAL to near zero beat with the aid of the Tek scope. If a signal is “dead on”, such as WWV, and there is no Doppler, the waveform will not “slip” in either direction but will just vary in amplitude with changes in signal strength. If the waveform is moving >>>, the RACAL is tuned BELOW exact zero beat. If the waveform is moving <<<, the RACAL is tuned ABOVE exact zero beat. If “stepping” the receiver 1Hz causes the direction of "slip" to reverse, the exact frequency is BETWEEN the current and former readout, i.e. between two concurrent 1Hz points. Also, estimates closer than 0.5Hz, of the actual
frequency, can be made by comparing the "speed" of the ABOVE and BELOW "slip." The slower "slip" is less than 0.5Hz from that 1Hz point.
For more accurate “in-between-Hz” measurements, the RACAL is tuned ABOVE the unknown frequency and the HP3336B is then tuned upward, in 0.001Hz steps, from 455,000.000Hz until the right to left waveform movement stops. The milliHz DELTA is then SUBTRACTED from the RACAL readout to give the “exact” frequency.
The AF DELTA is simply the result of subtracting the lower frequency from
the higher frequency.
Using a regular scope waveform allows very quick “zeroing” of signals, even those that are nearly in the noise. I find it much easier to use than the Lissajous
pattern. It is also very interesting to note the carrier frequency variances that occur during Doppler shift.|
Soapbox: EARLY SESSION:
40m; minimal Doppler; Signal -80dBm QSB to -100dBm. Spanish SSB QRM @ -110dBm. Not a problem!
80m; minimal Doppler; Signal -60dBm light QSB.
40m; moderate Doppler; Signal -80dBm QSB to -100dBm. Lid sweeping across K5CM frequency. No problem.
80m; moderate Doppler; Signal -50dBm light QSB
40m DELTA between EARLY & LATE SESSION +0.04Hz
80m DELTA between EARLY & LATE SESSION +0.16Hz
I posted the EARLY SESSION numbers because of the very minimal Doppler. TWO VERY STABLE paths from K5CM, for a change!
TNX to Connie for his hard work!
|Method: Flex 6600 with Fldigi4.0.18|
Soapbox: Thanks for the FMT. It's always fun.
|Method: 2-trace phase matching + SpectrumLab assistance|
Soapbox: Thanks again. Later signals significantly doppler widened
|Method: An LPDA antenna feeds a coherent receiver whose 2nd IF is feeding an HP-5370B counter, and all clocks are locked to a GPS disciplined 10 MHz standard. A PC samples the counter twice a second, then post-pass a least-squares fit determines an estimate of the received frequency. WWV is used pre, and post pass to determine a best guess at the Doppler correction to apply with the same setup.|
Soapbox: Thanks to the sponsors for another fun FMT. All signals both early and late put about -75 dBm into my receiver where the noise level was -105 dBm, so I had a beautiful SNR, and no excuse to get it wrong. I was a bit surprised to find the 40 meter signal right on top of an Asian SSB QSO, but they didn't interfere too much.
|Method: HP Z3801Z GPS receiver 10 MHz output driving HP3586B selective level meter|
Soapbox: K5CM signal was very weak on the early run. 80 meter early run signal was good. On the late run both 40 meter and 80 meter signal were good but with the usual Doppler shift.
|Method: Shera GPS Locked 10mhz, Icom 756 locked to the GPS with providing 1hz Digital Readout, Spectrum Lab and SignaLink GPS, 80 meter Dipole at 70 feet.|
Soapbox: This new format was NICE. I earlier FMT tests, we on the west coast have to contend with frequency shifts caused by sunset that cause real shifts in frequency. These can be all over the place. This double sets transmissions on each frequency allow us to chose which time seems best either early or later. It also allows us to see the actual shifts caused by the time of day. Should be interesting when the actual results are released.
Connie, Thanks for all the work and your time spent. I is greatly appreciated.
|Method: IC-703+, FLDigi in Freq Analysis mode, 40m dipole @30 ft.|
Soapbox: 80m was challenging as there was a nearby digital signal. 40m clean copy.
|Method: IC-718, SignaLink USB, fldigi, indoor wire antenna. Calculated error function with WWV. Applied to direct measurement from fldigi.|
Soapbox: Heard both east- and west-directed signals on 40-meters. I may have heard the very weak east-directed signal on 80-meters, but nothing on the later west-directed signal. Opposite of what I expected. Thanks to K5CM for providing this; I always enjoy it.
|Method: IC-706 MKIIG with 30 MHz phase-locked to GPS with Shera controller. Spectrum Lab for Fourier Transform processing.|
Soapbox: The double transmission timing from K5CM is a great idea. Conditions were general good. Thanks for the event.
|Method: Modified HP 3586B Selective Level Meter, HP 3326A Synthesizer,and GPS disciplined oscillator. (with the exception of the GPSDO -boat anchors all)|
Soapbox: Always fun. Signals good here in southern Maryland for both time slots. Thanks to K5CM for the early/late session FMT. Gives me two chances to screw this thing up.
|Method: Elecraft K3 calibrated vs. 10 MHz WWV. Fractional Hz are guesses.|
Soapbox: K5CM strong, WWV very weak.
|Method: FlexRadio 6500 locked to Trimble Thunderbolt GPSDO. HP3325A synthesizer also locked to GPS produces a 2 kHz reference audio tone, used to calibrate the sampling rate of a Denon professional SD card audio recorder. This time I simultaneously recorded WWV at 5 MHz in a second receiver slice to try to get a correction factor for ionospheric Doppler. Different frequency, different path, but may be better than nothing. But WWV was solid, very low Doppler. Got two measurements of 5000000.00 and two measurements of 4999999.99 Hz. Rewrote my GNU Octave code (which does the frequency measurements) to measure WWV and apply ionospheric corrections, but there was almost nothing to correct.|
Soapbox: Great fun as always. This time I measured WWV on 5 MHz simultaneously with K5CM. Saw much more frequency variation on K5CM than on WWV. Theories (1) vertical antenna at WWV versus horizontal at K5CM or (2) WWV signal was stronger (less noise) resulting in less frequency variation measurement on WWV signal. WWV measured spot-on, 0.01 Hz or less error, so maybe there wasn't much Doppler, at least not at 5 MHz.
Soapbox: The early readings were 80 MTR signal –90 dBm with S/N 55 dB and no QSB. 40 MTR signal -100 dBm with S/N 30 dB and no QSB. The late signal strength readings were almost identical to the early. The 40 MTR late frequency was almost identical to the early. The 80 MTR late frequency was read as almost 190 mHz higher. My submitted readings were the early tests for both 40 MTR and 80 MTR.
|Method: Icom 706MkIIG CW mode 600Hz through filter to scope comparing to 600Hz from DDS generator for roughin. Final reading done with WSJT-X in freq cal mode.|
Soapbox: Signals were strong here on both frequencies during early session. (Went to sleep before late session.)
|Method: HP 3586, HP 3336, locked to gps, scope. Spotting with Apache 7000.|
Soapbox: Excellent propagation - very little doppler - no qrm.
|Method: Icom IC 7300, OCF antenna at 40ft.|
|Method: 40 M dipole HP SLM Spectrum Lab as a tuning indicator...|
Soapbox: Thanks for the FMT Connie. No early 80 m signal heard. I usually hear it pretty well. 73
|Method: IC-7610 in TX operated by remote control from east coast. Spectrum Lab audio frequency analyzer. 5 MHz WWV calibration.|
Soapbox: Murphy showed up with all his friends. I had to be on a business trip that evening so the initial plan was to use my IC-7610 via remote control. But initial testing showed about 2.5 Hz of drift and frequency wander with this rig. So shifted to remote control of W5YE's IC-7300 via Remote Hams software, but unknown internet issues prevented establishing a stable link. So went back to my 7610 and just made the best of it.
|Method: Rig: Icom IC-R8600
Antenna: SteppIR DB-36
GPSDO: TrueTime XL-DC
Ref Generator: HP-3336C
Software: Spectrum Lab Vers. 2.92
Measured signal and reference sequentially during the FMT. Set up SL with a 10 Hz measurement window prior to the FMT, but widened it 20 20 Hz for 0500Z measurements. Used the call-up period to tune the signal so that it was roughly centered in the measurement window and then adjusted the reference signal to also be within the window. Did not attempt to correct for doppler shift.|
Soapbox: 0500Z 40m signal was 30 dB greater than 0200Z 40m. Was much more confident in 0500Z measurements due to better SNR, but had identical 40m measurements, while 80m measurements only differed by 0.19 Hz, so I reported the 80m average.
|Method: Elecraft KX3 Receiver in USB with Spectrum Labs software & QRP Labs Signal Generator locked to 10 MHz Rubidium Frequency Reference. Used frequency difference method based on VE2AZX & K5CM tutorial for setting up and using SpecLab to compute difference.|
Soapbox: Call-ups were weak. Lots of local QRN but narrow FFT BWs resulted in 20+ dB of test signal SNR. I should have practiced more with setting up SpecLab to write freq. difference data directly to a file, rather than scribbling numbers. Post-processed data to determine average to subtract from siggen frequency. Thanks guys for providing the FMT. PS: Tutorial screen shots need to be updated.
|Method: FT-991A using wsjtx frequency cal mode|
Soapbox: Interesting test. The frequency I measured bounced around a bit, with these numbers being averages. I averaged over 26 seconds in 2 sec increments on 40 and only 18 seconds in 2 second increments on 80.
|Method: HP3586B, Spectrum Lab, 80m dipole.|
Soapbox: TNX to K5CM and others.
|Method: Used Les Ettenger's (W6JKA) QTH. Flex6700 with horizontal loop and my HP3336 synthesizer, Tek 224 scope and Jackson Labs GDO for precision 10MHz going to the Flex and the HP synthesizer. We did out best to match the phase of the receive audio to the synthesizer.|
Soapbox: I noticed more Doppler shift during the second session than the first. We had good signal to noise for both sessions with stronger signal for the second session.
|Method: Icom 718 and sound card.|
Soapbox: Really enjoy trying this! It’s a real challenge for me, been waiting since April!
|Method: IC-718 controlled by Flrig and Fldigi.Calibrated sound card and warmed equipment for 2-3 days. Monitored WWV 5000 and WWV 3330 frequencies to find most stable reference by looking at error average,error max and error min. Take the stable frequencies and calculate error per hertz. Calculate hertz difference between reference frequency and key down frequency. Then add or subtract error per hertz to key down frequency. This seems to get me close with a shoe string budget. Constantly improving my spreadsheet so its lots of fun. Thanks to Connie for his time and equipment. I just did the east coast run and all signals were very strong in Kentucky.|
Soapbox: Not sure why this is so much fun? Did the early run and Connie's signals in Kentucky were very strong. I think I have a reliable method now as long as I don't make a mistake.Thanks again Connie and I look forward to April!
|Method: Used Kenwood TS-180s in CW mode with 500 Hz filter as the receiver, and zero beat K5CM using a DDS Signal Generator and measured the signal generators frequency real time using my home brew frequency counter that was calibrated (about 9 months ago) to WWV using my 10 MHz WWV TRF receiver. Immediately after the FMT I checked the calibration of my frequency counter using my 5 MHz WWV TRF receiver and my frequency counter was less than 1 hertz off. Frequency counter resolution = 1 Hertz.|
Soapbox: Decided a few hours before the FMT to take part in my first ever FMT using the very basic zero beat technique and really enjoyed the event. Also enjoyed working Connie (K5CM) right after the Late Run on 80 meters to record the moment in my log book. Hope to be within a few Hertz of reality. 73, Don (wd8dsb)
|Method: Flex 5000, external Rubidium standard, FLDigi operated in Frequency analysis mode.|
Soapbox: Signals were very loud in the s7-s9 range for all sessions. Kind of nice for the November session. Very little QSB and no QRM. My noise level at this QTH on 40m and 80m is about an s3. I enjoy these tests but I missed the last couple. Sometimes I forget so this time I set my alarm as a reminder.
|Method: HPSDR Mercury & CHU calibrations|
Soapbox: first time!
|Method: Elecraft K3 receiver, FLDIGI and 'Spectrum Laboratory' software, Fluke 1953A counter with super precision crystal oven option used as 10 MHz timebase for HP 3325B frequency synthesizer for reference signals. The error from the reference signal was subtracted from the obtained results to reduce errors from the receiver and sound card clocks.|
Soapbox: This is my first FMT contest!
|Method: KX3 & spectrumlab|
Soapbox: Thank y'all for putting this on. It's fun.