The 2002 FMT
In the 2002 test, I copied three of the four test frequencies. Here are my results versus the official measured frequency on those bands:
| Band | N8UR Measurement | ARRL Measurement | Error |
| 80M | 3 579 334.69 | 3 579 335.2 | +0.5Hz -- +1.4x10-7 |
| 40M | 7 046 882.39 | 7 046 882.3 | +0.09Hz -- +1.2x10-8 |
| 20M | 14 049 072.27 | 14 049 083.8 | -11.5Hz -- -8.2x10-7 |
If you're interested, here's my ARRL Submission Letter (pdf).
I made a horrendous error on 20M; I either copied a reading incorrectly, or screwed up the math somehow. On 80M, the error was primarily due to speading of the received signal due to multipath. On 40M, the results were quite respectable.
I used my ham rig, a Kachina 505DSP, as the receiver with an HP8920B service monitor as the signal generator. The 505DSP has a very good receiver, with a wide selection of bandwidths, but its inability to disable the AGC was a serious disadvantage; it was necessary to adjust the signal generator's level quite carefully to keep it from swamping the W1AW signal. I used an HP Z3801A GPS-disciplined oscillator as the reference controlling the 8920B's frequency.