A common design architecture in modern transceivers is that all ALL receive oscillators and tx sidetone oscillator are synthesised from ONE master oscillator. If that one oscillator is correctly on frequency, everything about the transceiver frequency is correct.
When switched to CW mode, the tx sidetone frequency should be
exactly the same as the sidetone from a carrier at the
displayed frequency. So, for example, if the receiver is tuned to
10.000000MHz in CW mode, the beat note produced from
WWV (provided that it is strong enough to be heard) should be at
exactly the same frequency as the tx sidetone
oscillator. An alternative to WWV is to use a local frequency standard
such as a GPS disciplined oscillator, rubidium standard etc.
If the transceiver VOX is turned off, so that the transceiver does not transmit with key down, but produces sidetone, if the receive frequency is adjusted in increments of 1Hz, a point will be found where the sidetone oscillator beat with the beat note from WWV's carrier is less than 1Hz. The difference in the displayed frequency and the known WWV frequency is the error in the frequency calibration.
For a demonstration with commentary, listen here. The demonstration is performed on a TS2000 from the front panel without removing covers. The TS2000 can tune in 1Hz steps, but uses the RIT/XIT control for the fine steps. Note that some aspects of the demonstration (eg sidetone frequency, actual error) are dependent on the operating configuration, but the system works the same on any TS2000, regardless for instance of the sidetone choice.
Most modern Icom transceivers provide simple access to 1Hz stepping though the TS (tuning speed) button and direct readout, see your manual. Though Icom produces some radios that are difficult to calibrate (eg IC-910 which requires not only removal of covers allowing the unit to cool, but also removal of the 23cm module for access to the adjustment), the IC-7000 permits frequency calibration without removing the covers, and without tools.
It is not unusual that a supplementary crystal oscillator is used in FM tx because of issues with modulation of the signals from a DDS, so frequency accuracy on FM tx may depend on one further adjustment.
In transceivers that use DSP modulation and demodulation, ie a DSP at Intermediate Frequency, the reference clock used by the DSP block influences frequency calibration.
It is not uncommon that the DSP clock is not derived from the master oscillator, and that there is no provision for adjustment of the DSP clock frequency. In such cases, the DSP clock will contribute a constant error component at all receiver frequencies.
For example, measured frequency error of a TS2000 at 10MHz is 0.43Hz, and at 150MHz is 43.8Hz. Fitting these to a linear model y=mx+b gives m=(43.8-0.43)/(150-10)=0.31, and b=y1-mx1=0.43-0.31*10=-2.67Hz, so error=0.31*f-2.67Hz. The -2.67Hz is the error component due to the IF DSP clock which is not adjustable in the TS2000.
To calibrate a transceiver such as this, measurement of frequency
error at two widely spaced frequencies allows extraction of the DSP
clock error component, and calibration of the master oscillator in the
presence of the DSP clock error. A simpler technique is to just
calibrate the transceiver at a very high frequency where the DSP clock
error is less significant. In the case of my TS2000, I usually
calibrate it at 150MHz to within 10Hz using the fifteenth harmonic of a
10MHz frequency standard.
Note that PLL and DDS circuits do not have infinitely fine frequency resolution, but most implementations should have sufficient resolution to support calibration to 1Hz resolution, although the calibration will not hold due to the short and long term drift of the oscillator.
The first step to using this technique is to understand your transceiver, how it generates the various oscillators. As time passes and we dumb down, user manuals are less likely to contain the necessary information which by its omission is deemed to not be of interest to users, you may need to locate a service manual (often available at no charge on the 'net).
© Copyright: Owen Duffy 1995, 2019. All rights reserved. Disclaimer.