Even the much vaunted Bird Model 43 only claims an accuracy of 10% of the full scale reading of the particular slug in use. The accuracy of the bridge and detector network used in amateur amplifiers can be fairly good, if care is taken during construction and setup. But, without calibration against known standards, all of the common ham power meters fall short of mil-spec. The point to remember is that the difference between 1000 watts and either 900 or 1100 watts is very difficult to discern in actual on the air practice. In my lab days, I have spent the most part of the morning optimizing couplers with a network analyzer on the bench, and it gets more difficult if the coupler has to have the same coupling factor and directivity over a wide frequency span such as the HF spectrum. If you keep a technical log, it makes sense to record maximum output power into a dummy load on each band using the onboard meter. Years later, if you suspect the final amplifier tubes are going soft, repeat the test and draw conclusions. But don't expect the onboard meter to be lab standard.
73 de Norm W1ITT
Nope. The bird + CD slugs are +/-5% of full scale....anywhere on the scale. (with the pep kit, on the CD meter, it's then +/- 7% of full scale...anywhere on the scale. With the pep kit on the bird meter, it's then +/-8% of full scale...anywhere on the scale.
Bird / CD meter's are typ calibrated at 21 mhz, when their 2-30 mhz slugs are used.
The problem with the bird / cd meter's is..... you need to use the smallest slug for the job.
IE: +/- 5% of a 1000 watt slug (with meter displaying 1000 watts) is +/- 50 watts. IE: when meter reads...'1000 watts', the real power could be anywhere from 950 watts..to as high as 1050 watts. If meter reads ...'500watts', real power could be 450-550 watts (now as much as +/-10% outa whack).
IF the same 1kw slug used to read just say 100 watts, the real power could be anywhere from 50-150 watts. (now as much as +/- 50% outa whack).
On my L4B, in low power CW mode, with a 5 kw slug it reads 700 watts. With a 2.5 kw slug, it reads 650 watts. With a 1kw slug, it reads 625 watts. When testing using a calibrated scope across the 50 ohm DL, it works out to 620 watts.
On CW/ SSB / AM, without the pep kit, they just show quasi average PO.
Both (now discontinued) Alpha wattmeter's, they used a freq counter...and a look up table, for a separate calibration for each band. It would sense RF, and switch to correct calibration for that band..and within just 1 dot at 60 wpm if on CW. On
SSB, cough into the VOX, and bam...new calibration point in use.
On my Array solutions 'power master' (pep only at all times) wattmeter's, there is an option to 'auto subtract' the RVS power from the forward power...to display...'net power'. There is also software controlled separate calibration adjustments for frwd power...and also rvs power. Both frwd + rvs calibration points can be adjusted in 1% increments... from -15% up to +15%. That is done on the display unit, and no pots are used anywhere. Each HF coupler is hand calibrated to nist standards, and the frwd + rvs calibration points are marked on each coupler...then programmed into the mating display unit.
Back then, most of the amps had a ...... 'relative PO' meter, used for tweaking the tune + load for max PO, that's it. IE: SB-200/220/230 / TL-922, FL-2100B, etc