Nichols: The Two Bird Experiment – a Simsmith model

I have received comments from several people regarding Nichols: The Two Bird Experiment, mostly stating their belief that Nichols is correct.

Above is Nichols’ test setup, simple enough.

Let’s look at a Simsmith model of the test configuration with practical values for a Transmatch, ie this is what you should expect to measure with real world components.

But first a reminder that ForwardPower and ReflectedPower are notional quantities that express the actual voltage and current at some point in terms of ‘power wave’ in some reference Zref that MUST be purely real (ie Xref=0).

Since Nichols used a Bird Wattmeter and did not state the reference impedance explicitly, but did imply that Wattmeter #1 used Zref=50+j0 (ie a 50 ohm element) by his reported measurements, and I assume so did Wattmeter #2.

The low loss Transmatch is modelled as C1, L1 and C2 and adjusted for a near perfect match at the source G, and they are configured for real world loss.

Some additional quantities were calculated as above.

The key points are that:

  • forward power pfC1 and reflected power prC1 are calculated looking into C1, pfC1=100W and prC1=269nW=insignificant, and pf-pr=100W;
  • forward power pfLoad and reflected power prLoad are calculated looking into Load, pfLoad=154.6W and prLoad=55.66W and pf-pr=98.95W;
  • the network is low loss, and power is close to 100W though the whole network.

Simsmith has correctly solved this network.

Returning to Nichols:

With the transmitter keyed, the transmatch is adjusted to show zero reflected power on Bird Wattmeter #1. Transmitter is then adjusted to generate exactly 100 watts of forward power indicated on Bird Wattmeter #1. Bird directional wattmeter #2 indicates about 36 watts of REFLECTED power.

Nichols’ solution above is clearly wrong though he has assured us elsewhere that he confirmed his claims with measurement (I always base my claims on actual measurements).  Reflected power at the Load is 36% of forward power (pfLoad=154.6W) and so reflected power is 55.66W (prLoad=55.66W).