Monday, February 6, 2023

The utility of Elecraft's CP1 bi-directional coupler with the tinySA

Set-up for measuring spectral output of 100W rig. Note the 100W dummy load behind the tinySA

This is an example of where the availability of a new item adds value to something that's been around a while. The new item is the tinySA - many hams now have one of these small spectrum analyzers/signal generators - and the old item is Elecraft's CP1 bidirectional coupler.

Questions I get whenever I post something related to the tinySA are:

  • Can I use it with non-QRP equipment?, and
  • How do I measure a transmitter's output to make sure I don't exceed the tinySA's 6dBm (4mW) max input?

I decided to use my tinySA ULTRA to measure the spectral output of my Yaesu FT-891 when operated at full power: 100 watts.

100 watts, expressed in units that are more helpful, is 50dBm. That means I need to attenuate the FT-891's output by at least 44dB for safe input into the tinySA. Personally, I'd rather shoot for an input to the tinySA of 0 dBm (1 milliwatt) - this means 50dB of attenuation.

Most fixed and/or step attenuators cannot handle a 100W input - but Elecraft's CP1 has a maximum input of 250 watts, provided it's built to provide a -30dB sample (It can also be built to provide a -20dB sample but it's maximum allowable input then decreases to 25 watts).

Mine is built for a -30dB sample. This means that my 100-watt input produces a 100mW output at the forward sample port of the CP1. This is a low enough level to now apply to other attenuators for further reduction in power.

Put another way, 50dBm (100 watts) from the FT-891 has now been reduced by 30dB to 20dBm (100mW) - so we need 20 more dB of attenuation to get to 0dBm (1 mW).

A step attenuator makes it easy to dial in this or any other value within the attenuator's range.

At this point, it's obvious that we aren't measuring the 0dBm output from the attenuation chain for input into the tinySA - we are calculating it based on a known (or measured) 100 watt output from the rig going into known levels of attenuation.

100 watts, attenuated by 50dB, is 1 milliwatt. As they said in Titanic, "It's a mathematical certainty".

What we do have to know is the accuracy of the attenuation. This is where a nanoVNA comes into play - it's the perfect device to verify attenuation levels at the frequency of your application.

Shown below is a nanoVNA sweep of my Elecraft CP1 from 1-30 MHz:

Elecraft CP1's attenuation characteristics from 1-30 MHz

Keep in mind that the CP1 still has 100 watts going into it - that power must be dissipated at the CP1's output port, therefore a dummy load sufficiently sized will be needed (shown in top photo).

So how does the FT-891's spectral output look at full power? In a word, excellent:

In the photo above, the 2nd harmonic is 63.5dB below the fundamental frequency.

The tinySA (and most spectrum analyzers) also display the power of wherever a marker is placed. Marker 1 shows -19.7dBm at 14.054 MHz. This is after going through a total of 70dB (30dB from CP1 and 40dB dialed in with step attenuator) of attenaution, so the real power out of the ft-891 is shown to be:

-19.7dBm + 70dB = 50.3dBm, or 107 watts.

On 80m:

Again, excellent harmonic suppression - and the FT-891's transmitter output can be calculated to show exactly 100 watts out, based on 70dB of attenuation and -20dBm of power shown at Marker 1.




  1. That was cool, if you so desire, under the menu setting "Level" and "Ext Gain", type in with a minus sign the value of the total attenuation applied, and the marker will read the output of the radio directly... I really enjoy the articles about the NanoVNA and the TinySA, different things to try... Thanks again, I enjoy your posts everyday..

    1. That's true, the tinySA will do the math for you. I usually do it myself since I'm varying the attenuation to arrive at the threshold rather than varying the level out of the tinySA. Maybe I'm talking the long way around the barn...