Power output from CH0 of the nanoVNA as a function of frequency
Bryan, WA5VAH 10/08/19 #4369 Folks, For some measurements it is important to know how much power is coming out of CH0 on the nanoVNA. I have seen some data elsewhere in this forum and a few spectrum analyzer pictures with max hold enabled. However, I have seen nothing that showed the information I sought. So, I created a set of measured data over the full range of frequencies from 1 MHz to 900 MHz and plotted it as shown in the attached graph, "nanoVNAOutputPower.jpg." I connected the nanoVNA to the power meter using a very wide-band 10-dB pad to ensure the load was consistent and removed the loss in the pad from the power measurements. The data was collected every 1 MHz from 1 MHz to 30 MHz, every 10 MHz from 30 MHz to 350 MHz, every 25 MHz from 350 MHz to 450 MHz, and every 50 MHz from 450 MHz to 900 MHz. All data was collected in CW mode. For me, there were a few surprises in this data. I have read that the maximum power out of the device should be -9 dBm or less. This is true over the range of frequencies below 300 MHz. As you can see in the graph, I observed more power coming out of CH0 with CW frequencies above 300 MHz. I was surprised to learn that for frequencies above 300 MHz, the 3rd harmonic is always assumed in CW mode. This is evident in the power levels shown in the attached graph. I confirmed with a fast oscilloscope that this is indeed the case. When 750 MHz is set, a 250 MHz waveform is coming out of CH0. Perhaps the reasons for this in the firmware is that it produces more power at 750 MHz than using the 5th harmonic and setting the fundamental to 150 MHz. In fact, the output amplitude at the fundamental frequency is increased above 300 MHz and never gets back down to the level I observed at 1-30 MHz. The maximum power I measured was a little less than -7 dBm decreasing to a little more than -10 dBm at 900 MHz as the set frequency. I have read that the 5th harmonic is used above 600 MHz for swept measurements; however, that is not what I saw in the CW data. I assume this is an artifact of the CW measurement. Over the range from 1 to 300 MHz, I notice that the change in output power I saw (about 9 dB) is consistent with the change in the noise floor for S21 measurements when the ports of the VNA are terminated in loads. A copy of S21 with both ports terminated on my nanoVNA using nanoVNASaver 0.1.0 set for 5 sweeps from 50 kHz to 900 MHz and 20 averages is also attached, "S21 NoiseFloor.png." Clearly, we cannot be very precise with this type of noise floor data but I think it is consistent with the reduction in power out of CH0 I observed from 50 kHz to 300 MHz. So, I really do think there is about a 9-dB variation in the output power on CH0 from 1 to 300 MHz. At frequencies above 300 MHz the noise floor level continues to increase so the increase in power from CH0 must compensate for the use of the third harmonic pretty well in my VNA. I do see a substantial change in the noise floor below 600 MHz. I had thought this should happen at 600 MHz not below 600 MHz. -- Bryan, WA5VAH   |