开云体育

I just acquired a NanoVNA-H4. My main use will be as an amateur radio antenna analyzer. S11 measurements for HF through 450 MHz, particularly 2m and 70cm, where I have no tools other than an analog SWR meter. I have seen articles and videos where people calibrate the NanoVNA once from 50 kHz to 900 MHz, and leave it at that. I have also seen articles where people claim that you must re-calibrate every single time you change the frequency range. This seems like overkill for my uses. I see very little difference in the SWR, reflection loss and Smith chart readings between the C0 factory calibration and my own 2 meter band calibration. The HF readings I've done so far agree with my RigExpert HF-only analyzer.

My current thinking is that I should be fine with a calibration for 3-30 MHz and another for about 100-500 MHz, or maybe split the latter into two calibrations for 2m and 70cm. But maybe you folks know better. Please enlighten me. Again, I'm talking about real-world ham radio antenna tweaking, not theoretical physics.

Thanks and 73,
--Peter

This, it seems to me, is a question of how closely you need to examine the results. One of the limitations of the nanoVNA has been the restriction to 100 data points between the start and stop points. If a 100th part of the spectrum between the start and stop points gives you enough precision for your purposes. then...fine! If a 100th part is too big...if you need more precision...recalibrate for a smaller span.

Thanks Gordon. (Hey, we're practically neighbors!). I'm hoping that the NanoVNA interpolates between calibration points if you "zoom in" in frequency. For example, I calibrated for 130-160 MHz to check my 2m band antennas. Then, if I change the start/stop frequencies to, say 142 and 150 MHz to get a closer look at the 2m antenna , do I really need to re-calibrate? If the NanoVNA just interpolates smoothly between the wider calibration values and comes up with intermediate values that are good enough to zero out the pigtail and connectors, then all's well. On the other hand, if the NanoVNA can't deal with the differences between the calibration data points and the more narrowly-spaced data points when I measure the antenna within a narrower frequency range, then, yes, I'd need to re-calibrate.

My guess is that for practical, ham radio antenna purposes, I'm fine. But since I don't know precisely how the NanoVNA works, I'm asking. Especially since those SMA connectors on the Short-Open-Load calibrators have a finite life.

On Thu, Apr 23, 2020 at 04:55 PM, Peter Klein - KD7MW wrote:

Thanks Gordon. (Hey, we're practically neighbors!). I'm hoping that the
NanoVNA interpolates between calibration points if you "zoom in" in frequency.
For example, I calibrated for 130-160 MHz to check my 2m band antennas. Then,
if I change the start/stop frequencies to, say 142 and 150 MHz to get a closer
look at the 2m antenna , do I really need to re-calibrate? If the NanoVNA just
interpolates smoothly between the wider calibration values and comes up with
intermediate values that are good enough to zero out the pigtail and
connectors, then all's well. On the other hand, if the NanoVNA can't deal
with the differences between the calibration data points and the more
narrowly-spaced data points when I measure the antenna within a narrower
frequency range, then, yes, I'd need to re-calibrate.

My guess is that for practical, ham radio antenna purposes, I'm fine. But
since I don't know precisely how the NanoVNA works, I'm asking. Especially
since those SMA connectors on the Short-Open-Load calibrators have a finite
life.

Yes, NanoVNA works as you think.

hugen

Peter,

You have 5 'saves', so use them.? I have done 3 saves, 1-50Mhz 140-150Mhz 430-450Mhz.? Using too wide a sweep

also is a problem, as the signal generator uses big steps, and can miss the fine problems you might see.

After I did band-specific calibration, it seems to be a whole lot better!

73

Glen K4KV

I did some measurements today on a vertical whip antenna tuned for the top end of the 2M amateur band. I wanted to see the effect of calibrating the NanoVNA on SWR and Return Loss measurements? The conventional wisdom is that the closer you calibrate to the frequency range of interest the more accurate the results. How close is close enough? That depends on how important accuracy is to you in your measurements. The first measurement was done using the factory default 50 to 900 MHz. calibration. The second was done by changing the Stimulus range (zooming) of the NanoVNA to 140 to 155 MHz. while still using the factory calibration. I then re-calibrated from 140 to 155 MHz. and repeated the measurement. Subsequent tests did zooming and calibrating to an even narrower range. You can clearly see the effect of calibration on the results...

Roger

Here is a second report on the effect of calibrating the NanoVNA on SWR and Return Loss measurements. This time a 20M dipole was used as the antenna. It was surprising to me to see how good the measurements were when using the default 50 kHz. to 900 MHz. calibration and setting the Stimulus range to 13 to 15 MHz. Interpolation works well under certain conditions.

Roger

On Mon, Apr 27, 2020 at 02:13 AM, Roger Need wrote:

I did some measurements today on a vertical whip antenna tuned for the top end
of the 2 m amateur band.

I have made a table to better see what is the result of the measurements from the PDF file.
For my understanding here the *precision of the frequency* is important.

Calibration MHz Scan MHz Step KHz SWR RL dB Lag KHz
0.05 - 900 0.05 - 900 9000 1.25 19.04 153.041-147.76= 5281
0.05 - 900 140 - 155 150 1.06 29.6 148.400-147.76= 800
140 - 155 140 - 155 150 1.07 29.13 147.600-147.76= 200
140 - 155 146 - 150 40 1.06 29.61 147.800-147.76= 40
146 - 150 146 - 150 40 1.06 30.34 0

*Result* : The frequency lag of maximum RL can not be (much) better, than the measuring step size.
So, watch the *Step Size* = scan width / 100.

73, Rudi DL5FA

On Mon, Apr 27, 2020 at 02:41 AM, Roger Need wrote:

Here is a second report on the effect of calibrating the NanoVNA on SWR and
Return Loss measurements. This time a 20M dipole was used as the antenna. It
was surprising to me to see how good the measurements were when using the
default 50 kHz. to 900 MHz. calibration and setting the Stimulus range to 13
to 15 MHz. Interpolation works well under certain conditions.

I have made a *table* to better see what is the result of the measurements from the PDF file.

Calibration MHz Scan MHz SWR RL dB
0.05 - 900 1 - 30 1.20 20.56
1 - 30 1 - 30 1.19 21.12
0.05 - 900 13 - 15 1.12 24.42 good result for practical use
1 - 30 13 - 15 1.12 24.74
13 - 15 13 - 15 1.12 24.67
14 - 14.35 14 - 14.35 1.12 24.74

*Result* : A full span calibration is sufficient for practical use.
How smaller the scan width, the better the SWR / Return Loss.

73, Rudi DL5FA