Anyone tried to connect their NanoVNA directly to their TinySA... Very interesting. See pics below. The VNA is set to sweep from 125MHz to 175MHz. The SA is wide open from 0 to 800MHz. I was surprised to see signals all over the band though most are concentrated in the sweep limits.

Notice LOGMAG is flat and the Smith chart shows no change across any range..

NOTE: There is 30dB external attenuation between these two... Just to be safe.

Pic 1 AVG 4
Pic 2 MAX HOLD
Pic 3 random shot no calc on trace

Remember that the NanoVNA output is a square wave - it uses internal filters to ignore the undesired harmonics.
The TinySA is showing you the fundamental frequency of the main sweep, and the harmonics - which naturally spread out as the frequency increases.
All is well. And the flat S11 LogMag says that the input of the TinySA is very near a constant 50 ohms.

Ah yes, I forgot about harmonics, and yes a rock solid 50 ohms and no reactance across the range. I have another picture with the VNA sweeping from 1 to 799 MHz. The chart stays rock solid.

The "rock solid 50 ohms" is probably due to the 30dB attenuator.
--John Gord

Maybe... The NanoVNA's signal is -13dBm so below the SA's max tolerance (+6dBm). But the SA is a very sensitive device. The extra -30dB cuts a ~0.00006 Watt signal to ~0.00000006 Watts. In those pics, the VA's LNA is on. You can see how clean the signals look...

One more... Stan said "harmonics" so I put the NanoVNA in "CW mode" at 150MHz. On the SA the signal on the left is the principle, and you can clearly see harmonics at 300, 450, 600, etc, right where they should be.

But interesting that the odd numbered harmonics (1 & 3) are half the strength of the even numbered 2 & 4!!!

Matt,

You’re counting the harmonics improperly. The first harmonic is the same as
the fundamental so your odd harmonics are indeed much greater than the
evens as they should be for a decent square wave.

Tony

On Mon, Oct 14, 2024 at 10:19 PM Matthew Rapaport via groups.io <quineatal=
gmail.com@groups.io> wrote:

One more... Stan said "harmonics" so I put the NanoVNA in "CW mode" at
150MHz. On the SA the signal on the left is the principle, and you can
clearly see harmonics at 300, 450, 600, etc, right where they should be.

But interesting that the odd numbered harmonics (1 & 3) are half the
strength of the even numbered 2 & 4!!!

Hi Tony... I never understood that counting convention, but so be it---I realize any frequency will be a harmonic of some other frequency, so perhaps that is the reason for it.

Interesting that square wave harmonics come out that way. I did not know that! Nice to see a demonstration

Now for another instructive experiment: set your nanovna to 350 or 400MHz cw and make the same measurements, and describe why you see what you see.

Hi Stan... I'll try though until now I've not opened the tinySA beyond 800MHz. I should also try the MEASURE>HARMONIC mode instead of looking at the open sweep

My students had to have it repeatedly beaten into their heads:

THe 2nd harmonic is 2X the fundamental (1st harmonic) and the 3rd harmonic
is 3x the fundamental.
And that is also the way that all textbooks treat it, It is fundamental to
the math when expanding equations of Fourier series.

On Tue, Oct 15, 2024 at 10:59 AM Matthew Rapaport via groups.io <quineatal=
gmail.com@groups.io> wrote:

Hi Tony... I never understood that counting convention, but so be it---I
realize any frequency will be a harmonic of some other frequency, so
perhaps that is the reason for it.

Interesting that square wave harmonics come out that way. I did not know
that! Nice to see a demonstration

No need to go above 800MHz on the tinysa to make the interesting observation. Just do the exact same thing.

Mathematicians count from 1 because it comes from an exponent in the standard taylor series expansion. The zeroth power is 1, or DC term.

C language programmers start from zero, because of the original convention that arrays (y= a[i]) were pointer+index (y=*(p+i)).

What's interesting is to take a pulse generator and change the pulse width and rise/fall times.

The duty cycle affects the harmonic distribution, rise and fall time affects how quickly it falls off.

Even better if you do it at audio frequencies, and hook it up to speaker and watch it.

Triangle sounds different than sawtooth sounds different than squarewave (all odd harmonics)

There are oscillators (defined as an energy input and resonant system) that have the harmonics stronger than the fundamental (violins and some wind instruments).
There's also a thing where the position of the energy input to a mechanical system (say, where a string is plucked or hit) that tends to enhance/suppress harmonics.

The 7th harmonic is dissonant, so pianos strike the string at 1/7th of the distance from the end, which tends to suppress it.

But we stray substantially from VNAs here.

Ok Stan, you got me! See picture. The marker is on 400MHz. I am seeing signals both above and below the fundamental which is not even the strongest signal. The other signals go in 1/2 harmonic steps (?) with smaller spikes to either side in 1/4 steps... (?)...

I notice a measured level of -10 dBm. I believe Eric recommends for best
performance and minimal aliasing to keep the input level at or below
roughly -30 dBm. You might try the same sweep with a 20 dB attenuator
installed between the source and the input to the VNA.

Dave - WØLEV

On Tue, Oct 15, 2024 at 8:37 PM Matthew Rapaport via groups.io <quineatal=
gmail.com@groups.io> wrote:

Ok Stan, you got me! See picture. The marker is on 400MHz. I am seeing
signals both above and below the fundamental which is not even the
strongest signal. The other signals go in 1/2 harmonic steps (?) with
smaller spikes to either side in 1/4 steps... (?)...

--

*Dave - WØLEV*


--
Dave - WØLEV

Fun, right? As Dave mentioned, you will get a more accurate reading if the level lower, but it will be fundamentally similar.
This is part of the clever design of the nanovna, analogous to the clever design used in the tinySA.
The frequency generator chip used in the nanovna can only give a good signal up to about 300MHz, so for all frequencies above 300MHz it uses a harmonic. So what you are seeing is the 133.3 MHz fundamental, and the third harmonic of 400MHz, and other harmonics, same as what you would see if you set the nanovna to 133.3MHz. This clever use of the square wave and its harmonics greatly extends the usefulness of this little device.
Stan

Thanks Stan for that explanation. Very interesting...

To you and Dave, I have had the same 30dB external attenuation between the SA and VNA in all these pictures. What varies is that I've told the SA there is -30dB of external attenuation (LEVEL>EXT GAIN) so the -10 dBm you see is really -40 dBm. I am a BIG fan of attenuation!