¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. Nanovna-Users
  2. Topics

Calibration issue causing ripples in SWR? #calibration

 

After calibrating, I connected my nanovna to my VHF antenna via a brand new 75 foot length of 50ohm DX400MAX coax. The SWR measurement oscillates up and down every 4MHz or so (see attached). These correspond exactly to the half-wavelength harmonics of the cable. Looking at the smith chart, it is clear that the impedance is tracing a circle around about 45 ohms rather than the constant-SWR circle centered at 50 ohms, resulting in a frequency-dependent SWR.

Is there is a good way to determine whether this is a problem with my nanovna, with the calibration standard, or with the cable?

One note: the calibration standard load measures 51 ohms DC resistance, which seems like a problem to me. Should my standard be exactly 50 ohms dc resistance?

IMG_20200518_133238866_2.jpg

 

The 50 ohms cables as far as I remember were rated by different
manufacturers as 50 to 52 ohms years ago, so my guess is that 51 ohms
is on the ballpark. It is also a standard resistor value.

Try to find out the cable impedance, it seems to me to be the culprit.
Find the frequency where the cable is lambda/4 as a RF short circuit
with the end of the cable physically open. Set half the frequency
found before on the VNA and read the reactance, that is the cable?s
impedance.

I am writing by heart, a mistake might have slipped...

Jos¨¦, CO2JA

On 5/21/20, ed@... <ed@...> wrote:
After calibrating, I connected my nanovna to my VHF antenna via a brand new
75 foot length of 50ohm DX400MAX coax. The SWR measurement oscillates up and
down every 4MHz or so (see attached). These correspond exactly to the
half-wavelength harmonics of the cable. Looking at the smith chart, it is
clear that the impedance is tracing a circle around about 45 ohms rather
than the constant-SWR circle centered at 50 ohms, resulting in a
frequency-dependent SWR.

Is there is a good way to determine whether this is a problem with my
nanovna, with the calibration standard, or with the cable?

One note: the calibration standard load measures 51 ohms DC resistance,
which seems like a problem to me. Should my standard be exactly 50 ohms dc
resistance?
aparent1/kb1gmx
 

Always perform a reset before setup and SOL.

I have a piece of that DX400max 50ft long I do not see that oddity.

That cable would have to be really poor to get that result.

Several possible things, adaptors, connectors, or cable was kinked.
also the load at the other end? If using a intermediate cable check
it and all adaptors.

Generally if called with a 51 ohm load and its used at the far end of the
cable its should not be as pronounced as you show.

Remotely possible is the cable is really off spec or someone substituted
75 ohm material.

Allison
-----------------
No direct email, it goes to bit bucket due address harvesting in groups.IO
 

Hi Ed
It is quite normal what you see. Please note that the number of point are only 100 so for the span used 50MHz you have a point for each 0.5MHz. Reduce the span to e.g. 140 to 150 MHz and see for every 100KHz the impedance / SWR. The load are OK it is not what you concern should be, It do not think you Ohm meter are accurate enough to determine the very exact DC resistance and will only create a very little SWR contribution.
I would guess you antenna does not have 50ohm impedance on 145MHz. The oscillation you see are natural for reflection from the antenna which is wandering forth and back in the cable.
Kind regards
Kurt

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af ed@...
Sendt: 21. maj 2020 07:35
Til: [email protected]
Emne: [nanovna-users] Calibration issue causing ripples in SWR? #calibration

After calibrating, I connected my nanovna to my VHF antenna via a brand new 75 foot length of 50ohm DX400MAX coax. The SWR measurement oscillates up and down every 4MHz or so (see attached). These correspond exactly to the half-wavelength harmonics of the cable. Looking at the smith chart, it is clear that the impedance is tracing a circle around about 45 ohms rather than the constant-SWR circle centered at 50 ohms, resulting in a frequency-dependent SWR.

Is there is a good way to determine whether this is a problem with my nanovna, with the calibration standard, or with the cable?

One note: the calibration standard load measures 51 ohms DC resistance, which seems like a problem to me. Should my standard be exactly 50 ohms dc resistance?
 

As previously stated, you're fine. What you are seeing is quite expected.

Take your cal standard of 51 ohms: Assuming your cal standard is truly is
51.0000000 ohms (DMM measurement tolerance), the SWR against
50.000000...... ohms would be 51.0000000 / 50.000000, or 1.020:1. I doubt
any amateur SWR meter would display this as anything other than 1:1.
Again, not a problem.

Take your coax: Try making an identical measurement with the antenna end
of the line terminated with your 50-ohm cal. load or another known good
resistor (connect with no leads and the resistor must be good at 146 MHz -
no series reactance) in place of the antenna. The ripple is likely due to
your antenna not being 50.000..... ohms over the entire bandwidth you are
measuring. Again, this is quite typical and not a problem.

Dave - W?LEV

On Thu, May 21, 2020 at 3:15 PM Kurt Poulsen <kurt@...> wrote:

Hi Ed
It is quite normal what you see. Please note that the number of point are
only 100 so for the span used 50MHz you have a point for each 0.5MHz.
Reduce the span to e.g. 140 to 150 MHz and see for every 100KHz the
impedance / SWR. The load are OK it is not what you concern should be, It
do not think you Ohm meter are accurate enough to determine the very exact
DC resistance and will only create a very little SWR contribution.
I would guess you antenna does not have 50ohm impedance on 145MHz. The
oscillation you see are natural for reflection from the antenna which is
wandering forth and back in the cable.
Kind regards
Kurt

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af
ed@...
Sendt: 21. maj 2020 07:35
Til: [email protected]
Emne: [nanovna-users] Calibration issue causing ripples in SWR?
#calibration

After calibrating, I connected my nanovna to my VHF antenna via a brand
new 75 foot length of 50ohm DX400MAX coax. The SWR measurement oscillates
up and down every 4MHz or so (see attached). These correspond exactly to
the half-wavelength harmonics of the cable. Looking at the smith chart, it
is clear that the impedance is tracing a circle around about 45 ohms rather
than the constant-SWR circle centered at 50 ohms, resulting in a
frequency-dependent SWR.

Is there is a good way to determine whether this is a problem with my
nanovna, with the calibration standard, or with the cable?

One note: the calibration standard load measures 51 ohms DC resistance,
which seems like a problem to me. Should my standard be exactly 50 ohms dc
resistance?







--
*Dave - W?LEV*
*Just Let Darwin Work*
 

Another opinion is offered here, along with a bit of mocking of you "online experts" - his words, not mine:



--
John AE5X
 

I did not see it as mocking, but rather a brilliant and feasible
explanation of what might be the cause of the VSWR 'ripple'.

On Sat, 23 May 2020, 04:21 John AE5X, <ae5x@...> wrote:

Another opinion is offered here, along with a bit of mocking of you
"online experts" - his words, not mine:



--
John AE5X
 

Thank you all so much for your help so far!

I've taken the cable down to analyze it. I removed all adaptors and intermediate cables as Allison suggested. Using the lambda/8 method suggested by Jose, I get an impedance of 54.1 ohms. I have also found the open circuit impedance at lambda/4 and lambda/2, which should average out to the characteristic, and gives me 53.2 ohms, so it looks like the cable impedance is somewhere around 53.6 ohms. I'm not sure if this is within usual tolerance, but it seems questionable for such expensive cable.

My antenna is definitely not matched. (I got the vna to design a matching network). As the blog post shared by John points out, properly matched coax should not cause the ripples that I see. It should cause a frequency-dependent phase shift in the impedance, but not affect the magnitude/SWR. So it seems the ripples are likely caused by the mismatch between the cable and vna, and greatly increased in magnitude by the mismatch between the cable and antenna.

So it looks like these SWR ripples can be explained without any problem with the calibration or vna. However, if anyone has suggestions for verifying the output impedance of my nanovna, I'd still like to do it!

Thanks again!
Ed
KC1DKY
 

Ed,

*1)* You've got a typo in your callsign: it should be KC1D*YK* (ROKBL!)

*2)* In terms of verifying the fundamental measuring impedance of the nanoVNA, that's a lot trickier. The essence of most VNAs is that the VNA is approximately correct, but to take an accurate measurement the VNA is compared to a set of reference impedances (the SOL calibration). [The nanoVNA S_11 bridge can be seen in the circuit diagram (schematic) for whichever model of nanoVNA you are using.]

*3)* Given that you are an M.Math, but I a mere subverter of otherwise pristine mathematics towards practical application (also known as Engineer), then you can probably imagine the system of matrices which make this more-or-less work.

*4)* On a very practical note, the cost of a good calibration set can, for good reason, easily exceed the cost of the VNA itself, since nanoVNAs are so very cheap, but accuracy is demanding.

*5)* I admire your experimenter's attitude of "Let's just measure it", which the nanoVNA readily allows one to do. The discrepancy that you've measured is one which, a few years ago, would simply have passed unnoticed, without access either to very expensive test gear, or to cheaper test gear that requires a slow manual process that few would undertake as a matter of course!

*6)* There is also the question of the impedance of CH1, for S_21 measurements, but that's a rabbit hole for another day.

*7)* In short, the measurements that one gets from the nanoVNA are only as good as the calibration references in use. Or if one prefers, the numerical results from a VNA directly come from the calibration set used for those measurements. The numbers *are* the calibration. [That's the essence of it.]

HTH, 73, Stay Safe,

Robin, G8DQX

On 23/05/2020 04:39, ed@... wrote:
So it looks like these SWR ripples can be explained without any problem with the calibration or vna. However, if anyone has suggestions for verifying the output impedance of my nanovna, I'd still like to do it!

Thanks again!
Ed
KC1DKY

1 - 9 of 9