开云体育

On Sun, Nov 8, 2020 at 05:24 PM, David Eckhardt wrote:

I could measure total
current on each conductor, and ultimately 'sum' the results to obtain the
expected zero with no CM current.

Hi Dave,

Great stuff, I'm pleased that you have been motivated to make the measurements, and I'm keen to discover the outcome.

Ideally you would be able to sum the individual amplitudes and phases of the two conductors forming the transmission line, but it is often easier to pass both conductors through the same RF current transformer in order to ensure that you are seeing the true picture.

One of the big Fair-Rite clip-on ferrites is ideal for this purpose.

It has been pointed out to me privately that many of the diagrams that can be found on-line, showing how common mode current flows and how a choke balun works, are actually flawed, as the common mode current is shown as only flowing on one side of the choke, which is impossible. I think this is simply an illustrative error, made in order to try and simplify the diagrams, but it can lead to further confusion.

A choke will impede the overall flow of common mode current (on both sides of the choke) and its reactive component may also lead to phase changes occurring, which will modify the field radiated by the conductor it is placed on. So it's not a totally predictable mix, as there are lots of different interactions going on, especially when the reactance of the choke is likely to change with frequency, and the conductor common mode standing wave pattern will also vary with frequency. Which is why it's a good idea to attempt to quantify the effectiveness of chokes when placed in circuit at the required frequency(s) of operation. More than one choke and different construction may be required in order to achieve satisfactory operation over a wide frequency range.

--
Regards,

Martin - G8JNJ

On Sun, Nov 8, 2020 at 08:23 PM, Stan Dye wrote:

Two simple notes to consider adding:

- The firmware gives reasonable defaults for the scale and reference line
for many use cases, so you only need to set them if you need them to be
different.
- When you save a calibration, it also saves the display settings, so you
can easily recall the whole setup.

(I note these because the list of setup tasks for each measurement seems
long and maybe intimidating if I have to do them for every measurement -
but it's not nearly as bad as it seems.)

Stan KC7XE

Very good idea. I will do that. Thank you Stan.

Also I have corrected some misspelled words - thank you Mike.

Martin 9A2JK

John

You don't state which manufacturer of cable but I know some years ago that Andrew used to number their coax every meter so if you're able to ascertain the right number on the cable [easy] and access the other end, you should be able to confirm your nanoVNA measurement!

Best of luck and I, for one would be interested on how accurate you find the nanoVNA compared to the expensive VNA.

73, Bob
G1ZJP

Thanks for the info, guys. We use heliax to feed a transponder antenna and the indoor end is connected to a special N connector made for the cable. We have a ton of N-to-SMA adapters. We also have a $$$ VNA at work but I also want to test it with the nano for comparison, just didn't know if the cable being coiled would matter.

Thanks again,

--
John AE5X

Yes, at the factory they are typically tested on the spool.
You are looking at internal fields, coiling the coax up makes no difference.
73 Kent WA5VJB

Fantastic job.
I have practicly every manual for the NanoVNA H that I could get my hands on to learn the NanoVNA. Not to degrade the other manuals, but yours seem to be the most useful for a old beginner like me.
Thank you and all others who help us to learn and use the NanoVNA to its fullest.
Now, we need one like yours for the NanoSA.
Again thank you.
73's
Clyde Lambert KC7BJE

On 11/8/20 12:17 PM, John AE5X wrote:

Hello all,
I don't have much experience using TDR's in general and am wondering if they can accurately measure cable length if the cable is coiled rather than deployed in a more typical manner? At my work QTH we have a length of heliax on a big wooden spool and no one knows how much of it there is, nor can we easily count the number of coils on this stuff. It is quite rigid and we'd like to not have to unroll it.
Thanks/73,

that is the perfect use for a TDR.

I am no expert, (I have only used TDR with Nanosaver on short lengths with multiple connections -every single connector showed up with a slight blip) but I see no reason for a TDR measurement not to work. It does not matter that the wire is coiled although there might be some uncertainty at the far end ; hopefully it is open circuit at the far cut end, but the last inch might be crushed. What you hope is that there is not a perfect 50 ohm load at the far end! You need to set the vna to the highest possible upper frequency for short lengths (1mm to a metre) but have the lowest start frequency for long lengths.

The biggest difficulty will be to get a decent connection to the heliax. I do not know how you connect Heliax to sma. I suppose very short wires soldered to the sma connection should be ok for the far end measurement. You might even get a clean enough signal to confirm that there is no damage somewhere on the coil.

Steve L

Hello all,

I don't have much experience using TDR's in general and am wondering if they can accurately measure cable length if the cable is coiled rather than deployed in a more typical manner? At my work QTH we have a length of heliax on a big wooden spool and no one knows how much of it there is, nor can we easily count the number of coils on this stuff. It is quite rigid and we'd like to not have to unroll it.

Thanks/73,
--
John AE5X

Martin can you send me this document
Tnx

--
Sent from my Android phone with GMX Mail. Please excuse my brevity.

The exact same case. While calibrating only "Y" axis is working. The horizontal "X" axis was always stuck at the very screen edge. I followed Bob's solution - re-solder the flex cable and now works like a charm. 73 de SV1ONK

I too send you my thanks for an excellent tutorial document. Very nicely
done.
Two simple notes to consider adding:

- The firmware gives reasonable defaults for the scale and reference line
for many use cases, so you only need to set them if you need them to be
different.
- When you save a calibration, it also saves the display settings, so you
can easily recall the whole setup.

(I note these because the list of setup tasks for each measurement seems
long and maybe intimidating if I have to do them for every measurement -
but it's not nearly as bad as it seems.)

Stan KC7XE

Martin, you peak my interest / curiosity in making even more measurements
on my system. In choosing what I have done with my 450-foot (137-meter)
doublet is based on two things:
1) Results as measured on the HP 8753C VNA for each choke, and
2) Suppression of CM noise from the (new) home appliances with all their
embedded SMPS's.

You motivate me to make a few more measurements with everything in place
with current probes (which I will have to make). I once had a piece of
test equipment I made with a current probe on each of the two conductors of
the open wire feeders as they left the shack. I could measure total
current on each conductor, and ultimately 'sum' the results to obtain the
expected zero with no CM current. Anything indicated with the 'sum' would
be CM current. I need to make another just like that as the original was
lost in a fire in 2012. Good suggestion! I shall also measure and present
the inductive reactance of my chokes. I have actually done this, but not
recorded values (bad engineer as I don't keep a rigorous 'engineering
notebook').

Dave - W?LEV

On Sun, Nov 8, 2020 at 2:36 PM Martin via groups.io <martin_ehrenfried=
[email protected]> wrote:

On Sat, Nov 7, 2020 at 08:27 PM, David Eckhardt wrote:

Just one I could dig up quickly. Most explanations of a balun simply

state

they are a 'transformer' (which they may or may not be) that bilaterally
transforms CM to DM. They do not dig into how they accomplish that
function.

Hi Dave,

Thanks for the links, unfortunately they only cover the cancellation of
Common Mode currents, and do not address the issue of reactive impedance,
and if it is desirable in a Common Mode Choke Balun.

I think the crux of the discussion is if it is better to have a common
mode choke with a predominantly resistive impedance to act as an absorptive
load for CM signals, or as a predominantly reactive load that is reflective
to CM signals.

Ideally it would be better to have a high enough value of resistive CM
impedance so that CM signals are blocked, are not dissipating significant
amounts of energy as waste heat, and are not creating unpredictable CM
standing waves elsewhere.

However you don't seem to be able to achieve this, as you are running high
power and seem to have a lot of CM to deal with, so you are resorting to
using type 2 'Iron Powder' because the CM impedance is primarily reactive,
and the material is no longer over-dissipating.

But this isn't dealing with what happens to the highly reactive CM
component that is reflected back from the balun. Simply because you are not
observing over-dissipation in the Balun, or RF in the shack, it doesn't
mean that the balun is actually working effectively, it maybe just moving
the problem, and the common mode RF could still be dissipated as heat
somewhere else in the antenna system.

I wonder if you have been able to make any comparative measurements with a
RF current probe along the feed line, in order to determine the
effectiveness of the Balun ?

Regards,

Martin - G8JNJ

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

Indeed Martin, that's why I use a 31-core and not a 2-core for this frequency.

Enclosed is a a graph, measured by nanoVNA H3.2 for a 240-31 core and 15 windings of rg58 coax. (It's a CMC in my 3.5 - 7 - 10 MHz trap-dipole.)
Resonance is visible at 5 MHz but then there is the resistive part of the impedance that saves the day. ;-)

For the record, I did the same measurement with 15? turns of RG174 (the thin coax). Due to less capacitance between windings the resonance point goes up in frequency to just aroud 6 MHz.


73

Arie PA3A


Op 8-11-2020 om 15:46 schreef Martin via groups.io:

On Sun, Nov 8, 2020 at 08:36 AM, Arie Kleingeld PA3A wrote:

Let's suppose you want a clean 500 ohm inductive impedance. A bit more than 31
windings on a T200-2 core will do that on 7MHz

Hi Arie,

I quickly wound some wire on a T200-2 'Iron powder' core.

Yes it does indeed produce an inductive reactance of around 500 Ohms at 7MHz, but it also has a very low value of resistive component, a self resonance at around 10MHz and has capacitive reactance above the self resonant frequency.

The bandwidth around resonance where the value of reactance exceeds 1,000 Ohms is only 6MHz wide, so it's only got a small frequency range over which it is likely to be effective. The range over which the resistive part exceeds 1,000 Ohms is less than 0.5MHz.
--
Regards,

Martin - G8JNJ

On Sat, Nov 7, 2020 at 08:27 PM, David Eckhardt wrote:

Just one I could dig up quickly. Most explanations of a balun simply state
they are a 'transformer' (which they may or may not be) that bilaterally
transforms CM to DM. They do not dig into how they accomplish that
function.

Hi Dave,

Thanks for the links, unfortunately they only cover the cancellation of Common Mode currents, and do not address the issue of reactive impedance, and if it is desirable in a Common Mode Choke Balun.

I think the crux of the discussion is if it is better to have a common mode choke with a predominantly resistive impedance to act as an absorptive load for CM signals, or as a predominantly reactive load that is reflective to CM signals.

Ideally it would be better to have a high enough value of resistive CM impedance so that CM signals are blocked, are not dissipating significant amounts of energy as waste heat, and are not creating unpredictable CM standing waves elsewhere.

However you don't seem to be able to achieve this, as you are running high power and seem to have a lot of CM to deal with, so you are resorting to using type 2 'Iron Powder' because the CM impedance is primarily reactive, and the material is no longer over-dissipating.

But this isn't dealing with what happens to the highly reactive CM component that is reflected back from the balun. Simply because you are not observing over-dissipation in the Balun, or RF in the shack, it doesn't mean that the balun is actually working effectively, it maybe just moving the problem, and the common mode RF could still be dissipated as heat somewhere else in the antenna system.

I wonder if you have been able to make any comparative measurements with a RF current probe along the feed line, in order to determine the effectiveness of the Balun ?

Regards,

Martin - G8JNJ

Martin your a Rockstar!

Now we need one for the TinySA
John VE3IPS

Roger I was very fortunate in that my second job out of college (I got a BS in electrical engineering) was with HP. Back when HP was the instrumentation company. In those days (1982) I doubt anyone would disagree that HP was THE premier designer of RF test equipment. VERY active in everything RF.

As part of the training for sales people, so we could intelligently talk with our customers, they spent 9 weeks, in 3, 3 week segments, teaching us the basics of rf and of their instruments.

While much of what I learned has been relegated to the dust bin in my brain, a few things remain. One was that return loss was always positive. If you do the math this will be born out. If you just think about it, its a ratio of incident vs reflected rf (or maybe the other way around. I just don't recall). A ratio of two rf power levels should always be positive. But again, do the math.

Sure many many folks make it negative. Many, in their explanation, first say one thing then later in their article, the other. Of course confusing! And some conflate return loss with insertion loss. But, insertion loss will also always be positive so I've not understood that one.

Anyway, as you said, its an argument that's been around forever. And, one side will always be wrong :)

Ron
N4XD

Yes very well done Martin !

People always need help and info on how best to use the VNA's, especially when they are new to the VNA area, which we all are (hobby/pro's alike) to start with.

Thank you all for kind words and advice in this group and to email.
On page 8, in the text below the figure 13, should have been written
''on a scale of 20 ohms per VERTICAL division'' - thank you Ed.

The word ''CHANEL'' in figure 13 should be ''CHANNEL'' - thank you Bob.

Larry, thank you for your advice. I will try to expand the next version of the Guide.

73
Martin 9A2JK