¿ªÔÆÌåÓý

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

Through calibration #calibration

 

I have two short cables with alligator clips that I use for looking at baluns, transformers wound on toroids, tuned circuits and what not.

When calibrating should i just tune the the s11 side or should i connect the two together and calibrate to the end of both of them?
Thanks
Bryan
 

I perform a calibration with the cables in place.

I have a S11 cable with two alligator clips, input and ground and a S21 cable with one input alligator clip

The S11 clips are separated for the open cal and connected together for the short cal. I install 50 ohm load (two paralleled 100 ohms) for the load. cal.

I connect the S11 input to the S21 input for the thru cal.

Mike N2MS

On 03/11/2023 11:03 AM Bryan Curl <bc3910@...> wrote:


I have two short cables with alligator clips that I use for looking at baluns, transformers wound on toroids, tuned circuits and what not.

When calibrating should i just tune the the s11 side or should i connect the two together and calibrate to the end of both of them?
Thanks
Bryan
 

Hi Mike. Thanks for the reply.
Just to be clear, you calibrate at the end of the S11 cable (with both cables in place). Then you connect them together for the through part of calibration.

Regards.
Bryan
 

Bryan,

Yes. I use it to check the characteristics of my ferrites. In addition, I have different sizes of Fair-Rite 31 43 and 61 cores so I can compare the unknown cores to the Fair-Rite products.

Mike N2MS

On 03/11/2023 12:59 PM Bryan Curl <bc3910@...> wrote:


Hi Mike. Thanks for the reply.
Just to be clear, you calibrate at the end of the S11 cable (with both cables in place). Then you connect them together for the through part of calibration.

Regards.
Bryan
 

For the S11 calibration (Open, Short, Load), it does not matter whether
there is a cable attached to the second port or not.
For the Isolation and Thru calibration, the cables should be attached to
both ports. Ideally, for the isolation calibration, you individually
terminate both cables with a 50ohm load (at the clip end of the cable).
Then for the thru calibration, simply connect the ends of the cables
together. You must do the isolation and thru calibrations if you want
accurate results on any thru (S12) measurements, but it is not relevant for
S11 measurements.
Stan

On Sat, Mar 11, 2023 at 10:05?AM N2MS <mstangelo@...> wrote:

Bryan,

Yes. I use it to check the characteristics of my ferrites. In addition, I
have different sizes of Fair-Rite 31 43 and 61 cores so I can compare the
unknown cores to the Fair-Rite products.

Mike N2MS

On 03/11/2023 12:59 PM Bryan Curl <bc3910@...> wrote:


Hi Mike. Thanks for the reply.
Just to be clear, you calibrate at the end of the S11 cable (with both
cables in place). Then you connect them together for the through part of
calibration.

Regards.
Bryan
 

Correct. For thru the cables are connected together. For isolation I connect a 50 ohm load is to each cable.

Mike N2MS

On 03/11/2023 1:10 PM Stan Dye <standye@...> wrote:


For the S11 calibration (Open, Short, Load), it does not matter whether
there is a cable attached to the second port or not.
For the Isolation and Thru calibration, the cables should be attached to
both ports. Ideally, for the isolation calibration, you individually
terminate both cables with a 50ohm load (at the clip end of the cable).
Then for the thru calibration, simply connect the ends of the cables
together. You must do the isolation and thru calibrations if you want
accurate results on any thru (S12) measurements, but it is not relevant for
S11 measurements.
Stan
 

Thanks all. I just wasnt sure, for a thru cal where to place the Open Short and Load bits. Its pretty straight forward for using just S11 for reflected measurements.

Mike that's a great idea for identifying ferrites! I tore some out of an old PC power supply that were green, blue and black but have no idea what they are. Colors may or may not be an indication.

Thanks again everyone.
Bryan
 

Bryan,

Those power supply ferrites are useful for low frequency power or choke applications but are not that useful for HF balun, unun or transformer applications. The unknown ferrites I find useful are those split ferrites placed on cables for RFI attenuation. Most are designed for VHF attenuation and are similar to the Fair-Rite 43 mix.

There is no standard color code. The colors are probably unique to the manufacturer.

Mike N2MS

On 03/11/2023 4:25 PM Bryan Curl <bc3910@...> wrote:


Thanks all. I just wasnt sure, for a thru cal where to place the Open Short and Load bits. Its pretty straight forward for using just S11 for reflected measurements.

Mike that's a great idea for identifying ferrites! I tore some out of an old PC power supply that were green, blue and black but have no idea what they are. Colors may or may not be an indication.

Thanks again everyone.
Bryan
 

Using short coaxial cables with alligator clips on the end for S11 and S21 can easily lead to erroneous results. At best it is only useful for measurements in the HF range. There are several reasons for this...

- In order to get decent results you need to establish a good reference plane. In order to get a good reference plane the distance used between the measurement points must be the same for the SOL calibration. The measurement of the DUT must also be made at the reference plane. Using alligator clips that are at the end of leads connected to a coaxial cable requires at a minimum that everything be held in place for SOL and measurements. A different distance between the clips for the short, open and load and DUT measurement will result in errors in the measurement. How much error depends on what kind of measurement is being performed.

- Users have to be careful what they use for the 50 ohm calibration load. Leaded components have inductance in the wire leads and in the body of the resistor if carbon or metal film resistors are used. Never use a 50 ohm dummy load because they have considerable inductance.

- The wires coming out of the 50 ohm coaxial cable to the alligator clips form a higher impedance transmission line that increases the further the leads are separated. At VHF and above they act like a dipole antenna and you get common mode current on the outer surface of the shield. Try to "calibrate out" this arrangement can lead to errors.

- For S21 measurements stray capacitance and inductance will affect the measurement (especially the S21 phase). To see what I mean try connecting a 100 ohm resistor between the two clips and see of you get 6 dB loss with minimal phase shift.

Measuring components like resistors, inductors, ferrites, capacitors and baluns requires a test jig that is up to the task if you want results you can believe. If you want to find resonant frequencies of components like traps, inductors or capacitors you need a proper test setup. There have been many posts by knowledgeable users in this group on how to construct low cost test jigs that give accurate results up to 900 MHz. or so. This group contains a treasure trove of information on this subject in the wiki and the message archive.

Roger
 

Powdered iron cores are all color coded. In general, ferrite material is
not color coded, although some suppliers may.

Dave - W?LEV

On Sat, Mar 11, 2023 at 11:24?PM N2MS <mstangelo@...> wrote:

Bryan,

Those power supply ferrites are useful for low frequency power or choke
applications but are not that useful for HF balun, unun or transformer
applications. The unknown ferrites I find useful are those split ferrites
placed on cables for RFI attenuation. Most are designed for VHF attenuation
and are similar to the Fair-Rite 43 mix.

There is no standard color code. The colors are probably unique to the
manufacturer.

Mike N2MS


On 03/11/2023 4:25 PM Bryan Curl <bc3910@...> wrote:


Thanks all. I just wasnt sure, for a thru cal where to place the Open
Short and Load bits. Its pretty straight forward for using just S11 for
reflected measurements.

Mike that's a great idea for identifying ferrites! I tore some out of an
old PC power supply that were green, blue and black but have no idea what
they are. Colors may or may not be an indication.

Thanks again everyone.
Bryan




--
*Dave - W?LEV*
--
Dave - W?LEV
 

Those ferrites in power supplies are not iron core. The only place I've seen iron core chokes are in HF communications equipment.

Mike n2MS

On 03/11/2023 6:27 PM W0LEV <davearea51a@...> wrote:


Powdered iron cores are all color coded. In general, ferrite material is
not color coded, although some suppliers may.

Dave - W?LEV

On Sat, Mar 11, 2023 at 11:24?PM N2MS <mstangelo@...> wrote:

Bryan,

Those power supply ferrites are useful for low frequency power or choke
applications but are not that useful for HF balun, unun or transformer
applications. The unknown ferrites I find useful are those split ferrites
placed on cables for RFI attenuation. Most are designed for VHF attenuation
and are similar to the Fair-Rite 43 mix.

There is no standard color code. The colors are probably unique to the
manufacturer.

Mike N2MS
 

On Sat, Mar 11, 2023 at 06:26 PM, Roger Need wrote:


Measuring components like resistors, inductors, ferrites, capacitors and
baluns requires a test jig that is up to the task if you want results you can
believe.
The bottom line for most VNA calibrations is that improvements in "raw" performance, that is, the smaller the errors you are calibrating out, the better and more stable the measurements using those calibrations will be. Minimizing stray inductance(looking at you, clip leads!), capacitance and port-to-port coupling (looking at you, clip leads!) and eliminating random errors by making consistent connections(torque wrench, keeping items from moving between cal and measure, etc.) will pay off in better-quality measurements.
You can (often) get away with a lot of slop at HF, but going higher in frequency you need to be increasingly careful and meticulous.
73, Don N2VGU
 

For testing non-insertable devices (devices that don't have the appropriate connectors to maintain a Zo of 50 ohms) I have used microstrip transmission lines for calibration and testing. The Zo of the transmission lines are 50 ohms (G10 / FR4 double sided pcb boards etched to have the correct dimensions to establish Zo of 50 ohms). As Roger mentions you need to establish a proper measurement plane so this has been the best method that I have used with good results up to 500 MHz.
I use SMA pcb connectors designed for the thickness of the copper-clad board to get a pretty good transition from coax to the microstrip lines.
One pcb assy is basically a length of 50 ohm microstrip line with SMA's at each end. This establishes the "thru" for the calibration (thru pcb assy).
The next pcb is exactly one half as long as the thru pcb with and SMA at one end and at the far end it is open (no connector or component to establish and "open" for the calibration.
The next pcb is like the open but has a short across the far end (same exact physical distance to the open).
The next pcb is as above but with two 100 ohm SMT resistors where one is connected to one of the ground planes and the other connected to the opposite ground plane to make low reactance. The two 100 ohm smt resistors create the 50 ohm calibration load.
The next pcb is like the "open" cal assy where the DUT is soldered at the far end to be tested.
Other assys can be the exact same construction as the "thru" but have a break in the exact center to insert a DUT in series.
Lastly another pcb assy made exactly like the "thru" but for testing a device in parallel (shunted across at the center point).
This may all seem crazy but I depended on testing like this for my job before I retired. It was very interesting to come across strangely-behaving components especially with inductors. Using clip leads would have totally masked the behavior and added even more strangeness.
 

I made a jig so I can test Tesla coil components with. The first couple permutations of the design used alligator clips. I hade to fashion cal standards for SolT ISO, but in the end I believe calibrating all of the inherent resistance capacitance and inductance out of everything except the device or components under test.

For the above I¡¯m using 2.8¡± saa and a 4¡±v2 both with the trick FW. And calibrating from 10khz-3ghz.

They are remarkably close in measurement and consistency.


Ultimately, What I have found is using sma edge connectors, soldered directly to the components you want to test yields a more accurate result that is easily replicated.

Take that as you may. There are a ton people on this groups.io with far more experience and knowledge than I can ever have. If this didn¡¯t answer satisfactorily, keep at it, they¡¯ll get you the answer you need.

Just verify, verify, verify.

That being said¡­

Disclaimer::

Since I¡¯m still very new to all of this and as an RF/microwave hobbyist, having just taken it up in the COVID years, I¡¯ll be the first to tell you I don¡¯t know shit.

I¡¯m blown away every day by something I come across, a paper I read or one of these videos on YouTube.

Though I have nearly 25 years of professional experience as a 15A Tech(chassis/Body electrical)/Drivability Tech (emissions, diagnostics) in GM dealerships, Private Auto Shops and at least a few months of each of those 25 years either setting up someone¡¯s bike for Willow or Seca , or hauling my ass around one the tracks. So I use to think I knew something¡­
I used to think I knew the nature of electricity, how to fix issues , etc.
As I learned, all too fast after taking up Antenna design and RF electronics , DC automotive electronics are the kiddie pool of the Field of Electronics and RF.

AC electricity, circuits, components, etc , a. nauseum are just this side of magic and it blows me away more and more the further into the topic I get. To end the disclaimer¡­. If I give the wrong info, I apologize. please correct me, that¡¯s how we learn.

Good Luck,


JasonBurchell
@BurchellPerformanceTuning.Com
2445 Balmoral Ct
Camarillo, Ca. 93010
1-805-766-5595
JasonBurchell76@...

On Mar 11, 2023, at 9:59 AM, Bryan Curl <bc3910@...> wrote:
?Hi Mike. Thanks for the reply.
Just to be clear, you calibrate at the end of the S11 cable (with both cables in place). Then you connect them together for the through part of calibration.

Regards.
Bryan
 

Great info all. I was aware that using coax with clip leads would add to some errors. I was working HF band below 30mhz so thought I might get away with it. Perhaps not.

Lots to learn about making chokes, transformer baluns and s21 measurements. Ill look for those articles on the test fixtures.

Bryan,n0luf
 

Bryan,

I don't use coax but loose wires connected to a connector. I use this to characterize ferrites up to 60 MHz.

Mike N2MS

On 03/12/2023 11:27 AM Bryan Curl <bc3910@...> wrote:


Great info all. I was aware that using coax with clip leads would add to some errors. I was working HF band below 30mhz so thought I might get away with it. Perhaps not.

Lots to learn about making chokes, transformer baluns and s21 measurements. Ill look for those articles on the test fixtures.

Bryan,n0luf
 

If you are measuring baluns and large components like this below 30 MHz you surely can get away with a lot. I guess when you run into situations at higher frequencies you can over do it a bit for these large very widely varying components. As others above soldering these to BNC's or SMA's should be fine but still not alligator clips:) 73, Pete
 

Can we go back to Fahnestock clips? :-\

73,
Gary - W6GVS

At 06:49 PM 3/12/2023, you wrote:
If you are measuring baluns and large components like this below 30 MHz you surely can get away with a lot. I guess when you run into situations at higher frequencies you can over do it a bit for these large very widely varying components. As others above soldering these to BNC's or SMA's should be fine but still not alligator clips:) 73, Pete
 

We have gone down the rabbit hole. Fahnstock clips no longer exist, and effectively never existed! :-)

On Mar 12, 2023, at 7:23 PM, Gary - W6GVS <gvsmith5@...> wrote:

?Can we go back to Fahnestock clips? :-\

73,
Gary - W6GVS

At 06:49 PM 3/12/2023, you wrote:
If you are measuring baluns and large components like this below 30 MHz you surely can get away with a lot. I guess when you run into situations at higher frequencies you can over do it a bit for these large very widely varying components. As others above soldering these to BNC's or SMA's should be fine but still not alligator clips:) 73, Pete
 

No longer exist?

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Don Rolph via groups.io
Sent: Sunday, March 12, 2023 8:25 PM
To: [email protected]
Subject: Re: [nanovna-users] Through calibration

We have gone down the rabbit hole. Fahnstock clips no longer exist, and effectively never existed! :-)
On Mar 12, 2023, at 7:23 PM, Gary - W6GVS <gvsmith5@...> wrote:

?Can we go back to Fahnestock clips? :-\

73,
Gary - W6GVS

At 06:49 PM 3/12/2023, you wrote:
If you are measuring baluns and large components like this below 30 MHz you surely can get away with a lot. I guess when you run into situations at higher frequencies you can over do it a bit for these large very widely varying components. As others above soldering these to BNC's or SMA's should be fine but still not alligator clips:) 73, Pete

1 - 20 of 65