The AM Forum
March 28, 2024, 11:15:02 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: Calculating a series resistance / VNWA Calculation  (Read 5403 times)
0 Members and 1 Guest are viewing this topic.
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« on: August 13, 2018, 11:16:52 PM »

OK,

Bought a new timewaster...  err...  tool.  DG8SAQ VNWA.  What an amazing instrument!

Have a question.  Ran an S21 on a resistor.  It gives me 3.86 dB of attenuation on a straight 50 ohm system.

How the H@ll do I calculate the value of the resistor?  I <<know>> it's a 56 ohm resistor, and the input and output ports are (supposedly) exactly 50 ohms (which I've verified with a 6 ghz calibration standard).

I feel like I should know this formula, but even after a half hour googling, I still couldn't find a formula that gives the result.  I can add a custom trace that will figure it in the VNWA software, but I'd like to know HOW it's being done.

Incidentally, this is to calibrate it for S11 measurements.  You can find the value of your cal resistor via s21, then do an s11 calibration and be within a couple tenths of a percent, supposedly. 


I don't recommend anyone purchase one of these.  I haven't turned a radio on in a week, since it arrived! LOL.  Have made a couple test jigs, tested a ton of crystals, measured a few MURATA 455 filters and found some of of a CB that give almost 8 kHZ each side of the skirt!  I've found optimum impedances of unmarked filters, etc. 

It also has a spectrum analyzer, TDA, crystal filter matching....  And I've not gotten nearly into it.  Can be run into subaudible to 1300 mhz.  600 and change shipped to the door, but well worth it. 

Oh, it's an antenna analyzer, too.

--Shane
KD6VXI
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #1 on: August 18, 2018, 11:01:02 AM »

Bueller?  Bueller? 

Nobody?

Still searching for the answer here.

--Shane
KD6VXI
Logged
w3jn
Johnny Novice
Administrator
Member

Offline Offline

Posts: 4620



« Reply #2 on: August 18, 2018, 09:03:58 PM »

Strictly speaking I don't think you can determine the *resistance* of the resistor with just a S21 measurement.  The resistor is going to present a complex impedance.

However, you can determine the impedance using the formula here:  https://www.keysight.com/main/editorial.jspx?cc=CA&lc=eng&ckey=1536926&nid=-11143.0.00&id=1536926

Z=(2Z0(1-S21))/S21

Remember in that formula S21= V1/V2, not the dB value from your analyzer, so S21(dB)= 20log(V1/V2)

So S21 for the above formula =10^( S21(dB)/20) - in your case, -3.86 dB = .64

Plugging the whole thing in - Z= [100(1-.64)]/.64 = 56.06 ohms.  Pretty close; that's a pretty good resistor and/or youre measuring it at a relatively low frequency.



Logged

FCC:  "The record is devoid of a demonstrated nexus between Morse code proficiency and on-the-air conduct."
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #3 on: August 18, 2018, 09:19:04 PM »

I learned with the mfj 259 variants LONG ago, measure as low as they will go when using as an lcr.

It's a garden variety 56 ohm 1/8 watt MOF.  Nothing special, but I swept from 1khz to 10 kHz :-)

That formula was what I was looking for, appreciate it.  As I said, this is the biggest time waster on the planet (right now).

--Shane
KD6VXI
Logged
w3jn
Johnny Novice
Administrator
Member

Offline Offline

Posts: 4620



« Reply #4 on: August 19, 2018, 04:13:35 AM »

Yeah, you're not going to really measure much parasitic L or C at 10 KHz...
Logged

FCC:  "The record is devoid of a demonstrated nexus between Morse code proficiency and on-the-air conduct."
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #5 on: August 21, 2018, 11:33:41 AM »

JN,

Just tried this out with some dummy load / calibration resistors I had made years ago for my MFJ259 (when I blew the detector diodes out.  Had to replace and then recalibrate).

I <<knew>> my resistors where not precision, they where garden variety Radio Shack 1/8 watt MOF (all we had locally).  I clipped the leads as short as I could and soldered one end into the pl259 center pin and the other lead was brought out one of the holes for soldering the braid to.

However, the one marked 200 ohms, after sweeping and plugging the numbers in, comes out to 202.1148 ohms. 

So....  It works!

Again, wanted to say thanks. 

--Shane
KD6VXI


(incidentally, as to the calibration of the 259b:  These resistors made it work, and they made it fairly accurate in the HF area.  However, the X function was slightly off.  My digital SWR meter would show a perfect match 1.0 to 1, but the mfj would show 50 ohms and a few ohms of reactance.  This was at the upper end of HF.  As it went lower, it was more accurate.  As you approached the 170 some mhz limit, it got more inaccurate.   For reference, however.  The R reading pretty much agrees with the VNA.  It was the X reading that was fubar...  And usually by just a few ohms of reactance.  Accurate enough for antenna work, but maybe not for other things.  Maybe this will help others.)
Logged
w3jn
Johnny Novice
Administrator
Member

Offline Offline

Posts: 4620



« Reply #6 on: August 21, 2018, 05:15:05 PM »

I have a HP 4195A spectrum analyzer/VNA, and a HP 3577A VNA.  They are a lot of fun to play with and see the real world characteristics of components, cables, filters, etc.  The ability to characterize these things at RF is very instructive.  There's an accessory fixture for the 4195A that allows you to connect components right up (which I don't have), and the instrument will do all the math for you.

The 4195A only goes up to 500MHz so ultra precision load/short/opens aren't a real issue, but I scored a big set of them on eBay for cheap.  If you want to go into the GHz range, a good set of HP open/load/shorts will cost you as much as a good used HP VNA - several thousand bux.  The 4195A cost me all of $150 at a hamfest; had a bad video RAM chip which was easily replaced.

The MJF sure has its place - simple, portable, and gives immediate relative readings.  Precision it ain't though - it's certainly impossible to get such results with PL-259s.

Logged

FCC:  "The record is devoid of a demonstrated nexus between Morse code proficiency and on-the-air conduct."
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #7 on: August 21, 2018, 08:05:25 PM »

If you haven't seen the DG8SAQ VNWA, it's well worth a look-see....  As well as to put a bug in Santas ear maybe ...  It's from the UK so I jumped at the thought of a 20 pct US style VAT / Excise / Tariff.  It was about 650, to the door.  Came with an Amphenol SOL calibration kit, and is good to 500 mhz at >100 dB and to 1.3 gig at I believe 70 dB....

The software is the real limitation / star.  I've had a few problems, but they seem to be Windows 10 related.  It does require a PC, however they do have it running in WINE.

I built a test jig after looking online at others both commercial and homebrew.....  Then I thought, do I actually NEED that many hundreds of dollars of accuracy? Smiley  NOPE!   I'll enclose a pic of the jig I homebrewed for giggles.

As to an accurate load.  Yeah, that's no joke.  I've seen people spending more on an accurate SOL load kit than the actual instrument itself.  Insane, and I can't see myself needing anywhere near that level of accuracy.

It has a crystal mode, where you can extrapolate crystal information, a filter sweep mode, spectrum analyzer, DTF/TDR, etc.  The only real rival was the VNA UHF that DX Engineering sells, but it's quite a bit more expensive.....  And about the same level of accuracy.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                   


* 20180821_165016.jpg (4601.71 KB, 4160x3120 - viewed 382 times.)
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #8 on: August 21, 2018, 08:12:35 PM »

Heres pics of scans of the 3 calibration standards I have.  Amphenol, Chinese and the homebrew PL259.

I'm really surprised at my PL259 load.  I expected worse, but it works!  To 40 MHZ, it's flat.  After that, it slowly rises, but I wouldn't trust it.

Anyway, time to go make dinner for the kids.  Thanks again for the formula.  Plugged it into an excel spreadsheet and it's part of the arsenal.


--Shane
KD6VXI


* Amphenol 1 to 1300 mhz.jpg (100.42 KB, 892x711 - viewed 263 times.)

* Chinese 1 to 1300 mhz.jpg (103.64 KB, 892x711 - viewed 256 times.)

* Homebrew PL259 1 to 1300 mhz.jpg (103.85 KB, 892x711 - viewed 265 times.)
Logged
M0VRF
Member

Offline Offline

Posts: 172


« Reply #9 on: August 22, 2018, 02:19:40 AM »

You would 'normally' use the S21 measurement to get 'insertion loss' thru' a circuit and you'd 'normally' use S11 on LH port (or S22 on RH port) to measure a components characteristics.

Using the 'correct' measurement method and using the LH port to measure S11 you'd expect a very good 'Return loss' as it's a 50R system. You have a minimum of 30dB's of return loss as long as it was a carbon/metal film resistor and the frequency was low.

Have fun with your new toy, not sure if you can see a smith chart on there but that's the one to use!

Regards

JB.



* sparams.png (31.22 KB, 830x526 - viewed 314 times.)
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #10 on: August 22, 2018, 08:21:49 PM »

It does have a smith chart function as well.  I didn't have it turned on for those screen grabs because it made the rest of the plots hard to follow:  IE, too busy on the screen.  Cool thing is, I can turn any / all functions off at will so I can step through to see pretty much anything wanted.

I did a couple scans of a Murata CFW455Ht.  One with the correct input and output terminations and another with 50 ohm termination.

The correct termination gave a nice, flat response.  The curve is what it's supposed to be.  The incorrectly terminated filter has a tilt to it, and the skirts are effected as can be seen by the bandwidth markers on the upper left corner.

The S21 measurement of the single component was to check the actual load.  In the manual, since the input and output are exactly 50 ohms, you can calculate the loss through a resistance inserted.  Then, you can use that value instead of just guestimating 50 ohms.  The manual touched on WHAT to do, but fully say how since the formula was truncated.

It really is a fascinating instrument.  And it opens up an entirely new world of looking at things.  Kind of like going from 2d to 3d, but in the realm of the components we use.

--Shane
KD6VXI


* Murata CF455 Correctly Terminated.jpg (97.07 KB, 892x711 - viewed 267 times.)

* Murata CF455 Incorrectly Terminated.jpg (96.68 KB, 892x711 - viewed 274 times.)
Logged
M0VRF
Member

Offline Offline

Posts: 172


« Reply #11 on: August 23, 2018, 02:20:37 AM »

Yes indeed an Net Ani tells all, nice that they're commonly available.

I'm fortunate to own an HP 8712C and use an R&S ZVA at work, somewhat of a different animal!

Any Net Ani makes antenna tuning and making filters a real breeze!

JB.
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #12 on: August 26, 2018, 12:57:07 PM »

<snip>  The incorrectly terminated filter has a tilt to it, and the skirts are effected as can be seen by the bandwidth markers on the upper left corner.

</snip>

--Shane
KD6VXI

By how much was it mis-terminated?

Also, how about a link/pix of this magic box? Curious.
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #13 on: August 26, 2018, 01:13:19 PM »

Bear,

It uses a 2.2k resistor on the input and output.  I bypassed the resistors for that test.

Sdr-kits is the seller.  Has to be ordered from the UK, as the guy who makes the Vna for AIM has threatened legal action.

It's the DG8SAQ VNWA.

--Shane
KD6VXI
Logged
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.087 seconds with 19 queries.