a 50-Ohm-source,e.g. a tx or a vna is an ideal voltage source (source resistance = zero) in series with a 50 Ohm resistor. Hence these 50 Ohm are in series with the antenna impedance no matter if you note the antenna impedance in series or parallel equivalent. If you are looking from the ideal source to the load then the source resistance combines with the load's impedance as seen by the ideal source. But in reality your 50 ohm source,e.g. tx or vna,is looking to the load. Therefore the load i.e. antenna impedance, as seen by the 50 Ohm source, is NOT changed in any way by the source resistance itself. The 50 ohm source resistance is not lying in the path to the antenna. It is so to speak behind the point where the vna is measuring phase and amplitudes from which it calculates the impedance.
Just connect a known resistor with e.g. (a few) hundred Ohm via an arbitrary length of any coax with your vna. Calibrate out the coax. You will measure again the true value of the resistor and nothing different like "Z new" .
WELL, sorry for that, but this is going to be a long one; so i'll make it into two postings. I removed some material which i though might not be pertinent to the VNA2180 cause in general, to shorten the text, but i feel the rest is pretty relevant to its use and operation.
CLEMENS, When i read the pdf and saw "voltage source", thought this was not going to influence the load being measured; but getting curious, the block diagram (which i usually believe more than the description) showed the Zout as being 50 ohms; i thought "not truly a voltage source, per definition"; why would the diagram actually state that: Zo=50 ohms. I think Bob actually means 1v pk through 50 ohms, to indicate the driving current available from the op-amp?
It took me this long just to find out about the effect of the source impedance on the load, because just after my posting we had icing and pretty bad. All wires including the 5 delta loops became covered with an inch (yes one inch) diameter of ice. The loops were not quite ready for this yet, and one tower steel support for the rope broke. Had to do some repair.
So in order to show that the VNA2180 does not affect the complex ant. impedance, I have done a back to back reading using a 15 ohms resistor in the coax line (and custom recalibrated). The Rs actually changed from 174 to 170 ohms; certainly not the parallel equivalent of 174 and 15 ohms. So this and in view of the high output Z of port A (and not a Zo of 50 ohms as shown on the VNA2180 diagram) i am satisfied the VNA is not a concern.
BOB: if you are reading this, I must say that i have really run the devil out of the 2180 in the last 3 weeks, compiling over 100 hours of use. So have found quite a few things you might be interested in knowing. I must say after those hours of continuous use that i get more and more confident and impressed with its performance.
1.--------------------------- First, you might take a look at "error 13270" which I first got after doing a custom calibration with a carbon comp. (the old type) 120ohms resistor in series with the output of the coax center lead piece, and going straight to the calibration standards. I think the VNA is looking along the coax and seeing average reflections till it gets to the end where it sees a bump of 126 ohms (a sudden reflection bump) then sees the 50 ohms standard. It tells me the "cal load may be in error....but may be used if those concerns are outside of the areas of interest". Guess it expected me to tell it, that the calib. resistor was a 176ohms resistor.
It keeps giving me that "error 13270", sometimes when i reload that stored custom calibration.
I also got a 23827 interpolation error, when it told me that the calibration file was invalid. But it still gave me the readings.
Later i got "error 13270" again after running "Point Data" (using a standard calibration file) and with that window still displaying, I go and add another "marker". It froze the scans for the active/connected antenna to port A; i could have shut down the pgm, but decided to just load a "good" scan. Then it was okay.
2-------------------- Is there a list of error definitions somewhere so you get an idea of what actually drives that error?
3-------------------- Later that night, while reviewing the scans stored during the day, i got a warning that the DC supply voltage was too high at 24vdc and that the power had been turned off (power still stayed on). I am using the dc module supplied with the vna; i measured with a dvm and was 12.58vdc with no load.....
VNA was still scanning and running okay though. Temp in shack was 14*C; (yes i keep it pretty cool in). Rebooted the 430 version and all was okay. Running Win XP PRO SP2.
4--------------------- After finishing my scans later in the day, i like to recall stored scans and look at the data on the markers. With rev. 430 running and with an active com port (should not matter) i recall a scan and call out the markers' data. It gives me swr 99, phase of minus 90, Rs of 0.000 on all the 3 markers used on the graph (all markers lie within the range of the scans and graphs).
The markers' Xs are -2.659, -2.692, -2.725 and the Zmag are those same values (since the Rs are all 0); those numbers have nothing to do with the recalled scans. Changing scan file, gives exactly the same results on all parameters.
5------------------ TREND does not take "1" for input. Going to zero seem to be the only number to give the proper parameter values. Using numbers from 2 to 20, keeps giving larger errors away from real values, as far as Xs =-213 (should be -1.8); phase of -35 instead of -0.7; other paramaters are also very very far from the real values.
Playing with Averaging does not improve/deteriorate the wrong results. Something strange (strange to me anyways), the current graph in use is from 3 to 4 mc; so, when you go to higher TREND numbers, a new scan will now stop short of the 4mc limit, as far as stopping at 3.677 for a Trend number of 20. A new scan will stop higher and higher as the numbers in Trend go from 20 to 0.
6------------------------ Drifting values: when you first turn the unit on, let it warm up for ten minutes to stabilize. The readings drift but stabilize; leave the unit on all day, if you are going to spend the day making measurements. I also noticed some changes if you run the unit on continuous mode, like the GRAPH RECYCLE, TUNE, or MARKER RECYCLE. Many will say it's not much, but almost as much as when you first turn on the unit. (maybe i've got a unit that drifts more than the "standard"). The major drifts are on Xs going inductive from -4.2 to -3.9 and the phase going from -1.1 to -0.9 (about ten percent).
Okay lets be practical we say; but when i magnetically couple two of the loops together (with no wind, cuz the swinging wires change the values), at quarter wave spacing, the inductive coupling gives a reactive change from an X of 0j to +40j; but at half wave and beyond spacing, this reactive change can be as little as +2j or +3j.
The resistive component change is even harder to measure, going from 210 to 206, an R change of 4 ohms or even less; and it is important to know which direction the shift is taking place.
I'd like to start with what can be reasonably obtained and minimize the errors here. This is a critical step to calculate my LC phasing network components. The coax stubs I will be using for the LCs, can get expensive to recut or throw away, not figuring out the time spent recutting.
Error code 13270 indicates a communication problem between the VNA and the PC while doing a scan on Port A. Make sure you have the latest usb driver from FTDI.
A comm problem can also cause the "high voltage" warning. The supply voltage in the VNA was not read correctly.
For best accuracy, a five minute warm up time is benificial. The latest version of the program leaves the power for the DDS chips on all the time but there is an option to set a "low current" mode if battery operation is employed. Low current mode turns the DDS's off between scans to reduce the power dissipation.
The latest program is version 440A.
When calibrating, always enter the actual value of the cal resistor, even if there are other circuits in series with the RF output. Standard Cal checks this value and warns if it seems to be inconsistent. Custom Cal does not check the load resistor, it assumes the value was entered correctly. Custom cal is different in this regard since the apparent resistance at 1MHz may be very different from the resistance value if there is a transmission line or a filter in the calibration path.
I'll take a look at the interpolation error. Send me the start/stop/delta freq values you were using for the custom cal and I'll try to reproduce the problem.
If the DDS chip power off feature is used, users should be aware that the readings may not be stable if they make several readings in a row. The chip then should pbly be left on to temp stabilize.
For anyone who cares to know, today i've decided to just run the VNAcontinuous to see what stability i can expect. I found out that it took 3 hours of continuous power on to reach the same value parameters as yesterday. I will list the first few readings and spare you details and only list half hour readings after that.
Yesterday when i shut off the VNA after running it all afternoon, the loop 5 parameters were 223.7-16.8j; phase at -4.25* and loop resonant freq given by phase crossing the x axis was 3537.6 It was +14*c outside and 16*c in the shack. Calibration steps at 0.020mc
Then i recycled the graph 217 times and got 225.8-17.1j; -4.40*; 3537.3kc pretty stable now.
Then i turn off VNA for 45minutes and the first few readings 226.0-11.9j; -3.0*; 3532.6kc
half hour later seems more stable and 226.8-17.5j; -4.41*; 3537.4kc guess it did not have enough time to cool off to the morning temp.
Seems that the short term variations are greater for reactance and phase. The resistive part show more variation with longer times.
So now i can explain a bit more my suspect readings after running outside to do loop changes and coming back in to find the readings were not what i expected them to be. It's easy enough to turn on the unit when i get up and after breakfast and equipment prep. the VNA and me are ready and pretty stable....
BOB tks for the replies. when i custom calibrated and got the error 23827, start/stop were set at 1 and 20mc and delta of 0.020mc.
Have now loaded and using ver 440a, and still not getting marker data when loading up a stored scan. Guess that feature works only when you have a live scan on a live port.
Also found a recently released driver for the FTDI. Ver 2.06.02 Mar 2010, The version 2.04.16 is not to be found on their site any longer.
VER 440A coming back from running Port B, i load a new cal port A only calibration file, and load an existing port A only scan.
Have the 220 ohms antenna connected to port A. I open the markers which give me the 220 ohms + Xj values correctly. Then i ask to run a scan A, and the data on the right of the graph is totally out of context like Rs 19ohms and Xs of -2.65; but the marker data updates with new CORRECT VALUES of the antenna.
Done this several times and the only way to get it scanning properly is shut down the program. SCAN, RESCAN OR RECYCLE do give slightly different values, but nowhere close to what the markers indicate (marker values are always correct).
.....AND ver 440a also does not remember the settings of "AVERAGE" value and "TREND" entered in the rh side window after shutting it down, but remembers "INTERPOLATION" setting. Scan also still stops short of the full freq. scan when you set "TREND" to any number other than 0 and do a rescan.
Remembering the "last value" is a double edged sword. If you are working on project "A" and you stop at the end of the day and start again tomorrow remembering the last settings is a GOOD thing. However if tomorrow you change yopur mind and work on project "B" I would much rather see the analyzer come up in a default state. I have hundreds of hours on both the 4170 and the 2180 and I have been burned more times by the analyzer remembering things than not remembering them. Very nasty things happen if you have "line extension" enabled and don't remember it. I would often like to see a PRESET button similar to the green button on HP/Agilent equipment.
LARRY, along those lines of "remembering things..." it becomes handy when you have to shut down/restart the program often. I also got caught by forgetting to put back the original set parameters before running next scan.
On my previous post, i mentioned that the program RH side readings of the cursor parameters, were wrong after coming back from PORT B, i now find out that this also occurs after you change calibration file within the same PORT A. Again, the "marker data" gives the new correct scan values, but the RH side cursor position readings stay pretty close to the last Calibration File values; this give a false impression that the new cal file does not affect your new scan. I lost precious hours till i found that out. I now end up shutting down/restarting the program dozens of times a day to make sure i get the correct readings.
THIS IS A MUST ISSUE CHANGE REQUIRED. Again, running VNA VER 440A WITH XP PRO OS.
Maybe all those preset and "new calibration file" screwing up readings, issues would be better solved by having a reset button in the bottom window, next to "HALT".
The next issue to address is on the DISTANCE TO FAULT. using 280 ft of 0.66VF rg59 and shorting the far end gives 280' But opening the end says cable lossy or terminated in a load; says "Maximum phase = 38.765 deg. It should be closer to 90 deg." This is okay if you have the coax in the shack, but not too handy if your cable is buried under 6 feet of snow!!!!...
Lastly, the program froze the whole system; i got the details of the cause of shutdown (files conditions when it froze, checksum, etc) and was able to take screen pix of those. If interested, i can email details to whomever might want.them.
SOLUTION TO THE NEW CAL FILE PARAMETER READINGS. To follow up on my previous post indicating that the values shown on the right of the curves for the cursor position, are wrong after changing CALIBRATION FILE, i posted that you had to reboot the pgm to get the correct values. WELL, after trying just about every combination to avoid rebooting the pgm after changing calibration files, i found out that hitting the "LIMITS" button (without having to change any parameters in that window), just hit "ENTER" and the program will now show the correct values of the scanned port.
And on that same subject of problems with SCAN values after changing CALIBRATION FILES, the MARKER DATA seems to average new readings with the readings from previous calibration file, to give the new values; hitting the REFRESH 2 times or more give the correct new parameter readings.
Lastly i have also found that FIREFOX, in particular "JQS.EXE" (a java routine), is one of the culprits interfering with the VNA program (freezing screens). In past two, three days have run considerable time on the pgm, keeping FIREFOX off without problems.
.....WELL, on that subject of using "LIMITS" to sort of make the graph cursor values correct after changing calibration file, i ran it all day trying this "so called solutions" i made over 75 calibration file changes today while tuing the 80m 5 loop array, and it worked okay 73 times (i checked it everytime, sort of not really trusting that solution; and glad i did....).
So i guess the "SOLUTION" is not really fool proof and we still need some sort of on screen reset, if we're not to turn off/on the pgm everytime. I still saved considerable time by using the "LIMITS" change "solution" though.
BTW, have not had a single "seizure" all day, thanks to turning off "firefox" and "jqs.exe".
Another lil crux, do you know that the rh scale numbers are not following the graticules? For example if you set the Xs or the Phase for scale of 5, it shows at half point the number 3. Funny, the swr (lh scale) gives the correct numbers according to the graticules.
So you have to fool it, by entering a max scale number that gives an integer when divided by 2.
But i must say that VNA2180 is a joy to use; i'd probably go crazy (or at least a bit more so!) if it was not so precise and user friendly.
VNA 2180 MULTIPLEXER? QST of march 2011 has a great review of the product. It mentions the availability of a six channel multiplexer for use in setting up antenna arrrays. Searched Array Solutions site and the web,but nothing found. Any info on the subject Bob?
BTW i finished phasing/tuning the five 80m delta loop array. Was quite a job to do, but with the 2180, it gave me a pass to success. I now enjoy almost 10 db gain to Europe in both receive and transmit mode. It is like having 10kw suddenly popping sigs into Europe...
Next project is a six element vertical array of CVA (closed vertical array) type. Sort of a Bobtail modified to give 50 ohms Z at feed point. That multiplexer may come as a help to ease the pain???
The six channel multiplexer will be available in about a month. We have the first one in operation and the design of the case is being finalized. In the latest edition of "Low Band DXing" ON4UN has information about this multiplexer. He used a prototype five channel mux for his data, which is similar to the six channel production model.
Each channel is calibrated, so if the cables to the current sensors are not exactly equal, they will be corrected. The mux is controlled by signals from the expansion port on the VNA2180. Ports that are not selected are terminated in 50 ohms.
TKU BOB FOR THE REPLY and for the screens; I want one of those!!!!
Will be impatiently waiting. Have they figured out a rough price yet? This might have saved me buying another 4 channel scope for the 80 meter 5 loops array. But i still have to figure out a new network for the summer impedances after the snow melts. This might prove an easier way to get to it. Tks again
It is possible to save all your settings in a State file so you can restore the scales and limits quickly when changing from one application to another. This file is like the ini file (which is saved automatically) but you can give it a unique name and there can be any number of State files. See the Files Menu.
The data on the right side that is readout by the cursor is obtained directly from memory for the last scan that was done. Loading a new cal file doesn't affect this data until another scan is done. Reading the markers does get new data but only at the marker frequencies.
When a new cal file is loaded it's not necessary to exit from the program. I often load a new cal file and click Rescan to see the effect of different cal files.