After an hour or more of pushing and twiddling various buttons and knobs on my VNA I have at least established two standards: I connect a short and the (polar) spot flys off to the extreme left-hand-side and sits stably on the central horizontal axis. I then remove the short and replace it with an open and the spot jumps across 180 degrees to the extreme right-hand side on the same plane. All well and good so far. Next, I replace the open with a lab-grade 50 ohm termination and I wind up with an arc on the outermost circle of the polar chart at about two-thirds of the way round the circumference (clockwise) from "12 o'clock" as opposed to a spot in the dead center of the display as I would expect. What have I done wrong? I have checked this with 3 different 50 ohm loads and they all show an arc at the 'south west' corner of the chart. What's goin' on here, guys??
You haven't fully described your setup, so it's hard to tell. What are you using for a bridge? If I recall correctly, your VNA is an HP8754A, which won't do S-parameter measurements by itself; it requires an external bridge (an S-parameter test set, or a reflection-transmission test set, for example).
The fact that you see an arc on the outermost circle, assuming that you have it set as a unit circle on the linear reflection coefficient plane (the Smith Chart outer circle), implies that you have a (nearly) unity reflection coefficient. Is something almost disconnected so it looks like a tiny series capacitance?
Okay, thanks. I must admit it is symptomatic of the load's center pin not quite making contact for some reason, even though the load is fully screwed home. But the same thing happens with 3 different loads! I'm using the T/R test set, by the way, although I don't see how that could explain the problem. I don't have the full S-parameter bridge but am in the market for one.
If you're using N connectors, check the center pins of the females. Make sure that they aren't splayed open.
The local repeater association I'm a part of has had repeated problems with splayed N connectors. They can be caused by a couple of problems. A botched install of an N male can cause its center pin to project too far forwards, out of the shell. We've also seen some N adapters (Chinese-made I believe) whose center pin was larger than usual in diameter, or was not tapered in the usual way.
In either case, if you screw an N male connector with a bad center pin into an N female, the male's pin can force the split segments of the center female pin outwards. Once this happens, a _good_ N male connector's pin may not make reliable contact with the female's damaged center pin. Intermittent nastiness can occur.
It's sometimes possible to bend the female's split pin sections back together with needle-nose pliers, creating a repair which may or may not be permanent. A better solution is to replace the damaged female connectors entirely, and inspect all N male plugs and discard any whose center pins are too large in diameter, not tapered properly, or project too far out of the shell.
I suppose that similar problems might occur on BNC or SMA connectors as well, for related reasons.
--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!
Okay, thanks for that guys. One clue that might assist in tracking down this problem is that the aforementioned arc that appears at about '37 minutes past the hour' on the outermost polar circle when the 50 ohm termination is fully screwed in first begins to appear at just before 12 o'clock and gradually circumambulates around this outer circle anticlockwise to its final position as it's fully screwed home. I'd have (intuitively) thought that once the centre pin makes proper contact with its counterpart, the trace would suddenly flick to close to the centre. But with no experience of VNAs, that's just a hunch. Any suggestions?
Dave Platt wrote, among other things, "In either case, if you screw an N male connector with a bad center pin into an N female, the male's pin can force the split segments of the center female pin outwards. Once this happens, a _good_ N male connector's pin may not make reliable contact with the female's damaged center pin. Intermittent nastiness can occur."
I've seen people suggest that you can mate an N plug with a BNC jack in a pinch, but what you describe is _exactly_ why you should never do that. The N plug center pin is a distinctly larger diameter than a BNC center pin.
Also note that there exist 75-ohm N connectors. Do not try to mate them with 50 ohm versions. A 75 ohm plug in a 50 ohm jack is likely to not make good contact, and a 50 ohm plug in a 75 ohm jack may cause damage to the jack.
Suggestions? You bet! Get out your RFSim99 and build a model of what you are putting on the RT test set test port, including any connecting cable. Play with the model to see what will get you the sort of performance you are seeing.
Note that you haven't really told us what you are plotting. I'm assuming it's a plot of the complex reflection coefficient on a display that's scaled to a unit circle. Are you using a standard Smith chart overlay?
The normal interpretation of the result you describe, however, doesn't make much sense, since the load seems to go from something like a nearly pure inductance with reactance equal to the reference resistance at the center of the plot (50 ohms?), and wanders around to become a very low impedance (at the left edge, -1+j0 reflection coefficient), and then around to become capacitive with moderate reactance (35 ohms?). Could you possibly be using some other scaling? What happens if you are trying to display a value beyond full scale on your display?
I suspect something is not what you think it is in the measuring instrument, and from past experience, I'd say it could be just about anything. Careful visual inspections often reveal problems.
Yes, thanks, Rene. But for something as basic as short, open and matched terminations, I'd assumed the spot displayed should just hop to each appropriate place along the horizontal center line. I'm not even doing any transmission tests at present; this is so far just reflections only...
Oh dear. It appears you may be right. I've just checked with the suppliers and they tell me the plugs I bought are 75 ohm ones. This really stinks as in their catalogue, although they don't specifiy the impedance of these plugs, they do clearly state that they're suitable for RG214 coax. AFAIK, RG214 is only available in 50 ohm - if anyone knows otherwise please tell me. It took me a good deal of time and trouble to make up these patch leads so all in all, I ain't exactly pleased. Here's the URL for the catalogue page so you guys can tell who's at fault here:
Since posting earlier I've been on to the actual manufacturers and they reckon these plugs are actually 50 ohms. So the supplier's insist they're 75 and the makers say they're 50! I need to know for sure, as one or the other is in error. How can one tell, by visual inspection and or physical measurement, which type is which? There must be something that's visibly different about the two types. THanks.
Yes, look at the thickness of the central pin. A 50 Ohm connector has a central pin of some 1.6mm thick. The 75 Ohm version is a mere needle, something like 0.8mm.
It's one of my long-time gripes about coaxial connectors that N and BNC series of the two impedance levels are sufficiently compatible to mate together, but not enough to do so without damage. In a lab with someone ignorant of this fact, it's a source of endless trouble.
Thanks, Jeroen. Well on that basis, all the plugs I have here are indeed 50 ohms, then. In fact I pulled a centre pin out of a new packet from the same batch and it fits the sockets in the VNA perfectly. They all accept the pin with a nice snug fit and what's more the sockets show no signs of ever sustaining any damage/spreading/flairing/bending and that goes for those on the T/R bridge, too. On reconnecting everything, however, the problem still persists, although the 50 ohm load is now showing as very close to a dead short; virtually the same spot on the Smith chart overlay as when the socket is shorted, in fact. Once again, the other loads do likewise and they all check out fine as 50 ohms with a DVM. Very strange! I wonder if there's some setting I may have got wrong somewhere. :-/
What would be the result of calibrating the system with an inductive "short"? If it can be done, results of testing with a good 50 Ohm load might be skewed.
Having had a look in my (junk)box with N-connectors I found that most silver plated (professional) units have 50 Ohms stamped into the body ,whereas the nickel plated units either have no identification or a specific manufacturer's code . The Farnell catalogue you referred says these N-connectors are nickel plated . Try to find any stamped code on your units . If any I'll be happy to compare with those on my connectors.
Frank GM0CSZ / KN6WH ==========================================
Good thinking, Fred. That's theoretically possible, I guess. But I'm carrying out these checks purposely at only around 5Mhz to minimise such high frequency effects. You may well be on the right lines, though!
Have you tried buzzing your home-made leads through with a DVM to confirm that they're ok (at dc) - both centre and shield of the coaxial connectors?
Try discarding the test-set for a moment and just connecting Ref to the A and B inputs on the VNA (configured to measure loss/gain). If you know that your coaxial leads are ok, then you'll be checking that all the I/O ports on the VNA are working correctly.
Have you confirmed that the test-set is correctly connected to the VNA?
Try terminating unused I/O ports with a 50 Ohm load or with a pad (a
10dB pad makes a reasonable 50 Ohm load) and then repeating the cal and measurement.
Try calibrating the VNA/test-set without the leads, just using a Male-Male adaptor instead. Then see what happens when you measure gamma with 50 Ohms dangling off the adaptor.
All the 75 ohm BNC connectors I have, have the same diameter center pin in the connection region as 50 ohm connectors. The difference is in the thickness of the dielectric around the pin in the connection region, being much thinner in the 75 ohm connectors. The pin of the 75 ohm connector is smaller only where it's in solid dielectric at the bottom of the connector.
There's lots of room for confusion. The Kings tool FAQ implies the center contact diameter got smaller when the pin was redesigned and the old crimp die set would not work:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Why won't my Kings 2025-X-9 die set work with the 2065-X-9 connectors?
With the redesign of the True 75 Ohm BNC connectors, the center contact diameter got smaller to maintain a constant 75 Ohm impedance through the connector. Consequently, the center contact crimp dimension got smaller also.
The problem with the inability of the die set to crimp this new connector is probably because it is the older, larger die dimension. The Kings website has a Cross-Reference Search that lists the applicable die set for most Kings connector part numbers.
formatting link
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ However, another manufacturer claims their new 75 ohm hardware is interchangeable with 50 ohm connectors:
Emerson Network Power Connectivity Solutions is introducing its premium line of TRUE 75 Ohm BNC connectors and adapters. These flexible cable 3 Piece BNC Connectors are essential for analog and digital high frequency signals.
They have been designed around MIL-STD-202 for up to a 3GHz frequency range, and are constructed of high performance materials including a machined nickel plated over brass housing, a teflon dielectric and gold plated brass contacts. These connectors are designed to eliminate distortion and impedance mismatching caused by using 50 ohm connectors on 75 ohm cables, and they safely inter-mate with standard 50 Ohm BNC connectors.
formatting link
Ohm.pdf
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ If your junkbox is anything like mine, it may have stuff purchased 20 years ago mixed with more recent purchases. So it might be worthwhile to check carefully before mating different hardware.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.