Some real world measurements, on coaxial stubs.
Jules Smith )
Sunday, 10 September 2017|
I recently had a spectrum analyser, with tracking generator, to try out. After testing every filter and duplexer I could find, I tried sweeping some simple coaxial T stubs. The results were pretty much as per the text books, but interesting to see in practice.
So out of interest, I though I'd share with you.
Note that the marker positions on the wide bandwith sweeps are a bit course, onl;y set roughly, so +- a few MHz is expected. This should not detract from the findings.
A quick explanation, in case you have not used spec-an's before....
A spec-an tracking generator, is a sweeping oscilator, that acts like a signal generatotr giving an output at the same frequency that the spec-an is receiving. So you can use it to test passive circuits, such as filters, or feed it into amplifiers etc, and characterise them. As well as looking at the outputs of units that generate frequencies, as you would do, when tuning up multiplier chains (if anyone does that anymore?) or testing the output of an exiter/TX.
OK, on to the stubs.
These were connected using a good quality N type T adaptor. As the cheaper ones can have the T connected with a spring. Which is extra inductance, which we don't want.
The stub was just an N type RG76 style patch lead, about 8" long. Found at random in my patch lead box.
An initial sweep with the stub end open circuit, showed a -30dB notch at 230MHz. (notch 230.jpg)
This looked about right, as the quater wave notch frequency (taking the velocity factor of the coax into account). But being a transmission line, is will also notch at each odd 1/4 wavelength interval.
(third harmonic notch 693 oc.jpg and full sweep oc.jpg)
It was interesting to show what happens when you use the same notch, but with the end short circuit.
The notch will move to the half wavelength frequency. In this case 439MHz. And now will pass the 230MHz, though with some attenuation still. The open circuit line is shown, (notch 439 sc.jpg and pass 230 sc.jpg)
For comparison:The open circuit stub, with the marker at the new 439MHz point so you can see it pass, with little attenuation.
(pass 439 oc.jpg)
OK, where could we actually use a stub?
Let's assume we are need to notch out the third harmonic of a transmitter. For this length stub, our TX would have to be around 73MHz. (pass 73 and 3db points.jpg)
In this case, the attenuation of our third harmonic is a healthy 30dB+, but the loss at the fundamental 73MHz, is 2dB. That is a lot of loss. Far too much for this stub to be useful in a power transmitting system. This is why stubs are not usually seen on transmitter outputs.
However, if it were to be used at a lower level point in a chain, say in a driver stage, where we may have some spare gain and be able to stand some loss, this may well be worth considering.
Coax stubs are more useful to notch out strong locally interfering signals on receive systems.
In this case let us assume that the interfering signal is at our notch of 230MHz, and we want to know how the stub will affect our wanted signals.
I have plotted the 3dB points of this stub. And they are at 168MHz and 300MHz. So trying to receive signals outside this range, will be attenuated by less than 3dB, whilst the strong 230MHz would be deeply attenuated.
(pass 73 and 3db points.jpg)
The losses below 168MHz, you can see, only trail off gradually. So you may not want to use this stub as an antenna filter directly. But stub filters may be still of use after a preamp stage. Where loss is less important.
Traditionally the problem with strong adjacent signals, has been that they can saturate preamps. But the high dynamic ranges of current devices means, that it is getting more common to put filtering after the gain stage. But this needs to be decided on a case by case basis.
As a real world case, this exact stub may help me reduce desense problems, from my local DAB radio transmitter, into my 145MHz preamp. If the DAB is not actually overloading the preamp itself, but getting through and causing problems at the rig, I could put this stub after the preamp. Or, if really strong, I could use it before the preamp and decide that losing 3db receive at 145MHz is better than an overloaded preamp. Though in that case a different filter may be a better bet. But if that were all I had available, something is better than nothing.
Though not shown with the markers, the 460MHz is hardly attenuated at all by the notch. So using it to attenuate an interfering signal at 230MHz from getting to into a 460-oddMHz receiver. And up here the attenuation of the wanted signal is negligable.
So for blocking the DAB transmitter from flattening our 430MHz receiver, this stub would work well, and hardly attenuate our wanted signals.
Similarly, if we used an appropriatly scaled stub on our 144MHz NFD contest station, it would be very effective against local strong signals from our other station on 70MHz, if we found that it so be overloading our front end.
Though if we were to try the same trick against the 50MHz station, we would find that the 3/4 wavelength effect of a 50MHz stub (notching also at 150MHz), would render our 144MHz station, useless.
If anyone is interested, the spec-an used, is the Siglent SSA3021X, with tracking gen option.
Till recently, spec-an's have been exeptionally expensive. The only way that Amateurs have been able to to own one, was to spens a few thousand pounds on a 25year old, used commercial unit. But over the last few years, companies such as Rigol have been producing quite respectable units for under £1500. Siglent have moved into this market too, but their offerings really are giving the likes of HP/Agilent a run for their money.
If you are looking for a budget, but very capable spec-an, do have a look at these.
(I have no connection with Siglent, except being a very impressed user!)
|Attachments: (Click to view/download):|
Email author |
Link to this page | Print | Log In
Get your FRARS sweatshirt, polo shirt or cap embroidered with your callsign!
FRARS runs regular courses to help you prepare for your amateur radio licence. |
For more information Click here
||UK Microwave Group
In early 2004 the UK Microwave Group (UKuG) became the representative voice of the UK amateur radio microwave enthusiast after the RSGB disbanded its own Microwave