Do you really need a scope for RF work?

[MrSlack]

I was going to buy a new 100MHz scope but I haven't really touched my current Tek 465B scope much in the last few of weeks while playing around with RF. I've been primarily using a homebrew RF power meter for peak reading, an RF probe and frequency counter and not much else in between. If I poke a scope on the signal I can see the carrier fine but my resolution as a human is finite when it comes to harmonic content etc. Also my scope bandwidth tops at 100MHz (or just past it).

I'm starting to think that a calibrated RF power meter and spectrum analyser with tracking generator are way more important. I've already got a half decent signal source (Marconi 2019A) which as proved invaluable even if it did try and kill moving it.

Are my assertions correct or have I missed a giant use case here? Is an older SA a better investment than new scope?

For ref,looking at 40m, 2m and 70cm bands.

[Kalvin]

You can use a cheap SDR dongle as a simple spectrum analyzer. The dynamic range is a bit limited, but you should be able to cover 100kHz - 1.7GHz with a modified SDR dongle. Add a suitable SMA/BNC-connector and use an attenuator as needed.

With the scope you can see the waveform shape alright, but typically in RF work the spectrum is more interesting than the waveform shape itself. Analysing the modulated signal in time domain is quite challenging, but in frequency domain it is very easy to see the modulated signal's bandwidth, harmonics etc.

[MrSlack]

Interesting idea. An cheap SDR costs virtually nothing as well.

[tggzzz]

Time domain instruments (e.g. scopes) are useful for analysing signals defined in the time domain; no surprises there! Ditto frequency domain instruments such as spectrum analysers and network analysers. There is a third class of "communications oriented" instruments that attempt to bridge the gap, sometimes called modulation-domain analysers, sometimes other things. Typically they have a general purpose core with add-on "personalities" for specific modulation schemes, e.g. GSM, CDMA etc.

My attitude is usually "don't get something until you have a demonstrable need for it". All equipment has "subomtimalities", a learning curve, a space requirement, and a lost-opportunity cost (financial and otherwise). So, if the Tek 465 is sufficient for now - that's great. When it isn't (e.g. debugging a PSU switch-on sequence) then get something sufficient for that purpose.

My current setup includes a scope, an SA, an SDR used as a cheap portable SA, a cheap sig gen ("sine" 2.2-4.4GHz, "square" below 2.2GHz), and a cheap noise source. I wouldn't mind a (vector) network analyser.

The noise source + SDR are (with limitations[1]) equivalent to a scalar network analyser suitable for measuring a filter or network's amplitude response. Using software to track the sig gen with the SDR would also be a scalar network analyser. FFI, see blog in my .signature.

[1] low clock spurs, amplitude-frequency response not calibrated, limited dynamic range

[MrSlack]

Fair point about frequency domain. It's damn obvious but I didn't make that association.

Interesting point with the noise source and SDR based SA. My main applications for an SA are testing oscillators and (with noise source/tracking generator) measuring frequency response so that would do the job nicely.

Will have a read of your blog - many thanks.

What I really want is a standalone 50 ohm SA with tracking generator for a tenner (how much my Marconi 2019A cost me) but that's not going to happen

[hendorog]

Are my assertions correct or have I missed a giant use case here? Is an older SA a better investment than new scope?

For ref,looking at 40m, 2m and 70cm bands.

I learned so much about RF from getting an SA that I wouldn't have otherwise. I think it is well worth it and would be more important to me than a better scope.

You can cobble together quite capable gear very cheaply, but it is hard to know what it is telling you if you don't have an instrument you can trust.

I got my 2019A for a tenner too! - well NZD so maybe 5 quid...

[Nuno_pt]

For an SA+TG it will be around $1500. Or the 815 from Rigol.
Another good thing that I'm looking for is an Communication test set, like Motorola, R&S, Marconi/IFR, this ones are of great value, since they bring many things for you to use in Ham or RF Work.

I'm still looking for an Sinal generator at an decent price, at least go to 1.5GHz.

[jpb]

My first job as a new graduate was working on GaAs monolithic microwave integrated circuits (MMICs) in the range 10GHz to 18GHz or so and we never used scopes (this was the early 80s). All characterization was done with S-parameter test gear (network analyzers) - these were rather expensive though about 20 times the cost of my flat! One reason though was that scopes didn't go up to those sort of frequencies then (I think).

I would have thought that RF was mostly in the frequency domain whilst scopes are really useful in the time domain.

Having said that, I love having a scope now, but I don't really work in RF.

[VK5RC]

+1 re the SA,  as a ham with interests in the above 1GHz,  I use my SA much more often than a scope.  I went the SDR way first and as my interest grew (and also got a bit annoyed chasing 'birdies' and other false 'signals' )  I got a second hand but good HP SA.  Never regretted it, but it is the most expensive bit of gear I own, (a DC block lives on the input).

[vk6zgo]

"RF" like happiness in the old song,is "different things to different people".

If you are working on RF in the region from just above audio,up to around 30 or so MHz,a  standard 100MHz oscilloscope will be a useful device.
Above that,it becomes less useful,until at VHF,it reaches its frequency limits.

A Spectrum Analyser is the best device to look at RF,but doesn't lend itself to probing around inside stuff.
Many times,we just want to have "a look"not do an accurate measurement,& there is where the convenience of the 'scope becomes important.

It's nice to have both!
On one occasion,I was confronted by a seemingly anti-intuitive situation with an FM Broadcast Transmitter.

It had a FET PA  driver stage,driving a Tetrode vacuum tube as the final stage.
The driver had a meter showing its output power.
Intuitively,you would tune the input tuning of the tube stage for maximum driver output on that meter.

Maximum output of the driver should logically  correspond to maximum output of the final stage,but in fact,it coincided with a "dip" in final power output.

Luckily,there was a test point at the driver output,& placing a 200MHz Iwatsu Oscilloscope at that point showed a signal suffering from bad second harmonic distortion,with high,narrow peaks.
The "power meter" on the driver was a simple "peak-hold" detector,so read highest when the signal was most distorted.
The final stage was selective enough to reject the second harmonic,hence the Transmitter output power was reduced when tuned in that condition.

The 'scope made the fault condition immediately obvious,
A Spec An would obviously show the fault,too,but the association between the  high second harmonic & the high meter reading would not be as immediately apparent.

[MrSlack]

I've managed to borrow a Tek 494 for a couple of months while an ex colleague is on holiday. Will see how I get on

[C]

Back in the 70's a good scope for RF work was the one with the best X-Y-Z display. It was just a analog display for other test equipment.
Some of this test equipment used the scope sawtooth sweep output also.

One thing you do not see in a lot of radio equipment is test points added to make using test equipment easer with out loading the RF circuit.
For example, having a second or third buffered output to a mixer allows many things to be added in addition to test equipment.   

[MrSlack]

Yep definitely with respect to the XYZ situation. I've actually been looking at the SA's in amateur radio project literature that use a scope as a display and it's pretty achievable to knock one up that's good to 1GHz by the looks. They're not complicated at all. Difficult bits are VCO, IF filter and keeping noise and LO out. Probably £80 investment looking at the cost of log amps.

Nice full project description of one here: http://lea.hamradio.si/~s53mv/spectana/vco.html

[rfbroadband]

well in general it would be great to have both, but of course there is always the budget issue.

it looks like you want to stay below 500 MHz (70cm band). You could find scopes with good FFT functions that basically provide with a decent spectrum analyzer within a scope.

But a good 500MHz BW scope with a decent FFT function will cost ...:-(.

[AF6LJ]

I vote for getting a decent scope first, get your SA later.

[David Hess]

The oscilloscope will be useful for diagnosing other parts of your RF circuit like the power supplies and control signals and such and for that, your 465 will be fine so there is no reason to invest in another oscilloscope.

I *have* a high bandwidth oscilloscope which is quite capable of detecting harmonics but in practice, distortion levels are so low that they are invisible.