FET probes are usually designed to drive 50 Ohms, yes, but I wouldn't trust one to make any meaningful spectrum analyzer measurements. They just aren't good enough in terms of dynamic range and distortion to be of much use other than as a scope probe. OK, maybe you don't mind seeing harmonics and a raised noise floor that may not really be from the probed circuit.
Sure. You can also use a resistive (450 or 950 ohm) probe, which are cheap and very wideband. Spectrum analyzers have a lot of gain, so you can often see what you want with a 10:1 attenuation in the probe.
I'm not sure what you mean by that.
--
John Larkin Highland Technology, Inc
picosecond timing laser drivers and controllers
Of course the FET probe won't necessarily be that linear, since it's intended for use with scopes, so you may grow some spurs that aren't really there in the DUT. Also, spectrum analyzer inputs tend to be delicate--make very sure you don't put any DC there unless it's specified to be able to handle it.
I don't know about that. Tektronix does a good job at characterizing their active probes for flatness, linearity, and other properties. They act abo ut like you'd expect any other high-Z buffer to act.
Just for fun, I once measured the residual phase noise and AM noise of my P
6245. It had a flat broadband floor at around -150 dBc/Hz on both plots, w ith a 1/f corner around 10 Hz for PM and 100 Hz for AM. This was a 10x pro be, so that suggests that its input-referred noise figure would have been i n the neighborhood of 7 dB. (I also hooked it up to an NF meter and saw a similar figure at VHF but didn't bother recording any data.)
So the dynamic range from noise floor to clipping is at least competitive w ith the performance of most spectrum analyzers. Whether the probe's IMD pe rformance will let you take advantage of that, who knows... but I don't see any reason to suspect it's a problem.
Also, if you have an old-school analog scope it may have a vertical output jack on the back that will drive 50 ohms. Either way, use a DC block, obvi ously.
A 450 ohm resistor in series with a properly terminated 50 ohm coax makes a x10 probe, just the way the (9 Meg // 1.1 pF) and (1 meg // 10 pF) do in your average HiZ x10 probe.
The difference, as John L. says, is bandwidth.
Re: DC vulnerability--read the manual and find out for sure.
1 uF is too much for use with 0 dB of RF attenuation. It won't offer any protection against power transients. I'd use 0.1 uF at the very most (and at least 10 dB of attenuation.)
A few dB less attenuation, at the cost of loading down the circuit more. It's just a voltage divider, for better or worse.
First thing you need to do is disclose the HIGH end of the spectrum analyzer. What frequency ranges does your measurement require?
What level of input protection is provided by the analyzer? Putting a 1uF cap in series with a microwave spectrum analyzer and hitting it with a 9V battery can do some damage to your input mixer.
Probing is a very complex issue. The general case is a bitch. But, it's much easier in a controlled situation.
What are you trying to accomplish over what frequency range? If you're trying to take accurate measurements or calibrate something, you need to put some serious effort into fixturing. The thing you call the ground connection can have more serious implications than the details of the probe internals. Don't waste a lot of effort on probe fidelity if you can't control the ground connection. Capacitance at the input can overwhelm any resistance calculations you do. Tektronix had some app notes on probing back in the day. Google tektronix probe design for some info.
If you're poking around in a circuit to see what's broke, you can use most any old probing technique and learn what distortions it puts into your measurement. I'd worry more about how much damage you do to the circuit or your analyzer when you poke the probe at the wrong place.
Are you using the right tool for the job? Many oscilloscopes have built-in FFT capability...or software that can let your computer download the waveform and do it there.
Spectrum analyzers can be expensive and delicate. I'd never use mine to poke around in a live circuit. I want to know EXACTLY what's going in the hose before I connect it.
begin tangent... Two advantages to a spectrum analyzer are sensitivity and dynamic range. Take a tiny ferrite bead. Cut a slot in it...took me over a dozen tries to do it without busting the core...thinnest Dremel cutter was as thick as the hole diameter. Wind a few turns and connect it with zero lead length to the end of a 50 ohm coax to your spectrum analyzer. Justincase, keep some attenuation set in the input attenuator. You can use the probe to find high frequency signals in individual circuit board traces. Accuracy is crap, but often, that's not what you care about.
If you want 350MHz, I have got a deal for you! I designed this probe to do exactly what you're looking for, and I've tested it as good to 350MHz. Full design files here:
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.