EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: Mark on March 17, 2016, 10:35:19 am
-
I have been working on a circuit which cannot tolerate the extra loading of a 50 ohm spectrum analyzer input, so I found that a high-impedance probe (FET probe) is the answer to this. I built a little FET probe based on EDN's article "High-impedance FET probe extends RF-spectrum analyzer’s usable range", which works well. I just hook it up to the SA's USB port for power (+/- 5V are provided by a LTC3265IFE on my board).
http://m.eet.com/media/1130765/072105di.pdf (http://m.eet.com/media/1130765/072105di.pdf)
(https://www.eevblog.com/forum/projects/fet-probe-for-spectrum-analyzer-or-x10-scope-probe/?action=dlattach;attach=208919;image)
The test circuit is an amplifier at +/- 40V, so I am using my FET probe with a x100 scope probe. My question is therefore, for this application, can a simple x10 or x100 scope probe directly connected to the SA be used instead of the FET probe or are there downsides to this approach? The article linked above mentions DC components, but the Rigol DSA815TG is AC coupled.
-
That has a 15pF tip capacitance, which is obviously a frequency-dependent load. What is your working frequency and UUT's sensitivity to load?
If you need lower capacitance (<1pF) at your frequencies, consider a 500/1000/5000ohm Z0 "low impedance" probe.
See https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/ for references to DIY Z0 "low impedance" probe and others.
-
Mark, the X 10 and X 100 oscilloscope probes are designed to be connected to oscilloscopes with an input impedance at the BNC connector of 1 meghon // 20 pf or thereabouts. The X 10 probe can be thought of as a 10 meghom resistor in series with the oscilloscope input impedance. The X 100 probe will be a 100 meghom resistor (I am ignoring capacitances, of course).
The SA has a 50 ohm input impedance so a 200,000: 1 volt divider for the 10X probe and 2'000,000:1 volt divider for the 100X probe. Too much, I would think. Your FET probe looks like the ideal solution for your application.
-
Your FET probe looks like the ideal solution for your application.
You don't have sufficient information to say that.
In particular, it has an imput impedance of ~25ohms (i.e. <50ohms) at 500MHz. Z0 probes are usually preferable at these frequencies.
-
tggzzz, thanks for the clarification.
Mark says:
I have been working on a circuit which cannot tolerate the extra loading of a 50 ohm spectrum analyzer input, so I found that a high-impedance probe (FET probe) is the answer to this. I built a little FET probe based on EDN's article "High-impedance FET probe extends RF-spectrum analyzer’s usable range", which works well.
So we know that the circuit can not tolerate 50 ohm loading and that the FET probe works well. I interpreted that it works well on this particular application, indicating that the frequency dependent attenuation is not at issue. If the FET probe does not work well on this application I stand corrected.
-
Hello and thanks for the replies. The frequency of the amplifier is 500 kHz, sorry for not mentioning that in the OP. The harmonic contents relative to the carrier seem to match my oscilloscope's FFT function within 1dB, so it looks ok (mainly interested in 3rd harmonic).
-
Could have used a simple opamp in that case.