Simple RF measurement with a scope

Thread Starter

neddie

Joined Sep 29, 2011
7
Hi All.
What would be the best way to make a simple measurement of a 100MHz signal
with an oscilloscope.
I want to do some experimenting with some impedance matching circuits. I have a signal generator (50ohm)
and a scope (500MHz bandwidth). The problem that I am foreseeing is the loading effect of the scope
on the signal. The probe has a capacitance of 3.9pF , which is 397ohms @ 100Mhz.
If I'm trying to match 50 ohms to say 1K , how can I measure it without totally messing up the readings
by loading the circuit.
What would be the best way to tackle this with the available equipment?
Put some higher impedance voltage divider across the load and measure from that?
I'm sure there must be some standard way to tackle this issue :0)
Cheers
Rob
 

Thread Starter

neddie

Joined Sep 29, 2011
7
I just want to measure the basic response of the matching network. Measure the amplitude of the sine wave at various frequencies.
I'm just going to be doing some experimenting :0)
 

BR-549

Joined Sep 22, 2013
4,938
If you want to look at frequency, phase or quality(symmetry) of the signal...you may use a pick up coil nearby.

You can also frequency sweep the input......and watch the filter response with a rf probe on scope at output. Which will display bandpass of filter/circuit on scope.

A sweep generator is a marvelous tool. One can adjust filter/circuit response while watching in real time.
 

Thread Starter

neddie

Joined Sep 29, 2011
7
I'm just concerned with the scope loading the circuit. I don't have active probes , just normal passive probes.
 

ArakelTheDragon

Joined Nov 18, 2016
1,350
You should try with the osciloscope at the high impedance port of the PCB and see what will happen, just to test if you will get the same results.
 

Thread Starter

neddie

Joined Sep 29, 2011
7
Awesome ,thanks.
I was not looking to build an active probe , but this has piqued my interest :0)
Cheers
Rob
 

cariban

Joined Aug 14, 2018
69
Hi All.
What would be the best way to make a simple measurement of a 100MHz signal
with an oscilloscope.
I want to do some experimenting with some impedance matching circuits. I have a signal generator (50ohm)
and a scope (500MHz bandwidth). The problem that I am foreseeing is the loading effect of the scope
on the signal. The probe has a capacitance of 3.9pF , which is 397ohms @ 100Mhz.
If I'm trying to match 50 ohms to say 1K , how can I measure it without totally messing up the readings
by loading the circuit.
What would be the best way to tackle this with the available equipment?
Put some higher impedance voltage divider across the load and measure from that?
I'm sure there must be some standard way to tackle this issue :0)
Cheers
Rob
I would not use oscilloscope to measure the transfer function. The network analyzer is the right tool. You will get the full view of the matching network by measuring the S-parameters using network analyzer.
 

Janis59

Joined Aug 21, 2017
1,232
Danadak: thanks a lot, You economied my time rather much.
P.S. On the first link - PCB of Rogers(TM) - yes yes and more time yes. I had met it few years ago and am extremely happy with that thingy. Never even dont think about to use for GHz and kV anything else. For example, the best Rogers pcb types has tan(fi) o,oooo3 whilst ordinary FR4 the 0,15.
 

Janis59

Joined Aug 21, 2017
1,232
RE: ""not use oscilloscope to measure the transfer function. The network analyzer is the right tool.""
Yes, BUT.... it costs too much for amateur level work. Thus it may be substituted by cross-scale SWR meter, what allows to adjust the output filters and impedance matching circuit, as well maximize the power. Or sweep generator together with frequency analizer (however this may be rather expensive instrument, it is available in many not so expensive medium and high end USB osciloscopes and even in table DSO. Third and very nice possibility is to use a diy white noise generator (however at 2,4 GHz and above it is not very much trivial per se) plus SDR. Actually, at 2,4G there are only one choice available in markets, the HackRF for about 130 USD, because 10 USD cheap "dongle" capitulates exactly between 1,6 and 1,7 GHz. The best thing into is that You clearly see EVEN the parasythic (mirror) channels, both modulation widening, clearly see the signal strength, frequency etc etc with rather high precisity. However I may tell, the install process of drivers is real hell. Ir went more than a day to doing nothing as only this.
 

Janis59

Joined Aug 21, 2017
1,232
P.S. The VERY poor-man method from my teenager age, its a circular loop with microammeter, capacitor and two diodes. Practically everything needed for adjusting the transmitter output is possible to do with this highly robust and trustable instrument, but dont dream about comfort then.
 

cariban

Joined Aug 14, 2018
69
P.S. The VERY poor-man method from my teenager age, its a circular loop with microammeter, capacitor and two diodes. Practically everything needed for adjusting the transmitter output is possible to do with this highly robust and trustable instrument, but dont dream about comfort then.
Nice story! As a professional hardware/RF engineer, I have to use network analyzer to scan the whole band to make sure good matching in the band. For wideband signal, for example cable-TV 47MHz-862MHz, manual sweeping is not feasible.

Nevertheless I agree that for one frequency point or very narrow band, manually tuning may be possible just by monitoring the output amplitude.
 

Thread Starter

neddie

Joined Sep 29, 2011
7
just some simple L matching networks to start with. It's just for experimentation. A bit of a learning
experience. You can simulate this very easily , but it's not the same as actually building the stuff :0)
 

cariban

Joined Aug 14, 2018
69
just some simple L matching networks to start with. It's just for experimentation. A bit of a learning
experience. You can simulate this very easily , but it's not the same as actually building the stuff :0)
You said you want to match from 50 Ohm to 1000 Ohm. For such high ratio, it is better to use transformer. For example, you can use two 4:1 transformer from Mini-circuits to get nominal 800 Ohm impedance. Then use LC matching network for fine tuning.
https://ww2.minicircuits.com/pdfs/T4-2.pdf
 
Top