power supply noise measurement with 1Mohms input + 50 Ohms feedthrough termination

Thread Starter

malli_1729

Joined Nov 28, 2012
32
Hi,

We need to measure power supply noise measurement for 0.82V,1.0V,1.2V,1.8V rails.

Setup will be:

50Ω, RF176 Co-axial probe directly soldered across output capacitor, configure oscilloscope input to 50Ω and with lowest vertical division to measure the ripple.

Low end scopes like DSOX3024A have offset ranges up to ±2V @ 50Ω, but new scopes can't offer more offset with 50Ω impedance. With AC coupling impedance is 1MΩ which is not matched.


I am planning to use MXR104A(this is common scope which i can use for other measurements also) to do the Power supply noise ripple measurement, But with 50Ω Input impedance offset ranges are as below

≤55 mV/div: ±0.8 V,
≤120 mV/div: ±1.6 V,
≤260 mV/div: ±3.2 V.

with 1MΩ input impedance

<10 mV/div: ±5 V,
≤200 mV/div: ±20 V.



I am thinking to use 1MΩ input impedance on MXR104A so, that i will have more OFFSET range of ±5 V, but to have 50Ω input impedance for impedance matching use a 50Ω feedthrough termination(Fluke TRM50 Coaxial Feed-through BNC Male to BNC Female Cable Terminator) before MXR104A input.

I don't have option to use high cost Power rail probes/differential probes, only this Option i have.

Do you see any issue with this setup for measurement of ripple.
 

Deleted member 115935

Joined Dec 31, 1969
0
Use the standard 10:1 probe,
use a very good ground connection,
NOT the bit of wire on the side of the 10:1 probe, but to the ground tip of the probe.

look at around 19 minutes into this
 

AlbertHall

Joined Jun 4, 2014
12,345
You may also get problems with a ground loop if the supply output is connected to mains earth, and the scope ground is also earthed.
I have been involved in setting up this kind of measuremnt and the supplies in question had earthed outputs and I found that I had to use an isolated differential measurement system.
Also the specifications usually specify a bandwidth for the measurement and you observe that bandwidth in the measurement system.
 

Thread Starter

malli_1729

Joined Nov 28, 2012
32
Use the standard 10:1 probe,
use a very good ground connection,
NOT the bit of wire on the side of the 10:1 probe, but to the ground tip of the probe.

look at around 19 minutes into this
No, I am not going to use 10:1 probe. controlling GND lead length(understand spiral loops etc) is hard, due to their 10:1 they SNR is very bad to do noise measurement in terms of few mV.
 

tautech

Joined Oct 8, 2019
383
1x probe, AC input coupling, 20 MHz BW limit and some averaging engaged.

These very low ripple measurements are better performed with a scope with 500uV/div or better sensitivity but controlling the background noise can be challenging.
 
Top