Maker Pro
Maker Pro

Common mode noise questions

M

Mook Johnson

Folks,

I'm ealing with a switching power supply and when I connect the scope probe
ground to the board ground, the scope get noisy even before I touch the
probe tip to the signal I'm trying to measure. It not my first time running
into the problem and I'd like suggestions on scope hookups that would not be
as suseptable to this noise.

I've trised taking the ground wire off the probe and using a short piece of
buswire wrapped around the ground probe near the tip. It helps some but I
still get noise when I touch probe ground to PSU ground.

I've also tried floating the scope with little success.

Is a differential probe the only way around this?

All of the text books say itscommon mode noise and to use a common mode
noise filter to prevent this noise from radiating in the wires to the unit.
but that essentially traps the noise in the part I want to measure.

Any tips?
 
D

D from BC

Folks,

I'm ealing with a switching power supply and when I connect the scope probe
ground to the board ground, the scope get noisy even before I touch the
probe tip to the signal I'm trying to measure. It not my first time running
into the problem and I'd like suggestions on scope hookups that would not be
as suseptable to this noise.

I've trised taking the ground wire off the probe and using a short piece of
buswire wrapped around the ground probe near the tip. It helps some but I
still get noise when I touch probe ground to PSU ground.

I've also tried floating the scope with little success.

Is a differential probe the only way around this?

All of the text books say itscommon mode noise and to use a common mode
noise filter to prevent this noise from radiating in the wires to the unit.
but that essentially traps the noise in the part I want to measure.

Any tips?

How did you float the scope?
Whatzdo...block the ground prong? Geez... :(

I've looked at diff probes for smps work..
After seeing prices*, I opted to design my own active probe using a
video amp and a ferrite xformer.

*Anybody see Chinese diff probes at $5.00 each..???


D from BC
British Columbia
Canada.
 
T

Terry Given

Mook said:
Folks,

I'm ealing with a switching power supply and when I connect the scope probe
ground to the board ground, the scope get noisy even before I touch the
probe tip to the signal I'm trying to measure. It not my first time running
into the problem and I'd like suggestions on scope hookups that would not be
as suseptable to this noise.

I've trised taking the ground wire off the probe and using a short piece of
buswire wrapped around the ground probe near the tip. It helps some but I
still get noise when I touch probe ground to PSU ground.

this is a good start.
I've also tried floating the scope with little success.

Is a differential probe the only way around this?

nope, that probably wont help either.
All of the text books say itscommon mode noise and to use a common mode
noise filter to prevent this noise from radiating in the wires to the unit.
but that essentially traps the noise in the part I want to measure.

Any tips?

download & read linear technology AN47. then make sure your scope probe
has a BNC tip adaptor. I use tightly twisted wire to a BNC connector,
then the probe+adaptor. gets rid of almost all measurement noise.

Cheers
Terry
 
J

John Larkin

Folks,

I'm ealing with a switching power supply and when I connect the scope probe
ground to the board ground, the scope get noisy even before I touch the
probe tip to the signal I'm trying to measure. It not my first time running
into the problem and I'd like suggestions on scope hookups that would not be
as suseptable to this noise.

I've trised taking the ground wire off the probe and using a short piece of
buswire wrapped around the ground probe near the tip. It helps some but I
still get noise when I touch probe ground to PSU ground.

I've also tried floating the scope with little success.

Is a differential probe the only way around this?

All of the text books say itscommon mode noise and to use a common mode
noise filter to prevent this noise from radiating in the wires to the unit.
but that essentially traps the noise in the part I want to measure.

Any tips?

Get a hunk of coax with a BNC on one end. Plug that into the scope.
Solder the other end into the power supply, braid to ground and center
to the thing you want to measure. Clamp a couple of those ferrite EMI
things along the way.

That should be better than a probe.

John
 
J

Jamie Morken

John said:
Get a hunk of coax with a BNC on one end. Plug that into the scope.
Solder the other end into the power supply, braid to ground and center
to the thing you want to measure. Clamp a couple of those ferrite EMI
things along the way.

That should be better than a probe.

Hi,

I just noticed this ground noise for the first time yesterday with my
USB scope, I touched the ground clip of the probe to the PCB ground of a
switching power supply, and wow tons of noise, even when I put the probe
tip and the ground clip both to ground. So this is common mode noise?

Why does the cut coax perform better than a probe in this case?

Is this only a problem with circuit ground?

cheers,
Jamie
 
P

Paul Mathews

Hi,

I just noticed this ground noise for the first time yesterday with my
USB scope, I touched the ground clip of the probe to the PCB ground of a
switching power supply, and wow tons of noise, even when I put the probe
tip and the ground clip both to ground.  So this is common mode noise?

Why does the cut coax perform better than a probe in this case?

Is this only a problem with circuit ground?

cheers,
Jamie






- Show quoted text -- Hide quoted text -

- Show quoted text -

The 'noise' associated with probe grounding near the tip is not common-
mode. It is differential mode voltage induced in the loop at the end
of the probe. In addition, the associated inductance will cause
ringing (overshoot, undershoot) artifacts. The way to eliminate it is
to make the loop as small as possible. Very short ground leads help.
The referenced Linear Tech App Note is a must-read. You can also twist
the ground lead around the probe tip. One way or the other, you must
collapse or otherwise neutralize the loop to eliminate this source of
differential noise. Diff probes won't help at all unless you collapse
the loop.
There will also be some common mode 'noise' associated with observing
high dv/dt waveforms with a typical oscilloscope. The scope's Common
Mode Rejection Ratio is finite, and various parasitic and EMI-filter
capacitances carry currents through the grounding system that
aggravate the problem. Snap-on ferrite clamps placed over the scope
probe coax can be a big help here, provided that the 'noise' spectrum
is mostly above 50 MHz or so. If you choose a large ferrite, you can
thread several turns of the scope coax through it for more benefit.
Like any good common mode filter, this should have very little effect
on the underlying signal, which is easily verified.
Paul Mathews
 
Top