oscilloscope probe grounding

Thread Starter

petewh

Joined Mar 23, 2016
45
I've been using a 2 channel scope for low frequency work and have always grounding only one probe clip to the circuit under test. Now that I'm working with 1Mhz and greater circuits when I do that i get a lot of ringing in the second channel. Do I ground both oscilloscope ground clips to the circuit under test? I don't want to create a ground loop. I understand I would ground both clips to he same point. I've search a lot of material about probe grounding but do not find anything on using 2 channels and grounding.

Pete
 

R!f@@

Joined Apr 2, 2009
9,918
Welcome to AAC

You can use the probe ground as the ground.
At 1MHz, you need to make sure that the probe can measure that frequency.
 

Thread Starter

petewh

Joined Mar 23, 2016
45
Welcome to AAC

You can use the probe ground as the ground.
At 1MHz, you need to make sure that the probe can measure that frequency.
I understand that I use the ground probe as a ground but do I ground both probes to the same common point using a 2 channel scope?
Pete
 

Sensacell

Joined Jun 19, 2012
3,432
At higher frequencies you should strive for a ground connection that emulates a coaxial cable connection- like the attached solution.
Even the shortest length of ground clip can be a significant impedance at RF.
.pcb-probing-2.gif

 
Last edited:

OBW0549

Joined Mar 2, 2015
3,566
I've been using a 2 channel scope for low frequency work and have always grounding only one probe clip to the circuit under test. Now that I'm working with 1Mhz and greater circuits when I do that i get a lot of ringing in the second channel. Do I ground both oscilloscope ground clips to the circuit under test?
Yes, use both probe ground clips. The reason you're seeing ringing when you use only one probe ground is that the capacitance of the ungrounded probe and the inductance of the ground return path (from the ungrounded probe back to the 'scope, and from the 'scope back to the circuit under test via the grounded probe) form a series resonant LC circuit that can show substantial ringing when excited with sharp edges such as from logic signals.

Always use the shortest probe ground connections possible, for best viewed waveform quality.
 

crutschow

Joined Mar 14, 2008
34,280
Yes, you ground each probe.
And each should be near the point of measurement, not both to the same point.
The resulting ground loop should have little effect on the circuit performance.
 

KL7AJ

Joined Nov 4, 2008
2,229
I've been using a 2 channel scope for low frequency work and have always grounding only one probe clip to the circuit under test. Now that I'm working with 1Mhz and greater circuits when I do that i get a lot of ringing in the second channel. Do I ground both oscilloscope ground clips to the circuit under test? I don't want to create a ground loop. I understand I would ground both clips to he same point. I've search a lot of material about probe grounding but do not find anything on using 2 channels and grounding.

Pete
Not much danger of creating a ground loop, since both grounds go to the same instrument. Are you using the probes in the differential mode? If so, you may not to use EITHER ground clip. Experiment and see...you can't hurt anything (usually!)
 

ian field

Joined Oct 27, 2012
6,536
I've been using a 2 channel scope for low frequency work and have always grounding only one probe clip to the circuit under test. Now that I'm working with 1Mhz and greater circuits when I do that i get a lot of ringing in the second channel. Do I ground both oscilloscope ground clips to the circuit under test? I don't want to create a ground loop. I understand I would ground both clips to he same point. I've search a lot of material about probe grounding but do not find anything on using 2 channels and grounding.

Pete
If the device under test is also grounded, you can get an earth loop - if that causes a problem, you may have to power the target via an isolating transformer.

At higher frequencies, the length of the probe's ground lead is more likely to be a problem than whether you ground one or both.

Some professional probes have detatchable ground leads, I think you can get shorter ones that clip on.

Most probes have a grounded shell just back from the probe tip - if you can position that onto a convenient ground while taking a measurement it eliminates ground lead inductance.
 

Thread Starter

petewh

Joined Mar 23, 2016
45
Not much danger of creating a ground loop, since both grounds go to the same instrument. Are you using the probes in the differential mode? If so, you may not to use EITHER ground clip. Experiment and see...you can't hurt anything (usually!)
Thanks, I am starting to make differential mode measurements also. Can you clarify what you mean by "you may not to use EITHER ground clip"? I don't understand that.
 

Thread Starter

petewh

Joined Mar 23, 2016
45
Yes, you ground each probe.
And each should be near the point of measurement, not both to the same point.
The resulting ground loop should have little effect on the circuit performance.
Thanks, I did really wonder about using 2 ground clips. I'm surprised you say there can be 2 (slightly) different ground points. I am more worried about a small ground loop effect on the scope, not circuit performance. I don't want to damage the scope.
Pete
 

Thread Starter

petewh

Joined Mar 23, 2016
45
Why would you think connecting both ground clips to ground can damage an oscilloscope? They are already connected to ground and each other at the oscilloscope inputs.
It seemed to me that by creating ground loops from small potential differences on my ground would cause a problem. In regards to ground loops with regard to earth grounds and my circuits I've read it. Most my circuits have little potential differences on the ground line though when driving a display and an ADC on a protoboard I did measure 150mv to 200mv differences at various points. When I built a soldered version the differences reduced considerably.
 

#12

Joined Nov 30, 2010
18,224
So, you're afraid 200 millivolts is going to cook your scope?
Consider the impedance of the circuit you're working on. If it can provide several amps at 0.2 volts, it might warm up the ground leads.
How much current are you expecting?
 

Thread Starter

petewh

Joined Mar 23, 2016
45
So, you're afraid 200 millivolts is going to cook your scope?
Consider the impedance of the circuit you're working on. If it can provide several amps at 0.2 volts, it might warm up the ground leads.
How much current are you expecting?
I must say I don't have much experience of the internal workings of a scope. In the the circuit I had described I was using 3 dm74185 binary to bcd converters which each pulled 1/10 of an amp so potentially 350ma were on the return. To answer your question 100ma is what I'm at worst usually expecting though, seriously 20ma is more likely. If I am drawing say 1/2 an amp I have to consider the overall layout and what and separate supply lines and consider safety a little more. I did not know that the scope is grounded at the bnc inputs. That is I thought there could be a long ground run internal to the scope and cause some havoc to the scope's working.
However 0.2v and several amps warming the ground leads would be something I wish to avoid. I haven't built many circuits with a lot of ttl yet that may run up the amperage. While at the sacrifice of measurement I would think grounding both probes nearer each other may be a better idea. Do you disagree?
Pete
 

WBahn

Joined Mar 31, 2012
29,976
On most scopes the probe ground clips are electrically earth-referred via the safety ground of the scope's power cord. If the circuit you are testing is truly floating, then when you connect the scope probe's ground clip to the circuit you are establishing a connection between your circuit and earth ground. One thing that you need to be careful of is that you likely have several pieces of equipment, the scope, a power supply or three, a function generator, a computer or network connection, and it is very likely that more than one of these is earth-referred in some way. You need to ensure that these are all compatible. For instance, if your power supply is 12 Vdc and you split it on your board to get +/- 6 V supplies with an artificial "ground" midway, then if that 12 V supply is not floating and you connect your scope probe's ground clip to the circuit "ground" you will be in for a surprise.

Also, you can get significant ground loops unless all of your equipment is powered from the same building circuit. A good way to ensure this is to plug a single outlet strip into the wall and then plug all of your equipment into this single outlet strip.

Since both scope probes have their ground clips connected to the same point (whether or not earth-referred), you must make sure that both ground clips are connected to the same node in the test circuit. If they are not connected to the same point, you do risk causing a small ground loop. But you also want them to hit the ground plane very close to where the probe tip is placed. These are competing goals and you ability to strike a good balance depends on how well laid out the circuit it.

You cannot make differential measurements with most scopes by connecting the two probe tip to one point and the ground clip to the other -- you can only take to single-ended measurements and use the scope's functionality to mathematically emulate a differential measurement.
 

ian field

Joined Oct 27, 2012
6,536
Why would you think connecting both ground clips to ground can damage an oscilloscope? They are already connected to ground and each other at the oscilloscope inputs.
If you connect a properly grounded scope to a PSU with a bridge rectifier front end - it results in a fairly loud phutt.

Sometimes I've seen PCB tracks blown away, but it usually happens to the target equipment rather than the scope.
 
Top