how to place probe ground clips.

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

5v333

Well-known member
Joined
Jun 30, 2013
Messages
631
Location
Gothenburg
when i do sweeps while checking on scope i use two probes placing them
on each side of a part where i want to see freq/phase response and also check for resonances etc.
my generator goes to 40mhz, my scope goes to 100mhz so i check stuff up to 40mhz.
some times i get significant peaks in the rf range.
rearranging ground/screening, theese ringings can be dealt with and it clearly has effect on the sound.

however, depending on where i put the ground clips i get different response.
i usually use a ground buss point for the test point in question, so different ground points per probe, as long as the two test points doesnt share ground flow.

id like to get this right so i know when stuff is correct.

does my grouding makes sence?
should booth clips be grounded at all?
should they have the same ground point?

any tips are appretiated!!
 
carefully  ;D ;D ;D

I like to say ground is concept not a single voltage.

Place the ground clip (carefully) relative to what you are measuring.

JR

PS: It's spelled "bus".
 
ill try to remember the correct spelling further on!

my intuition tells me that when ever i put a ground clip in circuit i mess with ground flow.

so i try to use the last bus in a ground chain so that i dont mess with the flow through the whole system or at least that part of the circuit.

i just measured ohm between the ground clips and they are connected (wow sherlock!)
this also gives me an idea that two probes should perhaps either have the same ground point or just ground one of them...

 
5v333 said:
ill try to remember the correct spelling further on!
thank you,  a personal campaign of mine for decades...
my intuition tells me that whenever i put a ground clip in circuit i mess with ground flow.
In general the scope ground is just a minus input to the scope. With grounded chassis scopes that ground could be connected to outlet safety ground and corrupt ground to safety current flows. It is not uncommon to see 3-wire to 2-wire cheaters on scope line cords to lift that ground. It is also possible to inadvertently blow up sensitive circuitry by careless unintentional shorts to ground from a wayward ground clip. 
so i try to use the last bus in a ground chain so that i dont mess with the flow through the whole system or at least that part of the circuit.

i just measured ohm between the ground clips and they are connected (wow sherlock!)
this also gives me an idea that two probes should perhaps either have the same ground point or just ground one of them...
yes, using single ground clip is also common practice. In fact the ground clip can be removed from the probe to avoid risk of inadvertent shorts from a dangling ground clip touching something it shouldn't.

JR
 
would it be sane to somehow
- disconnect the screen of the probes, at the scope end
- use the clips for grounding/shielding in nearby of the testpoints
- and instead use ONE separate connection for ground reference between the unit in test and the scope?
 
5v333 said:
would it be sane to somehow
- disconnect the screen of the probes, at the scope end
- use the clips for grounding/shielding in nearby of the testpoints
- and instead use ONE separate connection for ground reference between the unit in test and the scope?

You could - if you wanted to go to all the bother - but I don't see you'd gain anything over using ground on one probe only.
Using one ground lead only has pros and cons. On one hand it removes a ground loop but on the other it makes the loop area of the test signal  that has the ungrounded probe larger and likely to ring and / or pick up noise. It really depends on the frequencies (and impedances) in question.
Try an ungrounded scope by all means but preferably a battery or DC powered model. Don't use a 'Cheater ' Adaptor !
Alternatively if you have a suitable isolation transformer you could try using that.
I guess JR might recommend something like his fault interrupting idea...
 
5v333 said:
would it be sane to somehow
- disconnect the screen of the probes, at the scope end
- use the clips for grounding/shielding in nearby of the testpoints
- and instead use ONE separate connection for ground reference between the unit in test and the scope?
I'd say if you're using two probes to look at different points in the circuit, you need both ground clips connected to some ground in the circuit. The ground you choose should be electrically "near" the signal the probe is connected to.

It's not so much "clearly having an effect on the ground" as the ground wire and clip having enough inductance to change what the scope input sees for a high-frequency signal. If you take the clip off the end of the probe, you'll see the pointy tip, and just back from that a metal ring. That ring is a "real" ground, and if you can connect that directly to a ground on your circuit while connecting the tip to a signal,  the scope will show a lot better representation of the signal. There's a picture of this sort of thing in the book "High Speed Digital Design," which is largely about designing PC boards and routing signals so high-speed electronics will work right.
 
But if i have two ground clips at two differnt locations, we loose isolation between the two ground loops.
Think transformer isolated circuits.


The ring behind the tip is connected to the rest of the probe ground/shield.. why is this a real ground?
 
5v333 said:
But if i have two ground clips at two differnt locations, we loose isolation between the two ground loops.
Think transformer isolated circuits.


The ring behind the tip is connected to the rest of the probe ground/shield.. why is this a real ground?

The ring ground is for probing very HF signals where the inductance of a wire clip is too much.

JR
 
The other way to do this, which is easy on a Tektronix scope, is to use a dual input vertical plugin and set the plugin to sum channel 1 and channel 2 and invert channel 2. Set channels 1 and 2 to the same sensitivity and you now have an ersatz differential probe. You can just probe two points and you get the voltage between then while grounding nothing. Typically though, your scope is grounded and the DUT is also grounded, perhaps through your lab bench supply or its own arrangement, so the circuit is not actually floating relative to the scope. But, you're removed of the task of clipping actual ground to a circuit node, causing possibly bad things to happen.

If you want to adjust the CMRR, probe the same point with both probes, set one of the inputs to UNCAL and adjust the gain to get no summed signal. While this is not as good as a real differential probe, it's usually available for free just by owning two probes of the same type and having a scope that can sum and invert channels, something that Tektronix has provided on many of its classic analog scopes.
 
Back
Top