Author Topic: Oscilloscope earths.... Bah!  (Read 8273 times)

0 Members and 1 Guest are viewing this topic.

Offline Lee697Topic starter

  • Regular Contributor
  • *
  • Posts: 82
  • Country: au
Oscilloscope earths.... Bah!
« on: June 28, 2014, 08:11:15 am »
After causing a few sparks last week with a circuit and a mains earth-referenced power supply, I watched Daves video #279, and thought I had it beat!
So today, whilst using a battery powered circuit (12v 9Ah SLA), I obviously thought I could place the probes with impunity - fair enough, until you use the second channel.... what I hadn't realised at the time, was that after I connected channel 1, I now had a mains referenced circuit again.... a few sparks (and thankfully a working scope) and I'm again wiser!
Lee
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 28136
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Oscilloscope earths.... Bah!
« Reply #1 on: June 28, 2014, 08:46:33 am »
Yep, most of us have done it.  :-DD

Mains earth and input BNC outers are all connected as common.
We refer to the probe flying lead as a GND or Earth lead, but it is and it isn't.
We also have called it a GND lead to remind us that it is connected to mains GND.

But it is the reference for the applied signal firstly.

What is also often forgotten is this reference lead is commoned with all other BNC outers.
This can lead to a little confusion as you have discovered.

My best advice is if using multiple channels, remove all but 1 of the reference leads and carefully select where you place the 1 remaining reference lead with regard to the signal/signals you are trying to capture and any potential mains GND loops.

Another way around this minefield is an "Isolated" input scope.
Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline sean0118

  • Frequent Contributor
  • **
  • Posts: 453
  • Country: au
Re: Oscilloscope earths.... Bah!
« Reply #2 on: June 28, 2014, 09:07:56 am »
wow, I've never done that, but I would have sooner or later without reading this thread. Only really obvious after you consider it. Thanks for saving my scope OP.  :-BROKE  :-+
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: Oscilloscope earths.... Bah!
« Reply #3 on: June 28, 2014, 09:24:14 am »
Another way around this minefield is an "Isolated" input scope.

Or a high voltage differential probe on each (or just the 2nd) channel like I showed in the recent TV repair video.
 

Offline mij59

  • Frequent Contributor
  • **
  • Posts: 693
  • Country: nl
Re: Oscilloscope earths.... Bah!
« Reply #4 on: June 28, 2014, 09:43:58 am »
Or you could use an (medical) isolation transformer.

I ordered a Testec Si9001 differential probe  "only" 284 Euro's free shipping to the EU.
 

Offline Lee697Topic starter

  • Regular Contributor
  • *
  • Posts: 82
  • Country: au
Re: Oscilloscope earths.... Bah!
« Reply #5 on: June 28, 2014, 09:58:16 am »
My best advice is if using multiple channels, remove all but 1 of the reference leads and carefully select where you place the 1 remaining reference lead with regard to the signal/signals you are trying to capture and any potential mains GND loops.

I hadn't ever thought of that - only one of the GND clips needs to be attached doesn't it?
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: Oscilloscope earths.... Bah!
« Reply #6 on: June 28, 2014, 10:00:56 am »
Or you could use an (medical) isolation transformer.

An isolation transformer doesn't solve the 2 channel common ground problem.
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 28136
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Oscilloscope earths.... Bah!
« Reply #7 on: June 28, 2014, 10:05:19 am »
My best advice is if using multiple channels, remove all but 1 of the reference leads and carefully select where you place the 1 remaining reference lead with regard to the signal/signals you are trying to capture and any potential mains GND loops.

I hadn't ever thought of that - only one of the GND clips needs to be attached doesn't it?
Yep only 1.
There are several ways to "extract" the signals but all are more expensive than a little thought.
Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: Oscilloscope earths.... Bah!
« Reply #8 on: June 28, 2014, 12:16:40 pm »
I hadn't ever thought of that - only one of the GND clips needs to be attached doesn't it?

Yes. But the ground reference of the two signals is still the same. If you wanted to connect the two different grounds leads to two different places to begin with, then leaving one off won't fix your problem, because you obviously wanted to different ground references.
In this case you have to use either an isolated scope or at least one high voltage isolated diff probe. This is why I recommend having at least one of these probes.
 

Offline What_NZ

  • Regular Contributor
  • *
  • Posts: 150
  • Country: nz
Re: Oscilloscope earths.... Bah!
« Reply #9 on: June 28, 2014, 11:41:35 pm »
I read the OP a few times just to make sure I wasn't missing something. He didn't actually say he connected the two Ground (GND) leads to two different places in the circuit but that must have been what he did.

So he knew Channel 1's flying lead on the Scope Probe is connected to ground (GND) and to the Mains (he didn't say Mains Earth but I'm sure that's what he meant) but the alarming thing is, he didn't know (or realised) that Channel 2's flying lead on the Scope Probe was connected to the same ground (GND)..... Really?

I hope he realises now, that all those grounds (GND) are connected together and to the metal cover/chassis of the Oscilloscope (if it has one) and to the Mains Earth?

I would say if he doesn't, then he should put the Oscilloscope aside and read the Instruction Manual for the Oscilloscope, also a beginners guide to using Oscilloscopes.

Playing with Circuits and Test Equipment when you don't understand the limitations or the safety aspects of either, is dangerous.
Before you do anything always ask yourself "Do I understand what I am doing and if I do this, is it safe, what will happen?"

Until then, stick to your Multimeter and certainly don't try to test anything with High voltages or High currents.

Some people should have a warning label attached, which reads "I maybe dangerous to my own health"

And if you think I'm being a bit OTT, what could happen, if one day he thought powering his Oscilloscope from an Isolating Transformer (to achieve Mains isolation) was a good idea!        BTW you should never do this as it is a potential safety hazard, to you and anyone around you.

I realise this information may not pertain to people with years of knowledge but I don't believe the OP falls into that category.
« Last Edit: June 29, 2014, 03:45:00 am by What_NZ »
 

Offline ignator

  • Regular Contributor
  • *
  • Posts: 206
  • Country: us
Re: Oscilloscope earths.... Bah!
« Reply #10 on: June 29, 2014, 01:23:10 am »
I've been using O'scopes for 35 years. I've always felt they are designed wrong with the BNC ground connecting to chassis ground, which is then connected to safety earth ground via the power plug.
OK so the radio engineers invented these scopes, and then every thing was chassis ground reference. But it seems this continues as a paradigm. Maybe there should be 2 types of scopes the current ones that RF folks use, and then the correct type where every input is differential, and isolated from any of the other channels.
This would require using the probes coax ground connected at the point of interest as a reference (to a differential input). The scope could have switches to keep the ground references isolated or connected to any one of the probes that has the ground of interest connected. The user still has to keep track of what the ground reference is connected too, to insure they don't cause a fault current to flow through the scope and probe returns (or automatic prevention of this).

Or maybe I'm just ignorant to how these should be designed, but I'm blaming the radio engineers, and the copy cats that continue this design error.
 

Offline ejeffrey

  • Super Contributor
  • ***
  • Posts: 3683
  • Country: us
Re: Oscilloscope earths.... Bah!
« Reply #11 on: June 29, 2014, 01:40:21 am »
Or maybe I'm just ignorant to how these should be designed, but I'm blaming the radio engineers, and the copy cats that continue this design error.

I don't see why this is that big of a deal.  You want a differential measurement you use a differential probe or a current transformer.  For the vast majority of troubleshooting the common ground systems work fine.  People also have different ideas of what they want when they talk about that.  Most manufacturers do have a line of isolated input scopes (frequently battery powered as well) for power supply / automotive applications which are the most common use case for this.  BNC is also not the greatest choice for this as the shield is exposed.  That means that a true floating input is potentially dangerous.  Isolated input scopes and probes use BNCs with plastic shells but they are less robust than the metal ones. 
 

Offline What_NZ

  • Regular Contributor
  • *
  • Posts: 150
  • Country: nz
Re: Oscilloscope earths.... Bah!
« Reply #12 on: June 29, 2014, 01:42:42 am »
I've been using O'scopes for 35 years. I've always felt they are designed wrong with the BNC ground connecting to chassis ground, which is then connected to safety earth ground via the power plug.
OK so the radio engineers invented these scopes, and then every thing was chassis ground reference. But it seems this continues as a paradigm. Maybe there should be 2 types of scopes the current ones that RF folks use, and then the correct type where every input is differential, and isolated from any of the other channels.
This would require using the probes coax ground connected at the point of interest as a reference (to a differential input). The scope could have switches to keep the ground references isolated or connected to any one of the probes that has the ground of interest connected. The user still has to keep track of what the ground reference is connected too, to insure they don't cause a fault current to flow through the scope and probe returns (or automatic prevention of this).

Or maybe I'm just ignorant to how these should be designed, but I'm blaming the radio engineers, and the copy cats that continue this design error.

As someone else said - this is when you have to use differential probes. I don't have a differential probe and never needed one.

I would be surprised if a home enthusiast needed one, then again for some people on the forum, money appears to be no object and it would look nice sitting next to their other flashy test gear. Then there are the people who actually need and use them...........
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16545
  • Country: us
  • DavidH
Re: Oscilloscope earths.... Bah!
« Reply #13 on: June 29, 2014, 03:57:29 am »
Another way around this minefield is an "Isolated" input scope.

Or a high voltage differential probe on each (or just the 2nd) channel like I showed in the recent TV repair video.

Or 2 channels configured in add and invert mode to make 1 differential input.  If you have a 4 channel oscilloscope, then that may allow 2 differential inputs.  The ground leads on the probes may be connected together or to the signal ground.  It is not as good as a high voltage differential probe but it is better than nothing.
 

Offline edy

  • Super Contributor
  • ***
  • Posts: 2385
  • Country: ca
    • DevHackMod Channel
Re: Oscilloscope earths.... Bah!
« Reply #14 on: June 29, 2014, 04:16:24 am »
I am trying to understand the scenarios in the original post that caused the sparking. For the first, mains referenced circuit power by supply, if probing across portions of the circuit you use 2 channels and attach ground clip for both channels to the same ground point on the device (which you confirm with continuity test to short with ground pin on power cable) and place tips of your 2 probes across the circuit element you are checking? And just do a subtraction of one channel from the other to get the differential? Did the OP connect probe tip and ground of 1 channel across some portion of the circuit, not realizing the place they connected ground was not ground in the circuit and had a potential they shorted? What if you don't connect the ground of your probe at all and only probe with the tip?

With second scenario, a battery-powered circuit was being probed. How do you spark with only 1 channel? There is no loop. When OP used the second channel, it had its ground connected to a different point than the ground for channel 1, which again created a short through the scope from those two non-equipotential points in the circuit? Is that what happened?
YouTube: www.devhackmod.com LBRY: https://lbry.tv/@winegaming:b Bandcamp Music Link
"Ye cannae change the laws of physics, captain" - Scotty
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 28136
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Oscilloscope earths.... Bah!
« Reply #15 on: June 29, 2014, 04:54:22 am »
What if you don't connect the ground of your probe at all and only probe with the tip?

With second scenario, a battery-powered circuit was being probed. How do you spark with only 1 channel? There is no loop. When OP used the second channel, it had its ground connected to a different point than the ground for channel 1, which again created a short through the scope from those two non-equipotential points in the circuit? Is that what happened?
The reference lead is sometimes needed to ensure signal integrity from interference. Only by trying it would one know. But yes in an EMI clean environment often you can use a probe without a lead if the POI is mains ground referenced

Second scenario yes this is what happened.

Now I want to have a bitch at all posters that refer to the reference lead as a GND lead.
With scopes that have BNC's at mains GND potential you might think this way but if you get to use an isolated channel scope it is referred to as a Reference lead.

Get this terminology in your heads.  :rant:

BTW a important use for an isolated channel scope is 3 phase where 3 differential probes would be needed or commonly 4 isolated channels.
Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16545
  • Country: us
  • DavidH
Re: Oscilloscope earths.... Bah!
« Reply #16 on: June 29, 2014, 05:35:57 am »
I am trying to understand the scenarios in the original post that caused the sparking. For the first, mains referenced circuit power by supply, if probing across portions of the circuit you use 2 channels and attach ground clip for both channels to the same ground point on the device (which you confirm with continuity test to short with ground pin on power cable) and place tips of your 2 probes across the circuit element you are checking? And just do a subtraction of one channel from the other to get the differential? Did the OP connect probe tip and ground of 1 channel across some portion of the circuit, not realizing the place they connected ground was not ground in the circuit and had a potential they shorted? What if you don't connect the ground of your probe at all and only probe with the tip?

A differential probe measures the difference between to signals with respect to the probe ground.  If the probe ground is not connected either directly or through the oscilloscope ground, then it may float possibly exceeding the common mode range of the differential probe.

Quote
With second scenario, a battery-powered circuit was being probed. How do you spark with only 1 channel? There is no loop. When OP used the second channel, it had its ground connected to a different point than the ground for channel 1, which again created a short through the scope from those two non-equipotential points in the circuit? Is that what happened?

There would not be unless the grounds from 2 channels were connected to different points or unless the battery-powered circuit had some other external ground connection like through a USB port.

Usually the problem with off-line circuits is that the neutral connects to ground back at the distribution panel but the current through the neutral raises its potential compared to ground so when you connect the oscilloscope probe ground to neutral, considerable current may flow or all of it if the neutral somehow gets disconnected.  Even if this does not damage something, the ground loop current will corrupt the probe signal.
 

Offline What_NZ

  • Regular Contributor
  • *
  • Posts: 150
  • Country: nz
Re: Oscilloscope earths.... Bah!
« Reply #17 on: June 29, 2014, 05:39:17 am »

Now I want to have a bitch at all posters that refer to the reference lead as a GND lead.
With scopes that have BNC's at mains GND potential you might think this way but if you get to use an isolated channel scope it is referred to as a Reference lead.

Get this terminology in your heads.  :rant:

BTW a important use for an isolated channel scope is 3 phase where 3 differential probes would be needed or commonly 4 isolated channels.

Huh? bitch all you like  ;)

I think of it as. What potential am I going to make that point of the circuit when I connect the fly lead to it. I can then rationalise if that is the right or safe thing to do.

<Quote> "With scopes that have BNC's at mains GND potential you might think this way" No, you have to think this way and because the OP didn't think this way he got into trouble, twice!

I have never used an isolated channel scope so I never knew it was called a "Reference lead",  thanks for that.
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 28136
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Oscilloscope earths.... Bah!
« Reply #18 on: June 29, 2014, 06:13:47 am »
Huh? bitch all you like  ;)

I think of it as. What potential am I going to make that point of the circuit when I connect the fly lead to it. I can then rationalise if that is the right or safe thing to do.

<Quote> "With scopes that have BNC's at mains GND potential you might think this way" No, you have to think this way and because the OP didn't think this way he got into trouble, twice!

I have never used an isolated channel scope so I never knew it was called a "Reference lead",  thanks for that.
Consider this fact.
It is only an oscilloscope probe and can be connected to any BNC input device whether or not the BNC is ground referenced.
On a BNC earthed input device this thinking might be considered adequate, but if a device is not Earth/GND-ed it is a reference lead. Pure and simple.

The scope or the probe is not the problem, it is the way they are used.

To add further confusion, often circuit GND is not mains GND and in this case it one can connect ones reference lead nearly anywhere after a quick meter test to check it is safe.

Knowledge and understanding is the key.
How many out there commonly have 100:1 probes?
How many out there know how to down-rate the max voltage of their probes with frequency?
There are many ways to acquire the signals, not all are considered safe in this forum, nevertheless are commonly done by techs that know their stuff and the risks.
Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline VK5RC

  • Supporter
  • ****
  • Posts: 2672
  • Country: au
Re: Oscilloscope earths.... Bah!
« Reply #19 on: June 29, 2014, 07:11:50 am »
Whilst not absolutely on thread,  as a ham interested in the GHz area,  when I test home brew equipment I often have many bits of gear out at the moment,  before I power up,  I often spend quite a time checking the specs of the gear,  attenuators,  splitters,  power levels etc.  It is amazing how fewer pieces of gear have the basic specs in an area you can get to easily eg on the outside cover,  digging in the cupboard for the manual is such a pain.
For me these fundamental specs should be in plain sight.
With regard to the OP we all stuff up,  hopefully not too seriously or too often,  but the really important part is to learn from them. :-)
Whoah! Watch where that landed we might need it later.
 

Offline Lee697Topic starter

  • Regular Contributor
  • *
  • Posts: 82
  • Country: au
Re: Oscilloscope earths.... Bah!
« Reply #20 on: June 29, 2014, 07:14:23 am »
With second scenario, a battery-powered circuit was being probed. How do you spark with only 1 channel? There is no loop. When OP used the second channel, it had its ground connected to a different point than the ground for channel 1, which again created a short through the scope from those two non-equipotential points in the circuit? Is that what happened?

Yep. The first channel was being used to measure the voltage drop across a resistor, so the GND/reference wasn't at zero. This was fine. I then thought I would measure the voltage drop at the battery at the same time - a spark occurred when the GND/reference clip touched the neg terminal of the battery.
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 28136
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Oscilloscope earths.... Bah!
« Reply #21 on: June 29, 2014, 09:59:26 am »
Whilst not absolutely on thread,  as a ham interested in the GHz area,  when I test home brew equipment I often have many bits of gear out at the moment,  before I power up,  I often spend quite a time checking the specs of the gear,  attenuators,  splitters,  power levels etc.  It is amazing how fewer pieces of gear have the basic specs in an area you can get to easily eg on the outside cover,  digging in the cupboard for the manual is such a pain.
For me these fundamental specs should be in plain sight.
With regard to the OP we all stuff up,  hopefully not too seriously or too often,  but the really important part is to learn from them. :-)
Quote
as a ham interested in the GHz area
:wtf:  :o  Black Magic stuff.  ;)
Quote
It is amazing how fewer pieces of gear have the basic specs in an area you can get to easily eg on the outside cover,  digging in the cupboard for the manual is such a pain.
For me these fundamental specs should be in plain sight.
I have many scope probes and I keep all the manuals/de-rating graphs within arms reach of my scope work station, for good reason. You can't put all that info on a probe.
The problem is that many 10:1 probes are only good for their(if your'e lucky) rated 600 V @ ~18 KHz and only ~20 V @ 100 MHz while a 100:1 probe 1.2 KV @ 2 MHz and ~150 V @ 250 MHz.
At risk is yourself, your scope and the probe. Never mind the DUT!
This is stuff we all must know about and have at hand if needed.
And don't get me started on scope input Vmax p-p ratings, I will go on for hours.  :rant:
I ask you all how long did it take before you were a competent driver, now apply at least a  similar time to scope use in various situations.
Be careful out there!
Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline VK5RC

  • Supporter
  • ****
  • Posts: 2672
  • Country: au
Re: Oscilloscope earths.... Bah!
« Reply #22 on: June 29, 2014, 11:03:07 am »
I still don't think of myself as fully competent driver of most electronic equipment ,  I still stuff up even using a dmm from time to time! ,  especially if rushed stressed etc. Daves test of putting volts into ohms mode is an indicator that I am not alone :-)
i agree you can't get all the specs on the bits but a few of the basics would be really nice even if only a prompt. More use than some logo.
Whoah! Watch where that landed we might need it later.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf