Did I ruin my oscilloscope?

[tautech]

My understanding of the scope inputs was that they are isolated, (-) independant from each other, and not even connected to ground. I was expecting that the scope measures the potential difference between (+) and (-). As it doesn't I will have to use a differential probe.
Differential probe is just safest, and in fact easier to understand, as I just want to measure the potential difference. It's a shame scopes aren't designed like that by default.

It's all about cost.
In an Isolated channel scope, not only are the BNC shells isolated from Mains ground, also from each other.
Scopes constructed in this manner are 4 Ch and often designed for use on 3 phase systems and therefore MUST have the appropriate CAT ratings as well.
Non-isolated scope design elimimates much cost but presumes the user has the knowledge to use a scope safely.

You also must refrain from thinking of signals in terms of + for probe and - for reference(GND).
If you have a negative voltage signal then the reference(Gnd) is more positive with respect to the signal.
Think probe TIP and Reference CLIP (Gnd)

[vk6zgo]

Here is a stupid question: If you always have to connect the tip to ground, as your (-) at the scope side is connected to ground, why do you need the tip at all, as you are always measuring the potential difference between your signal and ground, and the scope is connected to ground already?

I actually have seen once someone measuring with a scope, and he did not connect the tip. I found it rather funny, and was pointing out that he is measuring a potential difference, and therefore needs to connect both. But from the conclusions in this posting, my view on reality was wrong, as the scope seems to measure always referenced to ground. So now I am actually in doubt myself, and believe that the guy doing this measurement without tip, might not have been an idiot after all, or do I miss out something here =)

As tautech has pointed out,you have your terminology backwards.
The "tip" is the end of the probe which you attach to the point you wish to look at the voltage on,the "clip",or more correctly "earth clip"is the thing you attach to the "common" point of the circuit.

Most equipment contains internal power supplies.

Let us consider a device with +12v ,-12v,+5v,-5v supplies,all with one side of the supply connected to the common "rail".
This "common rail" is connected to the chassis frame/cabinet .

Forgetting the 'scope for a second,if you take a DMM,place its black lead on the chassis,& touch the red lead to the test points for the above supplies in succession,-----you will see +12v,-12v,etc.

Let us assume that you have your 'scope sitting alongside this other device,all plugged in & ready to go.
Now take your DMM & put the black lead on the outer part of the "scope's BNC connector.

Check your voltage test points again.
You will see:-+12v,-12v,etc!

What has happened?
 This device,as is common in much non-domestic Electronic equipment not only returns one side of the power supplies to the chassis,but also connects the Mains safety Earth to the chassis.

An experienced Tech,looking at the dc supplies in such a device,may set his "scope  to "DC coupled" at,say,5v/div,'free run" the display,centre the line vertically on the screen,& touch the probe tip to the dc supply test points,so the horizontal line will deflect +2.4 large divisions for +12v,-2.4  for -12v,+ one large div for +5v,etc.

Ideally,he would always attach the "earth clip" to the chassis/common rail,but for a quick check it is not unusual to rely upon the common connection via the Mains Earth system.

It is normal practice when using a 'scope to measure everything with respect to the common rail,which in mains operated equipment is normally Earth/chassis.
Measuring across a circuit with each side above Earth is seldom necessary,but can be achieved either using a purpose designed differential probe,or by the following method.

Thinking back to our device with the multiple supplies,what if we wanted to measure the difference in voltage between the +12v & -12v supplies?(I mean look at the complete 24v)

OK,we place both,(or only one--it doesn't matter) our "earth clips"on the common/rail/chassis(we shouldn't rely on the Mains Earth connection here,as it may introduce errors).

Connect CH1 probe to one supply,CH2 to the other.

You will,with 5v /div on each channel,see two lines 4.8 large divisions apart.
Make one line the reference,& you will read off + or - 24 volts depending on which you chose.

Perhaps the -ve & +ve lines need to be the same numerical value--you can then set the 'scope to "A+ B",& it will display the difference in that value.

I would suggest you get a couple of "AA" batteries & have a play around in the "dc coupled" mode.

[Andreas]

Hello Rene,

The things that you should test (with mains line disconnected from outlet).
- is the  minus line from CH1 connected to earth pin from mains (near zero ohms)
- is the  minus line from CH2 connected to earth pin from mains (near zero ohms)

(in this case also CH1 and CH2 are still connected together)

The connections have to be for security reasons because the minus lines can be touched by operator.
(And you would not expect to have 110 or 230 V on the housing of the osciloscope):

With best regards

Andreas

[miguelvp]

Using the ground clip to measure mains is a big NO NO. You'll be lucky if the only thing damaged is your scope.

Ground is earth, not minus and not negative.

Also it's AC so there is no minus. And even if English is not your language, minus means the same as in English and it's not ground/Earth.

[tautech]

Using the ground clip to measure mains is a big NO NO. You'll be lucky if the only thing damaged is your scope.

Ground is earth, not minus and not negative.

Also it's AC so there is no minus. And even if English is not your language, minus means the same as in English and it's not ground/Earth.

+1
I can't beleive so many refer to scope inputs and probes in this manner. 

There is ONLY Signal input and it's Reference, commonly GND(but not always)
No positive or negative inputs.... however the Signal might be positive or negative or BOTH and not AC. 

Where does this misunderstanding of scope inputs come from?
DMM's?

[idpromnut]

Using the ground clip to measure mains is a big NO NO. You'll be lucky if the only thing damaged is your scope.

I think Andreas was trying to say that the OP should check that the grounding on each channel is still working (i.e. with the scope not plugged into the wall, and checking the scope's mains ground pin against each channel's ground clip lead for low resistance).

Doesn't invalidate your comment; be careful when you stick anything into the mains and know your equipment before doing so!

[David Hess]

Almost certainly didn't break anything.
All the current flowed through your BNC shield connectors on the front, and that would have been limited by the series resistor on top.
Unless you are dealing with a massively high powered system, your scope will be fine.

It takes surprisingly low current to turn the micro-grabber style of probe ground lead white hot before melting it.  I only had to make that mistake once and the oscilloscope was not harmed.

[pascal_sweden]

Regarding the isolated front end option from Siglent:
http://www.siglent.eu/isolated-front-end.html

What is the deal with the bandwidth and peak to peak voltage?
Does this mean that your scope does no longer measure the original signal with the original bandwidth provided by the scope and at the right voltage levels?

I don't understand what they mean with 1 MHz bandwidth and 3 Vpp. Can someone explain this, as I plan to buy this to make sure I don't blow up my scope in the first week of usage

[Nerull]

It has a 200:1 attenuation ratio - a 600V input signal is output as 3V. These things are meant for protecting your scope from high voltages.

When measuring high speed signals, you have to consider how everything in the signal path effects the signal. Your scope has a maximum bandwidth rating, your probes have a maximum bandwidth rating, random pieces of coax have maximum bandwidth. This generally means that signals over a certain frequency are attenuated or distorted. This particular piece of equipment is rated for 1MHz. If you use it on anything higher than that you may not get good results, even if you hooked up to one of those 63GHz Agilent monsters. It's pretty rare to see high voltage and high frequencies together, and this thing isn't really intended for that.

[pascal_sweden]

Okey, then I have to look for another device.

I was planning to use this as an isolated front-end, to make sure that the clip on the probe is floating, and not connected to ground.

I want to have a system where I can connect the probe tip and clip to any point in my electronic design, without any risk of making a short circuit. I want to make an objective measurement of the potential difference between the two measure points (the tip and the clip), without affecting the signals in my electronic design, and without affecting the functionality of my scope.

If I understand well, this device from Siglent does it, but at same time provides attenuation, and limits the bandwidth to 1 MHz (at least does not affect signal characteristics below 1 MHz).

What I am looking for is basically the same device, but without the attenuation and without affecting the bandwidth. Just the isolation of the input signals, to make sure that the clip of my probe can not cause a short circuit.

Basically I am looking for an adapter (similar form factor as the Siglent adapter) which turns the standard Rigol probe (as delivered with the scope) into a differential probe, when I put it in between my scope and my probe. Does such an adapter exist?

I want to use my standard Rigol probe as a differential probe, where I can connect the tip and the clip to any measure point without causing a short circuit. This should even apply when measuring with the 2 channels at the same time. I want that the clip of probe 1 is independant of the clip of probe 2.

[tautech]

Okey, then I have to look for another device.

I was planning to use this as an isolated front-end, to make sure that the clip on the probe is floating, and not connected to ground.

I want to have a system where I can connect the probe tip and clip to any point in my electronic design, without any risk of making a short circuit. I want to make an objective measurement of the potential difference between the two measure points (the tip and the clip), without affecting the signals in my electronic design, and without affecting the functionality of my scope.

If I understand well, this device from Siglent does it, but at same time provides attenuation, and limits the bandwidth to 1 MHz (at least does not affect signal characteristics below 1 MHz).

What I am looking for is basically the same device, but without the attenuation and without affecting the bandwidth. Just the isolation of the input signals, to make sure that the clip of my probe can not cause a short circuit.

Basically I am looking for an adapter (similar form factor as the Siglent adapter) which turns the standard Rigol probe (as delivered with the scope) into a differential probe, when I put it in between my scope and my probe. Does such an adapter exist?

I want to use my standard Rigol probe as a differential probe, where I can connect the tip and the clip to any measure point without causing a short circuit. This should even apply when measuring with the 2 channels at the same time. I want that the clip of probe 1 is independant of the clip of probe 2.

You will have 4 options.
Buy an isolated input DSO.
Buy differential probes.
Buy a 4 Ch scope and use std probes without Reference(Gnd) clips for a 2 Ch result
Forget the idea and learn how to use a scope safely.

[pascal_sweden]

You will have 4 options.
Buy an isolated input DSO.
Buy differential probes.
Buy a 4 Ch scope and use std probes without Reference(Gnd) clips for a 2 Ch result
Forget the idea and learn how to use a scope safely.

Regarding the proposed options:
Option 1) Not an option, I want to stick with MSO2072A
Option 2) Where to get affordable/qualitative/reliable differential probe for MSO2072A?
Option 3) Do you mean that you then will use the tips only and use math operations from the scope to combine tip1 and tip2 for the 1st channel, and tip3 and tip4 for the 2nd channel? Nice suggestion actually.
Option 4) Even if you know what you are doing, you might make mistakes (e.g. on a Monday morning)

[Marco]

Using the ground clip to measure mains is a big NO NO. You'll be lucky if the only thing damaged is your scope.

Why? You might burn the ground clip I guess, but the rest of the path to ground should be chunky enough to survive till the fuse blows. The ground getting raised a bit shouldn't affect the circuitry either AFAICS.

[tautech]

Using the ground clip to measure mains is a big NO NO. You'll be lucky if the only thing damaged is your scope.

Why? You might burn the ground clip I guess, but the rest of the path to ground should be chunky enough to survive till the fuse blows. The ground getting raised a bit shouldn't affect the circuitry either AFAICS.

Depends on how big the fuse is.
The current return path is via the GND tracks on the scope PCB.
You can vaporize those tracks with enough current/energy..

[Marco]

The current return path is via the GND tracks on the scope PCB.

Isn't the BNC simply soldered to the ground plane, which is connected to the chassis at each standoff?

[David Hess]

Using the ground clip to measure mains is a big NO NO. You'll be lucky if the only thing damaged is your scope.

Why? You might burn the ground clip I guess, but the rest of the path to ground should be chunky enough to survive till the fuse blows. The ground getting raised a bit shouldn't affect the circuitry either AFAICS.

As a concrete example consider when working on the primary side of a switching power supply with the ground clip connected to neural which is a bad idea but accept it now for the sake of argument.  This creates a ground loop through the oscilloscope back to the distribution panel but the relatively high resistance of the ground lead connection limits the current compared to the return current through the neutral.  Now if the neutral fails for whatever reason, *all* of the neutral current is going to go back through the oscilloscope ground and if the neutral failed because of a short, that could place the power line AC voltage across the oscilloscope from the probe ground to the earth ground.

This is why I tend to use an isolation transformer even when using a differential probe.

[David Hess]

Option 2) Where to get affordable/qualitative/reliable differential probe for MSO2072A?

Pintek would be my first choice but I lack personal experience with their products.

Option 3) Do you mean that you then will use the tips only and use math operations from the scope to combine tip1 and tip2 for the 1st channel, and tip3 and tip4 for the 2nd channel? Nice suggestion actually.

This works in a pinch but lacks the common mode range and common mode rejection of a true differential probe limiting performance.  Whether this is acceptable or not depends on the application.  Some DSOs are unacceptably slow when operating like this.

[RedOctobyr]

Please forgive my asking a question in someone else's thread, but it seems related to the discussion. If this is poor form, just tell me I'm new to all this, just got my first scope, and will watch/read the items linked in post #2.

In the first post, Rene was measuring across 2 resistors.

I'd like to measure the voltage drop across a current shunt (to figure out the current), with a 2 channel scope. So a similar situation, but just one resistor. The circuit is battery powered, so the scope and circuit both being grounded shouldn't be an issue.

- I believe I could use the tip of the two channels, one before the shunt, one after. Then subtract one from the other to figure out the voltage drop across it. But using both channels feels like it might make things less accurate? At most, the drop across it should be 75mV, so fairly small.
- As the circuit is battery powered, could I put the tip of channel 1 before the shunt, and the clip of channel 1 after it? Thereby measuring the voltage drop "directly" with a single channel? I wouldn't be able to use channel 2 for anything, I assume, as that might cause a reference voltage difference problem.

Thank you.

[Nerull]

Please forgive my asking a question in someone else's thread, but it seems related to the discussion. If this is poor form, just tell me I'm new to all this, just got my first scope, and will watch/read the items linked in post #2.

In the first post, Rene was measuring across 2 resistors.

I'd like to measure the voltage drop across a current shunt (to figure out the current), with a 2 channel scope. So a similar situation, but just one resistor. The circuit is battery powered, so the scope and circuit both being grounded shouldn't be an issue.

- I believe I could use the tip of the two channels, one before the shunt, one after. Then subtract one from the other to figure out the voltage drop across it. But using both channels feels like it might make things less accurate? At most, the drop across it should be 75mV, so fairly small.
- As the circuit is battery powered, could I put the tip of channel 1 before the shunt, and the clip of channel 1 after it? Thereby measuring the voltage drop "directly" with a single channel? I wouldn't be able to use channel 2 for anything, I assume, as that might cause a reference voltage difference problem.

Thank you.

That sounds like the sort of job multimeters are designed for, but both methods will work fine. Beware that oscilloscopes are intended for accurate display of waveforms and not accurate reading of voltages and so their accuracy in this situation kind of sucks. It'll be close, but don't expect a better reading than a multimeter. The DS2072, for example, specs its DC gain accuracy at 2% full scale compared to 0.05% for a Fluke 87.

[tautech]

The current return path is via the GND tracks on the scope PCB.

Isn't the BNC simply soldered to the ground plane, which is connected to the chassis at each standoff?

I will not pretend to know the design or Gnd path PCB layout of all scopes available, but you just DO NOT expose your valuable test equipment to this type of abuse.

[pascal_sweden]

You will have 4 options.
Buy an isolated input DSO.
Buy differential probes.
Buy a 4 Ch scope and use std probes without Reference(Gnd) clips for a 2 Ch result
Forget the idea and learn how to use a scope safely.

Regarding the proposed options:
Option 1) Not an option, I want to stick with MSO2072A
Option 2) Where to get affordable/qualitative/reliable differential probe for MSO2072A?
Option 3) Do you mean that you then will use the tips only and use math operations from the scope to combine tip1 and tip2 for the 1st channel, and tip3 and tip4 for the 2nd channel? Nice suggestion actually.
Option 4) Even if you know what you are doing, you might make mistakes (e.g. on a Monday morning)

I found the following differential probe on the Rigol Europe website:
http://www.rigol.eu/products/accessories/rp1100d/

Why is it so expensive? Are all differential probes in this price range?
Is it because of the high voltage levels or because of the bandwidth?

Does it really take 700 euro to make a non-isolated 100 MHz scope safer for short circuits?

I don't need the protection for the high voltage, just the protection for short circuiting my probe in low voltage electronic designs (5V - 24V).

Are there differential probes for lower voltage levels that are much cheaper?
I need the bandwidth (at least 100 MHz), but not the high voltage range (just between 5V - 24V).

[tautech]

You will have 4 options.
Buy an isolated input DSO.
Buy differential probes.
Buy a 4 Ch scope and use std probes without Reference(Gnd) clips for a 2 Ch result
Forget the idea and learn how to use a scope safely.

Regarding the proposed options:
Option 1) Not an option, I want to stick with MSO2072A
Option 2) Where to get affordable/qualitative/reliable differential probe for MSO2072A?
Option 3) Do you mean that you then will use the tips only and use math operations from the scope to combine tip1 and tip2 for the 1st channel, and tip3 and tip4 for the 2nd channel? Nice suggestion actually.
Option 4) Even if you know what you are doing, you might make mistakes (e.g. on a Monday morning)

Without a dedicated Diferential probe you will only have 1 channel for use using this method from a 2 ch DSO.
Thats why a 4 ch DSO is preferable....no additional cost for differential measurements using std probes. But you only can display 2 waveforms in this way.

Option 4) Even if you know what you are doing, you might make mistakes (e.g. on a Monday morning)

Then give up this hobby and find something safer to do.

Why is it so expensive? Are all differential probes in this price range?
Is it because of the high voltage levels or because of the bandwidth?

CAT ratings.
No manufacturer can control how a user might use their equipment so all err on the side of safety.
The Siglent ISFE is by far the cheapest option albeit at 1 MHz bandwidth which surpisingly will do most of ones work when this type of isolation is needed.

BTW Siglent cheapest Differential probe:
DPB3050 50MHz, + /?800Vp?p, error of 2%, 10X/100X attenuation ~US $485 RRP plus taxes if any.

[RedOctobyr]

That sounds like the sort of job multimeters are designed for, but both methods will work fine. Beware that oscilloscopes are intended for accurate display of waveforms and not accurate reading of voltages and so their accuracy in this situation kind of sucks. It'll be close, but don't expect a better reading than a multimeter. The DS2072, for example, specs its DC gain accuracy at 2% full scale compared to 0.05% for a Fluke 87.

Thank you for the reply, I appreciate it.

Sorry, I didn't mention that I'm looking to measure current spikes as servos start to move. So I'm looking to measure a very brief event, likely too quick for a multimeter to capture. But a scope should show it.

My thought was to try putting a multimeter across the shunt, along with the scope, and flow a known, constant current through it. Letting me get some sense of "calibration" for the scope's reading, vs a multimeter. Not a perfect solution, but I'm trying to do the best I can with the tools I have access to.

[tautech]

That sounds like the sort of job multimeters are designed for, but both methods will work fine. Beware that oscilloscopes are intended for accurate display of waveforms and not accurate reading of voltages and so their accuracy in this situation kind of sucks. It'll be close, but don't expect a better reading than a multimeter. The DS2072, for example, specs its DC gain accuracy at 2% full scale compared to 0.05% for a Fluke 87.

Thank you for the reply, I appreciate it.

Sorry, I didn't mention that I'm looking to measure current spikes as servos start to move. So I'm looking to measure a very brief event, likely too quick for a multimeter to capture. But a scope should show it.

My thought was to try putting a multimeter across the shunt, along with the scope, and flow a known, constant current through it. Letting me get some sense of "calibration" for the scope's reading, vs a multimeter. Not a perfect solution, but I'm trying to do the best I can with the tools I have access to.

For precision measurements of this type one would use an oscilloscope Current probe.
AC versions are passive units whereas DC units are powered active devices, neither are cheap.

Owning a DSO is only the start, to use it to its full potential, one needs to acquire a collection of accesories.