Floating an oscilloscope for measurements

[System Error Message]

Hi it seems i absolutely have no choice but to disconnect oscilloscope from mains. I have AV2000 running and it has been feeding noise back into the oscilloscope and messing up readings. DC readings now appear as AC around 0V rather than the expected volts and the consistent negative voltage noise as well.

How would i go about floating the oscilloscope? The AV2000 uses the mains and ground.

[tggzzz]

Hi it seems i absolutely have no choice but to disconnect oscilloscope from mains. I have AV2000 running and it has been feeding noise back into the oscilloscope and messing up readings. DC readings now appear as AC around 0V rather than the expected volts and the consistent negative voltage noise as well.

How would i go about floating the oscilloscope? The AV2000 uses the mains and ground.

Precisely why do you have to float the scope?
What other techniques and technologies have you assessed and deemed to be inadequate, and why?

Who is making you indulge in dangerous practices?
Have you got it in writing that they are forcing you to indulge in dangerous practices? ... safely stored in a place your dependents can find it?

If it kills you, presumably innocent third parties will have to deal with your body. Will they also be at risk of the same electric shock, and if not, why not?

Have you made a will? The UK intestacy laws can be surprising.

[System Error Message]

I told you why, i have the powerline adapter av2000 running about.

I cannot make measurements on the same power grid with those adapters running as they communicate using the ground link also so it messes up measurements from the oscilloscope. So when measuring DC instead of say a +12V straight line with the usual noise i get a 2V peak to peak AC wave. I checked all the settings on the oscilloscope as well.

[nctnico]

Use a differential probe. Another solution would be to use a 600 Ohm telecom (modem) transformer between the DUT and your scope but this is AC only.
Still I can't imagine a power line modem using ground because ground may be absent. There may be another problem lurking like faulty mains wiring in the building.

[nfmax]

Are you trying to measure the AV2000, or is it simply that the presence of the AV2000 signals is interfering with the normal operation of your oscilloscope? If the latter, you need to replace either your oscilloscope with one that is not susceptible to the EMI generated by the AV2000, or disconnect the AV2000 (all of them) while you are measuring. These mains networking boxes are the devil's own invention, anyway

[System Error Message]

differential probes dont help, you have modulated digital signals on the ground that change as voltage is applied. Have any of you guys ever tried the av2000 powerline before? The AV2000 uses 2x2, the 2nd link using ground. Older powerline adapters do not use ground.

Using a transformer and differential probes on the front side of the scope isnt going to help as the problem is on the rear end so i will still get weird readings.

I didnt have this issue before installing the av2000. I also dont have a choice of removing them because im just renting so i cannot install cables for my network (already asked). Basically its modem -- powerline -- router. Is the bandwidth of the av2000 necessary? The answer is yes as it also has the lowest latency among powerline adapters. I actually get 1/3 of rated speed on the powerline adapter itself so downgrading will seriously hurt internet performance.

So i do need to remove the ground reference from the scope, so my question is how about would i do that? Do i get a UPS and run it on battery? Do i just use a 2 pin plug instead?

Im not trying to measure the av2000, the av2000 is interfering with the scope.

[tggzzz]

I told you why, i have the powerline adapter av2000 running about.

I have no idea what an "AV2000" is (as per your OP), and if you think I am going to search for a service manual and then read it to find how it workd, you will be sadly disappointed.

I cannot make measurements on the same power grid with those adapters running as they communicate using the ground link also so it messes up measurements from the oscilloscope. So when measuring DC instead of say a +12V straight line with the usual noise i get a 2V peak to peak AC wave. I checked all the settings on the oscilloscope as well.

If they lift the ground relative to something-or-other, then presumably that would also affect other equipment.

My other questions are rather important as well.

[tggzzz]

I didnt have this issue before installing the av2000. I also dont have a choice of removing them because im just renting so i cannot install cables for my network (already asked). Basically its modem -- powerline -- router. Is the bandwidth of the av2000 necessary? The answer is yes as it also has the lowest latency among powerline adapters. I actually get 1/3 of rated speed on the powerline adapter itself so downgrading will seriously hurt internet performance.

You should remove the source of the problem, not introduce other problems.

There are other newfangled ways of getting short-range network connectivity without cables. I suggest you investigate them.

Floating a scope may not affect internet performance, but it may seriously hurt other things' performance.

So i do need to remove the ground reference from the scope, so my question is how about would i do that? Do i get a UPS and run it on battery? Do i just use a 2 pin plug instead?

You can do anything you like, but not safely. The key question is "when a scope probe shield is accidentally connected to a dangerous voltage, what stops the entire external chassis being at that dangerous voltage?". Batteries won't make that safe!

Please consider the safety of the "first attenders" that may have to deal with your body.

[forrestc]

I told you why, i have the powerline adapter av2000 running about.

I cannot make measurements on the same power grid with those adapters running as they communicate using the ground link also so it messes up measurements from the oscilloscope. So when measuring DC instead of say a +12V straight line with the usual noise i get a 2V peak to peak AC wave. I checked all the settings on the oscilloscope as well.

Are you sure what you think is happening is what is actually happening here?  (Or stated differently:  Are you sure that you don't have something else broken?)

First of all, the scope circuitry itself should be isolated from the AC power line current carrying conductors completely.  This is a safety issue if it isn't.

The shield on the scope BNCs will most likely be attached to the ground - BUT... if used properly this shouldn't cause what you're describing - simply because the 'reference' for the measurement should be the ground connector on the probe itself, not the AC power line.  So measuring a DC power source should effectively reference the ground potential on the scope, eliminating any noise on the ground line.   Think about it this way:  The ground noise you're getting in the back becomes the reference for the scope.  It doesn't matter what that reference voltage *is*... it could be a very large sine wave.... once you connect ground from the front of the scope to your circuit, the reference voltage for the scope is identical to the reference voltage in your circuit.  (I'm ignoring potential safety issues with a large sign wave living on your ground, as that shouldn't ever happen).

Every time I've seen something like this, it's a situation of an improper or missing connection between the circuit under test and the shield on the BNC connectors.  Theoretically you could end up with some weird ground currents if your circuit is also ground referenced, and that *might* cause some problems.  But almost certainly not at the levels you're describing. 

I'd check your scope, and your probes, etc., before going down the path you think you need to go down.  Heck, start by finding a chunk of cable with a BNC connector on it and shorting it and then hooking it into the scope input.   

I apologize if you've already been down all of these paths, it's just that I'm very skeptical that ground noise on the AC line would cause this problem (with the magnitude you describe) if the cabling on the front of the scope is working and connected properly.   I think it's far more likely that either a) your probes are defective (or hooked up wrong) or b) Your scope needs some help.   What scope is this?  Can you post a pic of say the scope hooked up to a 9V battery and the accompanying waveform?   

Specifically in relation to ground isolation:  There are ways to do this, but very few are safe.  The scope ground is connected to the mains ground for good reason.  Only when you understand the reasons behind the connection, and how grounding keeps you safe.  Once you understand this, you'll also understand how to remove this isolation if you are willing to accept the risk.   I've never felt the need to do so.

[tautech]

SEM, you're between a rock and a hard place and I understand your dilemma but be assured what you're proposing has the potential to kill, so don't go there.
Find another safe solution no matter how painful it may be.

[singapol]

Hi it seems i absolutely have no choice but to disconnect oscilloscope from mains. I have AV2000 running and it has been feeding noise back into the oscilloscope and messing up readings. DC readings now appear as AC around 0V rather than the expected volts and the consistent negative voltage noise as well.

How would i go about floating the oscilloscope? The AV2000 uses the mains and ground.

You don't float the scope but use proper differential probe or differential amplifier which may be out of your budget.  Don't be reckless. Manufacturers of powerline networking don't follow industry standards and use different chipsets in their products resulting in some compatibility issues like this:

http://www.techhive.com/article/3011197/home-networking/zyxel-av2000-homeplug-av2-model-pla5456-powerline-adapter-review-speedy-hardware-dodgy-marketing.html

If you are curious to know how floating measurements are made then have a look at these: Remember that your life is precious.


Floating osciloscope measurements:

www.tek.com/dl/51W_10640_1.pdf

http://www.tek.com/document/application-note/three-facets-floating-measurement-solutions

http://www3.telus.net/Harley_Davidson/oscilloscope/differential%20techniques.pdf

http://www.davmar.org/TE/TekConcepts/TekProbeCircuits.pdf

[Alex Nikitin]

If a simple mains filter does not resolve this interference, you may have a problem with the ground connection on your mains. Under no circumstances you should have a noise, capable to interfere with a scope operation, to appear on the mains ground wires. Something is not right here.

Cheers

Alex

[System Error Message]

well whats not right is a DC voltage out from a bench PSU that i tested less than a month ago appearing as AC on the scope itself (scope is set to DC). I still havent tested plugging the scope behind the powerline adapter itself or a noise filter on ground. If the PSU's voltage was AC and less than 10 volts peak to peak it would not be able to power the equipment i was testing.

In the UK things are very tight in terms of space and supply that interference comes from various places. Short of a faraday cage for wireless interference it seems like i will need a filter for ground.

So aside from filters, plugging behind the PLA, what else can be done?

[Fungus]

well whats not right is a DC voltage out from a bench PSU that i tested less than a month ago appearing as AC on the scope itself (scope is set to DC).

Sure, but floating the 'scope won't fix that.

Why not float the PSU instead?

So aside from filters, plugging behind the PLA, what else can be done?

Check your mains wires. Maybe there's a loose wire in a plug or something.

Are both the 'scope and PSU plugged into the same power strip? If not, try that.

Have you got any new light fixtures in the last month. Cheap LED lights can do things like this (Dave made a video on it).

Take both the 'scope and PSU to somebody else's house and try it there.

[System Error Message]

Check your mains wires. Maybe there's a loose wire in a plug or something.

Are both the 'scope and PSU plugged into the same power strip? If not, try that.

Have you got any new light fixtures in the last month. Cheap LED lights can do things like this (Dave made a video on it).

Take both the 'scope and PSU to somebody else's house and try it there.

I use philips LED bulbs, not sure if you call them cheap.

If there is a loose wire it will spark as i have seen it before when dealing with poor quality terminals which you have to screw the cable down on. Cable got loose and there are a few sparks between the wire and terminal (very small gap, insignificant sparks).

I could certainly try the same power strip or test against batteries.

[Fungus]

Have you got any new light fixtures in the last month. Cheap LED lights can do things like this (Dave made a video on it).

I use philips LED bulbs, not sure if you call them cheap.

Did you get them recently?

One of them may be failing or have a bad capacitor in it. That can do all sorts of bad things to the power.

I could certainly try the same power strip or test against batteries.

Also try it at a friend's house (seriously!)

[System Error Message]

LEDs are easy to dismiss, just turn them off. They are LED bulbs, not the fancy LED light strips or lights that use pulsing and other tricks to be bright and power saving.

Nearest friend's house is extremely far away so not an option.

Edit: i just tested measuring a battery by both either using only 1 terminal on probe and both. With connecting both the probe and ground i got the accurate voltage, with just the probe itself the result is in the screenshot.

So now my concern is when i cannot use ground to probe the circuit.

[nfmax]

Have you tried temporarily unplugging the AV2000 units (both of them) and confirming the oscilloscope then operates correctly?

[System Error Message]

unplugging the AV2000 would be very very inconvenient especially for putting them back. Not only are things plugged into them and losing internet but to have to go down some floors unplug, and repeat is tedious including the annoyed looks i will get. I used to be able to just connect probe and not ground and get a near accurate voltage of what i was measuring.

[nfmax]

Wait - are you saying your standard way of measuring was to connect only the probe signal lead and not the ground lead, relying on a common mains earth to complete the circuit? That is just asking for trouble!

Edit to add:
Is either your scope or the UUT actually connected to mains through the feedthrough socket on the AV2000? Have you tested the AV2000 for continuity between mains earth in and mains earth out (while disconnected from the mains supply, of course!)

[Fungus]

Wait - are you saying your standard way of measuring was to connect only the probe signal lead and not the ground lead, relying on a common mains earth to complete the circuit? That is just asking for trouble!

I'm also worried about the "go down some floors and unplug" part.

I think we need more information about how things are connected.

[System Error Message]

Its not easy trying to hook up the ground of the probe at times to measure.
For example in this picture

There is a hole for you to insert probes but the ground for the probe is a crocodile clip so hooking it up and making sure the crocodile clip doesnt touch the board or circuit and accidentally cause a short is another because for one thing its an expensive equipment, another is that it has no ground connection. So trying to measure the voltage of it while it is powered up is tricky so i dont always connect the ground probe.

Besides isnt the ground probe itself connected to ground being more likely to be dangerous to connect both leads if the circuit is grounded?

[Alex Nikitin]

 

[nfmax]

well whats not right is a DC voltage out from a bench PSU that i tested less than a month ago appearing as AC on the scope itself (scope is set to DC). I still havent tested plugging the scope behind the powerline adapter itself or a noise filter on ground. If the PSU's voltage was AC and less than 10 volts peak to peak it would not be able to power the equipment i was testing.

Stop right there! If I understand what you are trying to say, it looks as if when you plug your oscilloscope into the output of your AV2000 the earth pin of the oscilloscope mains plug is no longer connected to mains earth. Stop. No I mean STOP! Disconnect the mains supply to the AV2000, disconnect the AV2000 completely from the mains and everything else, and check for continuity between the 'plug' earth pin and the 'socket' earth contact of the AV2000. I think you might have a faulty unit which could kill you. You should be able to return it for a correctly operating unit or a refund.

[System Error Message]

well whats not right is a DC voltage out from a bench PSU that i tested less than a month ago appearing as AC on the scope itself (scope is set to DC). I still havent tested plugging the scope behind the powerline adapter itself or a noise filter on ground. If the PSU's voltage was AC and less than 10 volts peak to peak it would not be able to power the equipment i was testing.

Stop right there! If I understand what you are trying to say, it looks as if when you plug your oscilloscope into the output of your AV2000 the earth pin of the oscilloscope mains plug is no longer connected to mains earth. Stop. No I mean STOP! Disconnect the mains supply to the AV2000, disconnect the AV2000 completely from the mains and everything else, and check for continuity between the 'plug' earth pin and the 'socket' earth contact of the AV2000. I think you might have a faulty unit which could kill you. You should be able to return it for a correctly operating unit or a refund.

Quite wrong, as i said i am not trying to measure the AV2000.

The oscilloscope and AV2000 are on seperate plugs, both are connected to earth. If i were to plug the oscilloscope into the plug on the av2000 they will both be connected to earth.