Author Topic: Rigol 2072A - ground leaking ?  (Read 7169 times)

0 Members and 1 Guest are viewing this topic.

Offline tiagobarachoTopic starter

  • Regular Contributor
  • *
  • Posts: 66
Rigol 2072A - ground leaking ?
« on: February 01, 2014, 05:50:31 am »
Hi guys !! I received my DS2072A .. I AM IN LOVE !!!! my first scope and seems amazing... I bought a differential probe.. Still waiting for delivery...

I have one doubt...
I did one test here.... The pron to connect to the outlet has 3 pins right ..... I did test with adapter to  not connect the earth pin...
On both positions,  ( live one pin and neutral on the other...vice and verse) when i touch on the ground of BNC i could fell a little bit of shock in my fingers...

This sometimes is common with few electronics device like computers, washing machine.... and if you plug it the other way, usually it stops..... So what i do is find this better position and then put the earth... ( usually the default position are the best)...
But on this rigol, i felt the shock on both position.... Is that normal ? Because i guess first without the ground it should be isolated the BNC ground to the mains input.... but if i am feeling small shock, it means that back on the power supply live part, there should be something going through both input wires to the ground...
This is the power supply from the teardown dave did... Look the C149 and C150 capacitors
http://www.flickr.com/photos/eevblog/8022086942/#in/set-72157631618295437
Both are connected to one side of the input and the other side goes to ground... So if ground is floating(not connected), the live wire( does not matter wich pins is it) will go through a capacitor and to the ground... I suppose that if I remove those caps the shock on the ground  would go away.
Look the picture i took just touching the probe ... It shows the 60  HZ signal from the 220v outlet because in reference to the scope BNC ground i am real earth ground and the scope ground is transmitting some of the input live...
its normal and the power supply needs the earth to work well ??
Thanks
 

Offline npelov

  • Frequent Contributor
  • **
  • Posts: 330
  • Country: bg
    • Microlab.info
Re: Rigol 2072A - ground leaking ?
« Reply #1 on: February 01, 2014, 09:36:23 am »
Do you not like your life? There is a simple device that detect live wire - I'm not sure of the name in english. If you try to find leakage you should use it or use a voltmeter with a large load resistor - 100k  to try to measure leakage voltage to ground. Measuring leakage with finger is not the best way. Especially on chinese stuff even though it's a relatively good branded one.

But ... I'll have to check that. The leakage shouldn't be enough to feel it (I won't touch it though).

But if you touch the probe you always get 50/60 Hz regardless of earth connection. That's the power of the earth :) ... or the difference in earth potentials. Everything leaks, but usually the current is too small to be felt.
 

Offline npelov

  • Frequent Contributor
  • **
  • Posts: 330
  • Country: bg
    • Microlab.info
Re: Rigol 2072A - ground leaking ?
« Reply #2 on: February 01, 2014, 09:56:43 am »
I know that measuring the leaking with a voltmeter won't work  because the insulation resistance will be different at high voltages, but you could try that. On my Brymen BM869 multimeter I get nothing in megaohm range, so it should be >50MOhm. I tried to measure the conductance - about 0.15nS - 6.67 G, but it's too noisy and it rises when I put my hand over it, so the reading is probably not acurate. But if you get anything <20MOhm between earth and any of the other power lines then you have serious leakage.

Btw I measured 7V/50Hz AC between earth connection on the power socket and my finger. Many people use that for dummy signal source :)
 

Offline tiagobarachoTopic starter

  • Regular Contributor
  • *
  • Posts: 66
Re: Rigol 2072A - ground leaking ?
« Reply #3 on: February 01, 2014, 11:53:44 am »
Do you not like your life? There is a simple device that detect live wire - I'm not sure of the name in english. If you try to find leakage you should use it or use a voltmeter with a large load resistor - 100k  to try to measure leakage voltage to ground. Measuring leakage with finger is not the best way. Especially on chinese stuff even though it's a relatively good branded one.

But ... I'll have to check that. The leakage shouldn't be enough to feel it (I won't touch it though).

But if you touch the probe you always get 50/60 Hz regardless of earth connection. That's the power of the earth :) ... or the difference in earth potentials. Everything leaks, but usually the current is too small to be felt.
I do like my life... That's why I have been fixing electronics for 13 years and never had any bad surprise.....
Is just a normal leakage, people...... Like computers without earth sometimes do.... Forgeting about my life a little bit, my doubt is if its normal on this device to even rotating the outlet position it still leaks on the ground of bnc enought to be felt?
I Will measure with a voltmeter and post the results in few minutes
 

Offline sean0118

  • Frequent Contributor
  • **
  • Posts: 453
  • Country: au
Re: Rigol 2072A - ground leaking ?
« Reply #4 on: February 01, 2014, 12:44:25 pm »
Shouldn't that make the RCD (Safety switch) on your house trip?  :o


« Last Edit: February 01, 2014, 12:51:50 pm by sean0118 »
 

Offline sigxcpu

  • Regular Contributor
  • *
  • Posts: 64
  • Country: ro
Re: Rigol 2072A - ground leaking ?
« Reply #5 on: February 01, 2014, 01:27:01 pm »
Isn't that the Y capacitor of a SMPS?
 

Offline commongrounder

  • Frequent Contributor
  • **
  • Posts: 372
  • Country: us
Re: Rigol 2072A - ground leaking ?
« Reply #6 on: February 01, 2014, 02:31:24 pm »
As sigxcpu said, those are safety capacitors across the AC line (both hot and neutral to ground) feeding the SMPS.  They must not be removed or you will expose your scope power supply to outside voltage spikes that could damage the circuitry.  They are also part of the noise filtering system to prevent noise from the scope supply itself radiating back into the AC line.  SMPS are different than linear power supplies with conventional transformers in that way.  I'm wondering why you would want to lift the ground pin, anyway?  When you get your differential probe, it will be required that you ground the scope for the probe to work properly.  It has to have a ground reference.  There is nothing wrong with the scope, IMHO.  The leakage current should measure under 6ma or so, to avoid tripping a ground fault circuit interrupter.
 

Offline npelov

  • Frequent Contributor
  • **
  • Posts: 330
  • Country: bg
    • Microlab.info
Re: Rigol 2072A - ground leaking ?
« Reply #7 on: February 01, 2014, 02:45:29 pm »
Is just a normal leakage, people......

I've seen people dancing around while holding two parts of a coaxial cable connected to old 10MBit network because one PC wasn't grounded well and the other one was. "a normal" leakage could from small to dead short between live and ground + bad ground. I don't have your experience but I've had many bad surprises. So if you've burned once you blow the cold soup too (doesn't sound that good in English)

@sean0118 not everyone has one - my apartment is 2 wires only and doesn't have safety switch. And these things trigger at 30 mA. Few micro/miliamps won't trigger it.

@tiagobaracho What's the model/price of the differential probe? What I did before is I used math with two probes to simulate differential probe. worked somewhat on my axiomet 1022C shitty scope. However Rigol DS2072A has really slow update rate of math - like 1-2 times per second. It might go up to 50 000 wfps in dot mode, but it's math is quite bad. FFT is the same speed - 1-2 fps.

 

Offline kxenos

  • Frequent Contributor
  • **
  • Posts: 284
  • Country: gr
Re: Rigol 2072A - ground leaking ?
« Reply #8 on: February 01, 2014, 02:54:59 pm »
...So if you've burned once you blow the cold soup too...
haha we have the same frase but we say the yogurt
 

Offline tiagobarachoTopic starter

  • Regular Contributor
  • *
  • Posts: 66
Re: Rigol 2072A - ground leaking ?
« Reply #9 on: February 01, 2014, 04:21:29 pm »
Guys... let me restart...

First.. i was not saying that i will remove anything from the power supply.. i just said that I GUESS this is the only path the energy may come through .. Because  the  SMPS ins complete isolated right ?the live phase with the secondary.... SMPS the transform is isolated and everything else goes through photocouplers...
So... How can it be that if I connect it on the outlet without earth, i can fell some electric leakage?  the secondary should be isolated and since the ground is not connected, it should have no voltage at all.... but it does.... I may simple plug the ground to earth and everything should work... But i am here trying to understand why ? why is it working like that ? Shouldn't it be completely isolated?


( update : guys.. i discovered !! the adapter that i used was a little bit wet inside .... i took it out of equation  and problem solved !I guess the woman that works here was cleaning the floor and spilled some whater)
I have just another doubt....
I have in my building 2 wires only... no real earth wire..... Usually i use the neutral as earth ( because i know they made it to earth on the junction box...  of the building)
It was installed a earth rod and jumped to the neutral ....
Should i use this neutral ? How do you guys do in this situation ?
Thanks
 

Offline cybermaus

  • Frequent Contributor
  • **
  • Posts: 674
  • Country: nl
Re: Rigol 2072A - ground leaking ?
« Reply #10 on: February 01, 2014, 06:08:25 pm »
So... You live in a place that does not provide ground, and also does not provide ground-fault protection? What country was that?
And rewiring your place up to code is not in your reach?

I am a really big fan of ground fault protection. While your trick of using neutral as ground will work to keep your ground-plane at ground potential (minus any residual potential caused by wire resistance, which could be several volts), it will not make anything safe for you. Not to mention what happens if you accidentally reverse a plug.


Do you have central heating? Or waterpipe? Can you check if any of those are properly grounded ? If so, maybe do the following:

Buy a ground-fault protection device somewhere. EU GF devices are for central mounting in the central power entry points on a D-rail and at 30milliAmps. US GF devices are more often decentralized, 6milliAmps and mounted in a wall-socket, though in-cord devices also exists. Don't get a cheap 2nd hand industrial device, they are (in EU) at 300milliAmps. If you by a 2nd hand device, do test it by using the test button or by connecting a 6K8Ohm resistor between life and ground.

Use the GF protection device to power everything on your workbench. Connect Hot to Hot, Neutral to Neutral and Ground to central heating or waterpipe. Or if you really do not have anything grounded, connect Ground to Neutral also, but on the incoming side of the GF protection, not on the usage side.

(Consider buying a second GF protection device, and power everything in your kitchen through it. It may safe your life one day, especially if your housekeeper splashes water onto electrics like that)
« Last Edit: February 01, 2014, 06:40:53 pm by cybermaus »
 

Offline leppie

  • Frequent Contributor
  • **
  • Posts: 269
  • Country: za
Re: Rigol 2072A - ground leaking ?
« Reply #11 on: February 01, 2014, 06:28:41 pm »
I recently purchased a Tektronix 465, and the grey beard told me he left earth disconnected.

He did state when measuring measuring power supplies with negative offsets could cause the the main breaker to trip if it was earthed.

I had it like that for a week, and got a few tiny little zaps (he assured me this was very low current 110V, me being in a 220V zone).

I got annoyed by the zaps and reconnected earth, and due to not working with negative voltages (I assume this has to be quite a bit negative) I have had no issues. No zaps either!

He did recommend however that if I want to connect the earth (which I ignored) to prevent mains tripping, to add an inline 10K 1-watt resistor.
 

Offline pomonabill221

  • Frequent Contributor
  • **
  • Posts: 252
  • Country: us
Re: Rigol 2072A - ground leaking ?
« Reply #12 on: February 01, 2014, 07:55:44 pm »
Many electronic devices use this type of filtering at the mains input.  They are common mode caps and should stay in place, but you SHOULD ALWAYS use a ground also!
The leakage current you are experiencing is from the two caps, but they are relying on the earth being connected!
Normally, household wiring is hot and neutral (or hot/hot for 220v) and the earth (ground) is carried by the flex or rigid conduit that is connected to the ground pin on the receptical, through the mounting screws into the box, to the distribution panel case.
The neutral should NEVER BE USED as a grounding conductor!!!!  Likewise, the ground is NOT a current carrying conductor and should NEVER BE USED inplace of the neutral.
You are really setting yourself up for injury/death/fire/etc., if you do this!
The ground is there for YOUR safety (it is called protective ground for a reason).

I know I have "floated my scope" when measuring line referenced supplies/equipment, but I know what I am doing.
If I can, I use an isolation transformer rather than floating my scope, but that isn't always possible.
For low current equipment, a 1:1 transformer that is adequately sized for the load is the way to go.
Normal variacs are NOT isolated though.
 

Offline tiagobarachoTopic starter

  • Regular Contributor
  • *
  • Posts: 66
Re: Rigol 2072A - ground leaking ?
« Reply #13 on: February 01, 2014, 08:02:52 pm »
So... You live in a place that does not provide ground, and also does not provide ground-fault protection? What country was that?
And rewiring your place up to code is not in your reach?

I am a really big fan of ground fault protection. While your trick of using neutral as ground will work to keep your ground-plane at ground potential (minus any residual potential caused by wire resistance, which could be several volts), it will not make anything safe for you. Not to mention what happens if you accidentally reverse a plug.


Do you have central heating? Or waterpipe? Can you check if any of those are properly grounded ? If so, maybe do the following:

Buy a ground-fault protection device somewhere. EU GF devices are for central mounting in the central power entry points on a D-rail and at 30milliAmps. US GF devices are more often decentralized, 6milliAmps and mounted in a wall-socket, though in-cord devices also exists. Don't get a cheap 2nd hand industrial device, they are (in EU) at 300milliAmps. If you by a 2nd hand device, do test it by using the test button or by connecting a 6K8Ohm resistor between life and ground.

Use the GF protection device to power everything on your workbench. Connect Hot to Hot, Neutral to Neutral and Ground to central heating or waterpipe. Or if you really do not have anything grounded, connect Ground to Neutral also, but on the incoming side of the GF protection, not on the usage side.

(Consider buying a second GF protection device, and power everything in your kitchen through it. It may safe your life one day, especially if your housekeeper splashes water onto electrics like that)
My building has 25 years and it was rebuilt the wires yet..
I live in a Brazil, but the problem is the old building i live..But anyways... from 10 year ago they start using real earth in every construction... Its law now..
On my brother's home he has the  ground protection..... I will buy one....
Here we dont have heating system...  The only " earth" that i might find is breaking the wall to reach for the construction rods... but i dont know if they work well as ground...
 

Offline tiagobarachoTopic starter

  • Regular Contributor
  • *
  • Posts: 66
Re: Rigol 2072A - ground leaking ?
« Reply #14 on: February 01, 2014, 08:07:41 pm »
Many electronic devices use this type of filtering at the mains input.  They are common mode caps and should stay in place, but you SHOULD ALWAYS use a ground also!
The leakage current you are experiencing is from the two caps, but they are relying on the earth being connected!
Normally, household wiring is hot and neutral (or hot/hot for 220v) and the earth (ground) is carried by the flex or rigid conduit that is connected to the ground pin on the receptical, through the mounting screws into the box, to the distribution panel case.
The neutral should NEVER BE USED as a grounding conductor!!!!  Likewise, the ground is NOT a current carrying conductor and should NEVER BE USED inplace of the neutral.
You are really setting yourself up for injury/death/fire/etc., if you do this!
The ground is there for YOUR safety (it is called protective ground for a reason).

I know I have "floated my scope" when measuring line referenced supplies/equipment, but I know what I am doing.
If I can, I use an isolation transformer rather than floating my scope, but that isn't always possible.
For low current equipment, a 1:1 transformer that is adequately sized for the load is the way to go.
Normal variacs are NOT isolated though.
Thanks for your reply...
But if you live in a building ( 10 floor) , old - 15 year, with no ground.. just neutral..... What wold you do ? Note use the neutral as ground or leave it open...
I cant use any ground here....there is no way to have here on the 10 floor
I bought one differential probe for safety and i will get one isolation transform... It can be any transform that has no impedance between the output and input wires, right ?
« Last Edit: February 01, 2014, 08:09:53 pm by tiagobaracho »
 

Offline cybermaus

  • Frequent Contributor
  • **
  • Posts: 674
  • Country: nl
Re: Rigol 2072A - ground leaking ?
« Reply #15 on: February 01, 2014, 08:23:25 pm »
If it has construction rods, they are likely in concrete and not conducting well.

Officially, ground wire is allowed only a few ohms from actual earth. The reason is that while ground is not allowed to be intentionally used as current conductor (as previous poster mentions) it should be powerful and thick enough to deduct away all current if there is a double fault (iow, if not only mains touches chassic but also the fuse is at fault and does not blow)

Not sure about your budget. Consider having someone place a real earthing pin. Measure and use water pipe. Connect to neutral may not be the best option, but still a possible choice. It is actually what the official installer does also, in the sealed part of your installation that you cannot reach normally. But as stated, connect them on the far side of the GF protection device. (and indeed, it is not officially allowed)

In any case, even if you have no ground wire at all, a ground fault device will still protect you. You see, a ground fault device does not rely on ground itself. It will trip if you touch mains, regardless if you have ground or if that ground is broken or not. That's why I think they are so great.

So do get that GF device, whatever they are called in Brazil, regardless if you get ground officially or illegally or not at all

And to keep your chassis at ground potential. Anything is better then nothing, but if you can spare the cost, consider bringing your installation up to code, though I guess on the 10th floor, placing your own grounding rod may indeed be a bit of a challenge.


 

Offline commongrounder

  • Frequent Contributor
  • **
  • Posts: 372
  • Country: us
Re: Rigol 2072A - ground leaking ?
« Reply #16 on: February 01, 2014, 09:00:12 pm »
In the Unites States, the National Electrical Code allows for earth bonding to concrete foundation and floor reinforcing rod and mesh.  This is normally a second connection along with a ground rod driven into soil, but if it is all that is available, then it is better than nothing.  A cold water pipe can also serve if it is continuous steel or copper all the way to the street.  Attaching the ground pin of a receptacle to the neutral pin is called a false or "bootleg" ground and is expressly forbidden here.
tiagobaracho  I am concerned about the use of a differential probe on your scope with it ungrounded.  The diff probe you mentioned you ordered earlier must have a earth ground for it to be safe, and work well, while probing hot circuits.  The input impedance on the probe leads are around 4megohms each to common, and are not isolated from ground (BNC shell).  An isolation transformer will help with this, but the scope chassis and ground connections (BNCs, etc,) could still potentially float to some unexpected voltage relative to ground.   I would find it to be difficult to have a reliable, safe electronic test environment without a good earth ground.
 

Offline npelov

  • Frequent Contributor
  • **
  • Posts: 330
  • Country: bg
    • Microlab.info
Re: Rigol 2072A - ground leaking ?
« Reply #17 on: February 02, 2014, 09:17:04 am »
In any case, even if you have no ground wire at all, a ground fault device will still protect you.

Are you sure? I heard that if you have 2 wire installation you can't make use of ground fault device. That's because ground fault device measures the difference of what's going out (well direction changes) of one wire and what goes in the other (that's for live and neutral). I don't know how current going in earth really work, but that's what I heard. You must have neutral connected to earth at one point in the building and then they should go separately (someone will correct me if I'm wrong). That's the only way ground fault could trigger.

What I thought is that if you touch the live wire the current will flow to ground which and because ground is connected to neutral before the ground fault protection it should trigger. But people I asked say that this is not the case and the building must have 3 wires or the protection won't work. I've powered my bench from a 3 wire cable  and I've put ground fault protection. I'm not sure if it'll work. If I connect the live to the wet floor it still doesn't trigger. Maybe the current is not enough.

@tiagobaracho so you had leakage in the contact, not the scope. That's a relief. But as you see surprises happen. It's good that they seal the plugs so if it gets wet it'll dry really fast.
 

Offline Rerouter

  • Super Contributor
  • ***
  • Posts: 4694
  • Country: au
  • Question Everything... Except This Statement
Re: Rigol 2072A - ground leaking ?
« Reply #18 on: February 02, 2014, 09:40:05 am »
the proper names are an RCD (residual current device) or GFCI (Ground Fault Current Interrupter) as they work by measuring the difference if you where to touch a local ground point current would flow through you, into the ground stake, through a few hundred meters of soil back up the ground stake at the transformer and return at that neutral point, so it would see a lower current returning through the measured neutral and trip

arguably this does nothing for you if you stick your hands between active an neutral, without some path to ground (be it concrete or wrist strap) as there is a minimal difference in current,
 

Offline cybermaus

  • Frequent Contributor
  • **
  • Posts: 674
  • Country: nl
Re: Rigol 2072A - ground leaking ?
« Reply #19 on: February 02, 2014, 09:44:10 am »
Yes, I am sure.

It measures the balance between Life and Neutral. Or L1 and L2 and Neutral if you have a 2-phase split US 240V dryer socket. Or L1,L2,L3,N if you have a 3-phase system.

In all cases, if those currents do not balance out, it trips. It does not actually need to measure if any current goes through ground. And thats a good thing to, because if you touch any hot wire, current would flow through you into the real ground, and not back through the ground wire.

It is indeed needed to have Neutral and Ground connected to have the GF device work, but that connection can be (and will be) all the way in the nearest power company owned trasformer. No actual ground wire needed in your premises. There are other reasons why you should have a ground wire, but not to make the GF device work.

----

Proper names are RCD in UK, or CFCI in US. Or ALCI or RCCB. And then you have the combined overload protection ones RCBO. They are called Alamat in NL. Who knows what they are called in Brazil.

Also, while I will not try it, I firmly believe that if you "stick your hand between" life and neutral, it is very unlikely the current will stay balanced, therefore it will trip. So while not designed for it, with great likelihood it will also protect against grabbing both wires. It will be a bigger burn due to the short path between wires, so don't try. But it will probably trip. (unless if you are also standing on a rubber plate)
« Last Edit: February 02, 2014, 10:20:37 am by cybermaus »
 

Offline johna

  • Contributor
  • Posts: 10
Re: Rigol 2072A - ground leaking ?
« Reply #20 on: February 04, 2014, 09:14:50 am »
@cybermaus Well, if I think  it is true. When I think the current still flows out of one power line into a path that's not going through the RCD. The difference should still trigger the protection. But it is 30mA current that triggers the protection. You could die from less if you have heart problems. But it's a good thing that I've put a RCD then. There is a test button, but maybe I should test it by connecting life wire to neutral before RCD through a resistor that'll provide more than 30mA current - like 4.7k (I have 4.7k/5W). The test button probably does just that, but  ... better safe then sorry.

@tiagobaracho can you tell me which probe did you order (a link, price?) and would you share experience when you get it.
 

Offline cybermaus

  • Frequent Contributor
  • **
  • Posts: 674
  • Country: nl
Re: Rigol 2072A - ground leaking ?
« Reply #21 on: February 04, 2014, 10:17:39 am »
Indeed, the test button is just a resistor too.

And yes, 30mA could still kill you (or so people tell me). So don't go grabbing the mains just to prove a point. However, I suspect (not sure, not claiming medical knowledge) to make it dangerous it would need to go over your heart and be maintained for a minimum duration . Not onl is the 30mA a maximum, it also has to shut down within some rated time, like 0.1 Second or something.

I do know (I have personally accidentally grabbed mains more then I am proud of) that without one you actually feel the 50Hz tingling and biting your fingers, with one you notice the lights went out (power is off) before you even realize you might just have felt something.

Still, as to the 30mA : I did once think to get a US one for my workbench, as those are 4~6mA. They are decentralized, one per power socket, so they do not have to assume accumulated leakage and thus can be stricter.
Problem is, their internal electronics run on 110V, even the 240V dryer socket ones need the mid-rail neutral for that reason, so I cannot use those in EU.

Maybe the ones in Brazil have a different rating too?
 

Offline Rerouter

  • Super Contributor
  • ***
  • Posts: 4694
  • Country: au
  • Question Everything... Except This Statement
Re: Rigol 2072A - ground leaking ?
« Reply #22 on: February 04, 2014, 10:35:55 am »
Further reading seems to point to the fact that 5 and 10mA units are used on 110V as there is less chance of a nuisance trip than with 240V, equally that most 30mA RCD's actually trip around the 20mA mark,

Look for a type 1 RCD, these are mainly for medical and have a trip point of 10mA, (line frequency does matter)
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf