Investigation into ground loop

[sahib]

I prepared this to bring some clarification to the discussion took place in this thread "Boutique" opamps from chinese vendors. Anyone tested?

Comments (in a civilized manner) from all forum members are naturally welcome.

I also would like to dedicate this post to John for the work he did on audio myths. I have not read it yet but I will. So I hope he accepts this with its truths or faults.

The discussion was around two main points.

1. Does the mains earth impedance has effect on the ground loop.
2. Does making the equipment safety earth wire impedance zero reduces the loop impedance to zero.

I said yes to 1 and no to 2.

However, in 1 I was wrong. I have suspended this view again as I believe the test carried out is not valid. So, I am going to device another test.

For 2, no it does not.

However, making path impedances/resistances very low the overall loop impedance can be made very low. This is explained in later posts.



Above diagram is re-drawn as impedances below.

 

[JohnRoberts]

oh-oh...

For starters how do you define a "ground loop" in the context of your discussion?

[edit: In my experience "ground loops" are a catch-all term applied to sundry ground noise problems. /edit]

JR

 

[moamps]

... Otherwise comments (in a civilized manner) from all forum members are naturally welcome.
.

I would suggest that you improve your model by adding interference currents that are injected into the ground points through the CY capacitors in the devices. The test voltage of 10V is approx. half of the standard max harmonic voltage sum according to European standards where the maximum allowed THD is 8% (say EN 50549) and where the 3rd harmonic 150Hz is 5%. etc..
And what you need to investigate is whether the current between points C and C' (screen wire) changes (whether it flows at all) due to impedance changes in path A and A' (changing the power cable). It is only one of several possible points of investigation.

 

[Newmarket]

oh-oh...

For starters how do you define a "ground loop" in the context of your discussion?

[edit: In my experience "ground loops" are a catch-all term applied to sundry ground noise problems. /edit]

JR

+1. This.
I was going to add a comment but you got there before me with the added edit.

 

[Voyager10]

I'd also point out that, if the physical area occupied by earth wire A, earth wire A' and the cable screen (the green loop above) has any changing magnetic fields in it, then a voltage will be induced, causing a current to flow around the loop.

This is in addition to any PSU leakage currents, and can be really tricky to remove from the environment.

 

[sahib]

oh-oh...

For starters how do you define a "ground loop" in the context of your discussion?

[edit: In my experience "ground loops" are a catch-all term applied to sundry ground noise problems. /edit]

JR

True. Although I have titled the thread with regards to the ground loop, I am not defining the ground loop at all and I would not attempt to anyway. I have addressed the two points I mentioned in the post and that is all.

 

[Speedskater]

a] the connection to Planet Earth has nothing to do with it.
b] reducing the Safety Ground/Protective Ground impedance from unit to unit will reduce the loop impedance, thus reduce the problem.

 

[sahib]

I'd also point out that, if the physical area occupied by earth wire A, earth wire A' and the cable screen (the green loop above) has any changing magnetic fields in it, then a voltage will be induced, causing a current to flow around the loop.

Thank you for that. I am well aware of this as I also have a goof background in electromagnetism.

This is in addition to any PSU leakage currents, and can be really tricky to remove from the environment.

Absolutely, and that is really the point whether simply plugging in low impedance power/signal cables will be the silver bullet.

I would suggest that you improve your model by adding interference currents that are injected into the ground points through the CY capacitors in the devices. The test voltage of 10V is approx. half of the standard max harmonic voltage sum according to European standards where the maximum allowed THD is 8% (say EN 50549) and where the 3rd harmonic 150Hz is 5%. etc..

Indeed. there is always room for improvement. However, as I mentioned before my aim was to look into the two points.

And what you need to investigate is whether the current between points C and C' (screen wire) changes (whether it flows at all) due to impedance changes in path A and A' (changing the power cable). It is only one of several possible points of investigation.

Well, if there is say, hum formed on both equipment when connected together, then a current must flow through the screen as it is one of the two paths forming the loop (the other being the mains earth).

However, have you seen the link Newmarket posted in the other thread? They investigate this in pro-audio application. They run a parallel low impedance (thicker) ground wire along side the (balanced) signal cable to divide the noise current going through the screen.

 

[sahib]

a] the connection to Planet Earth has nothing to do with it.
b] reducing the Safety Ground/Protective Ground impedance from unit to unit will reduce the loop impedance, thus reduce the problem.

If you read the last bit of my first post you will see that the mains earth impedance is having an effect on the loop. Just look the test result. I am not suggesting it is 100% correct, but if you think it is not, then it would be great to point out where I went wrong.

 

[moamps]

Indeed. there is always room for improvement. However, as I mentioned before my aim was to look into the two points.

My point is that you should include in your analysis the interferences that occur in individual devices and that contaminate the ground wire. This is one of the causes of the appearance of current through the shield of the audio cable and the reduction of the signal/noise ratio.

 

[sahib]

My point is that you should include in your analysis the interferences that occur in individual devices and that contaminate the ground wire. This is one of the causes of the appearance of current through the shield of the audio cable and the reduction of the signal/noise ratio.

I know what you are saying but that does not form part of the objectives of the investigation. All I am looking into whether the mains earth impedance have an effect on the loop impedance, and replacing the earth wires with zero impedance makes the loop impedance zero. These are the two objectives.

Otherwise, as you correctly point out this is also one of the culprits of noise currents flowing within the system. It would be great to investigate all of these without re-inventing the wheel.

 

[Newmarket]

If you read the last bit of my first post you will see that the mains earth impedance is having an effect on the loop. Just look the test result. I am not suggesting it is 100% correct, but if you think it is not, then it would be great to point out where I went wrong.

I think the point being made is that whilst those contact / wire impedances are real and can be shown to give the mathematical results you state, the actual connection to the Earth is irrelevant in the circuit shown. The wire / contact resistances exist without the connection to Earth.
Of course, if you were not making a connection to Earth you would not need that wiring and the loop would not be formed (as for unearthed Class II equipment).
But @Speedskater 's point is worth noting in general as there is often a good deal of misunderstanding the Earth as having some special significance in somehow "magically" absorbing or draining current when it is not actually part of the current path.

 

[JohnRoberts]

I think the point being made is that whilst those contact / wire impedances are real and can be shown to give the mathematical results you state, the actual connection to the Earth is irrelevant in the circuit shown. The wire / contact resistances exist without the connection to Earth.
Of course, if you were not making a connection to Earth you would not need that wiring and the loop would not be formed (as for unearthed Class II equipment).
But @Speedskater 's point is worth noting in general as there is often a good deal of misunderstanding the Earth as having some special significance in somehow "magically" absorbing or draining current when it is not actually part of the current path.

In the big picture lightning usually finds its way to earth (except for cloud to cloud strikes).

Small picture grounding deals with current flows and proper signal referencing.

JR

 

[sahib]

.....the actual connection to the Earth is irrelevant in the circuit shown.

Buddy, please. I will be banging my head now. I have a test result up there clearly saying that it does and you are still going on saying it does not. Please. Go back to my first post and read the result at the end. It is a very simple voltage dividing situation. If the mains earth impedance did not make any difference then I should be reading 0.0022V. Instead I am reading 0.007V. What else can I say.

Of course, if you were not making a connection to Earth you would not need that wiring and the loop would not be formed (as for unearthed Class II equipment).

Indeed but we are not dealing with that at the moment.

But @Speedskater 's point is worth noting in general as there is often a good deal of misunderstanding the Earth as having some special significance in somehow "magically" absorbing or draining current when it is not actually part of the current path.

Indeed there is no magic and no such thing as absorption, but if you are saying no current flows through the mains safety earth then I would respectfully ask you to revisit your mains earth knowledge. That is the purpose of mains safety earth, to provide low impedance path to ground so that one would not get zapped if there was leakage into the equipment chassis (say if neutral was broken or lightning). In return, if, say harmful or fault currents flow into the safety earth so does any other current.

When people think of earth wire they often think of mother earth. That's in a way true but the mains safety earth wire is connected to neutral at mains distribution and grounded. I have checked the continuity between the two at the office, and also the resistance which read around 120 milliohms. Of course this is here in the UK. I have no knowledge of US mains system.

However, just as a general info, I have also tested some IEC leads today. I have tested UK, US and Europe (Schuko). I have plugged them into a new IEC chassis plug an measured between the plug earth prong and and IEC prong to take the contact resistances into account. They were all fully moulded types.

New (straight out of the wrapping)

UK around 54 milliohms
US around 46 milliohms
Schuko around 75 milliohm

I have also tested used ones that I had lying about. These were left overs from my previous company. I measured four of each. here are the results.

UK varied between 189 milliohms to 890 milliohms.
US varied between 479 milliohms to 1.69 ohms (ouch!)
I had only one used Schuko and that measured 128 milliohms.

The measurements are of course dominated by the contact resistances,

This first tells me that one should be checking used ones as they can easily go out of legal spec. I was surprised with the US types as they were used but nowhere near as much as UK types.

So, there is no doubt that any improvement to mains cable and connectors will be beneficial. But in terms of cable resistance (hence the cable thickness) it will be completely swamped by the connector contact resistances. So, I would spend more effort on the connectors than the cable.


EDIT: Well, it would be a mature thing to accept the defeat if it is the case. So, I might indeed be wrong and the answer is in the mains lead measurements I made above. I will check the mains lead resistance of the function generator.

 

[Newmarket]

When people think of earth wire they often think of mother earth. That's in a way true but the mains safety earth wire is connected to neutral at mains distribution and grounded. I have checked the continuity between the two at the office, and also the resistance which read around 120 milliohms. Of course this is here in the UK. I have no knowledge of US mains system.

Indeed this is where things get "interesting" or to put it another way - disappear into the spark's world of TN,TS,IT, earthed distribution, unearthed distribution etc etc.

However, just as a general info, I have also tested some IEC leads today. I have tested UK, US and Europe (Schuko). I have plugged them into a new IEC chassis plug an measured between the plug earth prong and and IEC prong to take the contact resistances into account. They were all fully moulded types.

New (straight out of the wrapping)

UK around 54 milliohms
US around 46 milliohms
Schuko around 75 milliohm

I have also tested used ones that I had lying about. These were left overs from my previous company. I measured four of each. here are the results.

UK varied between 189 milliohms to 890 milliohms.
US varied between 479 milliohms to 1.69 ohms (ouch!)
I had only one used Schuko and that measured 128 milliohms.

The measurements are of course dominated by the contact resistances,

This first tells me that one should be checking used ones as they can easily go out of legal spec. I was surprised with the US types as they were used but nowhere near as much as UK types.

So, there is no doubt that any improvement to mains cable and connectors will be beneficial. But in terms of cable resistance (hence the cable thickness) it will be completely swamped by the connector contact resistances. So, I would spend more effort on the connectors than the cable.


EDIT: Well, it would be a mature thing to accept the defeat if it is the case. So, I might indeed be wrong and the answer is in the mains lead measurements I made above. I will check the mains lead resistance of the multimeter.

That is the purpose of mains safety earth, to provide low impedance path to ground so that one would not get zapped if there was leakage into the equipment chassis (say if neutral was broken or lightning). In return, if, say harmful or fault currents flow into the safety earth so does any other current.

Yes - the safety earth conducts fault current - thus unbalancing the L & N currents and tripping an RCD function (which may be a stand alone RCD or incorporated in an RCBO. (Different terminology may be used in other territories)
But fault current flows only when there is a errr...fault.
In normal operation there is no current in the safety earth conductor.
Okay - that's not actually true. There is always a finite amount of leakage current esp from emc filters.
And if it's connected to another piece of earthed kit we have the potential for our old friend ground loop current between chassis. But it does little to no harm if it doesn't flow in the audio ground.

When people think of earth wire they often think of mother earth. That's in a way true but the mains safety earth wire is connected to neutral at mains distribution and grounded. I have checked the continuity between the two at the office, and also the resistance which read around 120 milliohms. Of course this is here in the UK. I have no knowledge of US mains system.

Indeed this is where things get "interesting" or to put it another way - disappear into the spark's world of TN,TS,IT, earthed distribution, unearthed distribution etc etc. Disclaimer: I may have watched too many electrician videos on YouTube

 

[Speedskater]

In the US, the 'Safety Ground' or EGC (Equipment Grounding Conductor) is connected to the Neutral at the building's main breaker box. The EGC's main purpose is to trip the circuit breaker if there is a 'Ground Fault' (short circuit from Hot to chassis). That there is continuity from the EGC to Neutral to Planet Earth has little to do with this function. A secondary function of the Neutral to Planet Earth connection is to keep the Neutral near Planet Earth's potential. (say under 25 Volts)

 

[JohnRoberts]

I haven't been paying close attention but I suspect a mains lead with 1.69 ohms in the EGC (ground) conductor would be unlikely to receive safety agency (UL?) approval.

It has been decades since I was involved with this for my day job, but as I recall the safety agency "ground bonding" test dumped something like 50A into the ground path, and required only single digit voltage rise from IxR.

50A times 1.7 ohms will generate a more than a single digit voltage rise and fail UL safety testing.

JR

 

[sahib]

Firstly the error in my test for whether the mains earth affects loop impedance or not was indeed due to the mains lead of the function generator. I learned a good lesson. So, I was wrong. It does not affect the ground loop impedance. Just to clarify again that when I say ground loop impedance I am not defining ground loop. but simply the ground path that I sketched out in my diagram above.

However, before continuing with the rest of the stuff I would like to mention that I have initially amended my first post with regards to Thor. Whatever happened has happened and I would not like to go there. I am taking my share of the blame by dismissing his mains cable improvements and not concentrating well. Put it onto Covid that I had. He is welcome to comment if he wishes to. My motto has always been that there can be a good thing coming out of bad, and this has been an opportunity for me to look into something that I would normally have not done so.

I will also be amending my first post with regards to the errors. I will not delete any text. I will cross them over and update with the correct info.

However, I have made a bit of investigation on some hi-fi equipment that I had in my clutter shelf today. These all have figure of 8 or cable terminated mains power entry. So, older Class 2 and with no safety earth. I measured the DC resistance on the chassis between the mains entry point (where normally IEC and earth stud point would have been) and the ground/screen tab of the input/output RCA connector. Essentially I measured DC resistance over a straight line. For this see below diagram.

1. Rotel power amplifier:

On this amp the RCA connector has no electrical connection to the equipment casing/chassis. So the resistance on the chassis on a straight line between the RCA ground and the point near the mains entry measured around 9.5 milliohms. On the other hand, I opened the casing and measured the DC resistance between the ground input connection post on the PCB and RCA and it measured around 220 milliohm.

2. Technics power amplifier:

This actually is a modern surround amplifier that I had rescued from our local civic amenity a couple of years ago. I could not believe why would anybody throw away a thing like that.
Anyhow, on this one the RCA connectors have electrical connection to the equipment case. Again the resistance on the chassis on a straight line between the (input) RCA ground and the point near the mains entry measured around 69 milliohms. I have not taken any internal measurements on this.

3. Technics CD player:

This is a bit older equipment that was given to me by our downstairs neighbour. Again the RCA has an electrical connection to the chassis, and the resistance on the chassis on a straight line between the RCA ground and the point near the mains entry measured around 15.6 milliohms.

4. Turbosound crossover:

This is (naturally) a balanced equipment with IEC mains power entry. Surprisingly the PIN 1 on the input XLRs are not connected to the chassis at all. I would assume the reason is to eliminate possible ground loops to form. But there can be a price to pay for that too of course. However, the resistance between the earth tab on the IEC and PIN 1 on the output XLRs measured around 89 milliohms.

The distance between these measurement points are below 50cm. So, I measured the resistance of a 50cm length ordinary single strand 0.8mm2 hook-up wire, which gave about 16.6 milliohms.

I also had a couple of reels of medical grade 1mm2 multi strand Alpha hook-up wire that were used on a control system that I had designed (20 years ago) for an industrial dishwasher prototype. 50 cm length measured 9.65 milliohms.

Now, you'll probably anticipate where I am going with these wires. It seems clear that even on old Class 2 audio equipment the internal ground paths (Ground Path A and A' on the diagram in my first post) can be vastly reduced in impedance. Of course such modification would warrant compliance. So, why not replace mains entry with IEC for Class 1. Particularly on Rotel. The chassis resistance will shunt the actual ground path and give a better path impedance figure.

In terms of external reduction method for Class 1 by using low impedance earth wire, as I mentioned before, it is clear that the contact resistances dominate. I have no idea on what sort of premium connectors are available and their contact resistances are. I'll need to investigate.

As for the signal connections between two (ore more) equipment, an external low impedance wire in parallel to the screen as in the link given by Newmarket also clearly brings considerable improvement.

Finally, credit should be given where it is due. In my case to Thor for paving way to the subject.

EDIT: The first post became too cluttered. I have edited it completely but with an amended diagram.

 

[Speedskater]

OK, try this on each unit:
1] Safety Ground/Protective Earth attached to chassis near where AC power enter the chassis.
2] DC supply common attached to chassis (at a single point) near the signal input connectors.
3] Audio circuit common attached to chassis at that same single point.
4] each cable Screen/Shield attached to chassis at it's connector.

 

[sahib]

I am not sure what you mean by each unit. Is this for h-fi or pro-audio application? If it is for hi-fi application I am not involved in that line. But for pro-audio that is the standard practice that I have been doing since I started doing pro-audio (about 15 years ago).

However, I do not intend to get religious about this, but at the same time there is definitely something there in terms of increasing performance and I would like to experiment. The only problem is that I do not have a proper listening set-up. So, at the moment all I can do is to do it on paper, see how the maths work out and discuss it with the heavy hitters of the forum.

I have googled hi-fi connectors etc and came across so called high-end brands like Cardas and ETI but none of them seem to have any technical data on their IEC plugs/sockets. They are very fancy looking things with all singing and dancing chrome plated covers and so on, and cost £150 - £160, yet Bulgin with less than 10milliohm contact resistance costs under £4. Same goes for the cables. Cardas has a say ground cable at 17.5AWG (around 1mm2) and I would like to know what makes it so high-performance than a good quality standard 1mm2 wire.

Perhaps experienced members could chime in with what cables/connectors they use.