Leakage Currents in audio

[jkeny]

I see low frequency, low impedance noise & interference current getting mixed with high frequency, high impedance noise & interference voltage. These are two different problems and require two different solutions.

Please explain! You mean you have seen this in measurements or are seeing a confusion here?
What different approaches do you recommend for these different issues?

Swenson does say

"Remember the leakage from an SMPS contains both high impedance and low impedance components, the ground shunt gets rid of the high impedance components. The low impedance components are still left. These components are usually still greater in amplitude than many linears, but some linears might be a bit more.

BUT since what is left is the low impedance components, they CAN easily be blocked."

 

[Speedskater]

Please explain! You mean you have seen this in measurements or are seeing a confusion here?
Swenson does say

I see it in this and other threads often in other forums. Different problems and different situations all mingled together.

* * * * * * * * * * * * * * *
As for measurements, they need to be made at the loudspeaker terminals of the power amp. Best if done with a battery powered true RMS AC DMM (or lap-top with a good ADC card) and a battery powered O-scope.

 

[jkeny]

I see it in this and other threads often in other forums. Different problems and different situations all mingled together.

* * * * * * * * * * * * * * *
As for measurements, they need to be made at the loudspeaker terminals of the power amp. Best if done with a battery powered true RMS AC DMM (or lap-top with a good ADC card) and a battery powered O-scope.

I guess there may well be confusion but as I said, my focus is on PS stability which also implies ground stability & anything that disturbs this I consider detrimental to SQ, no matter what frequency or what name it is given.

 

[RogerD]

I guess there may well be confusion but as I said, my focus is on PS stability which also implies ground stability & anything that disturbs this I consider detrimental to SQ, no matter what frequency or what name it is given.

John, you are correct in my experience with chassis grounding. Since you are bonding components in a star grounding configuration,any internal or any loops created should be mitigated,maybe not fully though.
Also the parallel circuit that intersects the IC's will effect the audio signal SQ. I really don't understand how The IC circuit is not reset by these ground boxes. Since the impedance of the entire circuit is effected. In my experience there is no need to do any grounding involving the speakers,as the audio signal is perfected to a greater extent already.

 

[LL21]

John, you are correct in my experience with chassis grounding. Since you are bonding components in a star grounding configuration,any internal or any loops created should be mitigated,maybe not fully though.
Also the parallel circuit that intersects the IC's will effect the audio signal SQ. I really don't understand how The IC circuit is not reset by these ground boxes. Since the impedance of the entire circuit is effected. In my experience there is no need to do any grounding involving the speakers,as the audio signal is perfected to a greater extent already.

Hi Roger,

Question: I believe that Tannoy is one of the few speaker companies that has had a grounding post on the back for many, many years. I was told that this is because there is a lot of EMI/RFI and perhaps other forms of 'distortion related elements' around speakers, and that grounding removes a lot of hash intrinsic to speakers. Any thoughts on this?

 

[RogerD]

Hi Roger,

Question: I believe that Tannoy is one of the few speaker companies that has had a grounding post on the back for many, many years. I was told that this is because there is a lot of EMI/RFI and perhaps other forms of 'distortion related elements' around speakers, and that grounding removes a lot of hash intrinsic to speakers. Any thoughts on this?

Hi Lee,
I had a good friend that had the large Tannoy monitors...nice speakers. In my system I have no need to do this. As I said above when the chassis grounding has reached a level where it approaches "noiseless" the audio signal to the speakers should be perfected. At that point I really don't think exotic caps used in the passive crossover matter,but that's another subject. When the noise floor is so low that I can hear orchestra players breathe on DG recordings that's good enough for me,but try it,you never know.
I continue to believe that EMI RFi in the general sense is a secondary problem. The current induced noise is the majority of why SQ is degraded.

 

[opus112]

I see it in this and other threads often in other forums. Different problems and different situations all mingled together.

I for one would appreciate it if you'd untangle them, I might learn something useful.

 

[LL21]

Hi Lee,
I had a good friend that had the large Tannoy monitors...nice speakers. In my system I have no need to do this. As I said above when the chassis grounding has reached a level where it approaches "noiseless" the audio signal to the speakers should be perfected. At that point I really don't think exotic caps used in the passive crossover matter,but that's another subject. When the noise floor is so low that I can hear orchestra players breathe on DG recordings that's good enough for me,but try it,you never know.
I continue to believe that EMI RFi in the general sense is a secondary problem. The current induced noise is the majority of why SQ is degraded.

interesting...when i bought my Gryphon Colosseum, it came with VIP Reference powercable and the Gryphon Exorcist (some kind of demagnetizing instrument you attach to an RCA input...whereby it sends a high-pitched signal for 30 seconds). You could actually hear it change if you sent the signal right thru immediately a 2nd and 3rd time...at which point things would sound clearer, more cleaned up.

However, after installing Tripoint/Entreq, i was surprised to find that the Gryphon Exorcist no longer made any differences. I wonder if it is because it is removing the EMI/RFI or whatever from the system already? No engineer here, but it was an interesting discovery. I used to use the Gryphon Exorcist weekly, and now it sits in the drawer.

I hear good things about grounding speakers, but perhaps your observation is a bigger version of my little one about the Gryphon Exorcist no longer being necessary.

 

[jkeny]

Yes, please untangle for me, too

I see it in this and other threads often in other forums. Different problems and different situations all mingled together.

* * * * * * * * * * * * * * *
As for measurements, they need to be made at the loudspeaker terminals of the power amp. Best if done with a battery powered true RMS AC DMM (or lap-top with a good ADC card) and a battery powered O-scope.

 

[jkeny]

I guess that the untangling answers are to be found elsewhere - Swenson is doing the untangling:

SMPS leakage current contains two parts: a very high impedance part that goes through just about anything, including Ethernet transformers and a low impedance part that is easier to block.

And he seems to have identified what causes differences between ethernet switches as far as leakage currents are concerned - he has surmised that how the centre tap of the ethernet transformers is connected to ground determines whether the high impedance part can be shunted to ground

The part that varied from switch to switch is how it deals with leakage coming from OTHER boxes, ie over the Ethernet cable connected to the switch.

My guess is that the transformers in the box have a common connection and THAT goes through the caps, resistors etc to the ground plane. Thus it can go from one channel to the other in the same box, but if it has to go between boxes it HAS to go through the ground plane and that shunts it to ground.

 

[jkeny]

So back to considerations about the various results - what can we glean from some of the grounding (& non-grounding) efforts

- Swenson's measurements themselves are confusing as his plots show a reduction in the low frequency marker @63Hz when grounding the negative pole of the SMPS output? According to his text, this grounding only shunts HF components to ground, why? Why do LF components not get shunted too?

- in RogerD's chassis bonding or PEC, it may be that there are two effects occurring with various degrees of effectiveness - a shunting to ground of high impedance part of leakage currents & some re-routing of low impedance part through heavy gauge star grounded strapping between chassis.

- Don't know what could be happening with the various grounding boxes but it would seem something is happening?

- One other set of measurements we have seen which has a direct bearing on this (& are the reason for Swenson's investigations) - is Amir's measurements of Uptone's devices, which only showed the effect of the SMPS on the measurements. The limitation of these measurements need to be taken into account but we see low frequency, mains related frequency spurs appearing when the SMPS is used.

This all could do with untangling too!!