GoodSoundClub - Romy the Cat's Audio Site - Ah, the common mode noise!

11-24-2015

Post does not mapped to Knowledge Tree

JOHNR
Posts 24
Joined on 10-12-2015

Post #:	1481
Post ID:	22258
Reply to:	22255

MDB grounding

Paul,
Please be careful here because good ol' USA is quite different when using residential pole mounted feeder transformers.
What you may have is a neutral wire coming off of the 0 volts tap of the 120-0-120 secondary. This center tap is connected in turn to a ground rod using a substantial conductor. This is designed to handle a transformer failure in the event of the primary coming in contact with the secondary. This happens due to surges or lunatics stealing the above mentioned ground rod. The source voltage ranges from 9 to 15 kilovolts on those installations that I have come across.

The DB ground is most likely to be a ground rod located near the main cable entrance.
I would prefer to use this ground path simply because no matter what happens to the commercial supply, surge energy cannot get to this path.

In the case of Europe, this DB ground comes from a tap taken from the incoming cable armored sheath. This is designed to handle fault currents only.
It used to be grounded only at the feeder transformer but this has been relaxed but you are not allowed to fit your own ground. Besides, it is incredibly noisy.

The fundamental reason for the different methods used is the sheer length of such feeder systems. The power companies can support large areas using a single high voltage wire. The trade off is that there is no local voltage regulation and why hifi is such a problem when it needs clean power. Moving to SMPS based systems does away with all that as an issue.
Regards
John

11-24-2015

Post does not mapped to Knowledge Tree

Paul S
San Diego, California, USA
Posts 2,656
Joined on 10-12-2006

Post #:	1482
Post ID:	22259
Reply to:	22258

Center Tap Potential vs Safety

Yes, John, US residential service is center tapped, with a "true ground", ideally 0 Volts, a legacy of Thomas Edison's old DC system. And obviously, with split single phase, we also need to allow for load imbalances. It is not until just now that I realized that the European standard "neutral" was so lax, and put in the context of lightning, it would indeed make one think twice about creating the best ground path for one's apartment block!

So, why is it that Europeans on this board claim that their electricity is seldom (or never) a problem for them?

And what of "galvanic" issues in the "absence of ground"?

Best regards,
Paul

11-24-2015

Post does not mapped to Knowledge Tree

rowuk

Germany
Posts 454
Joined on 07-05-2012

Post #:	1483
Post ID:	22260
Reply to:	22259

Double the voltage, half the amperage, half the noise?

Maybe we in Europe benefit from 230V with half the amperage and not the grounding scheme? Seems to me that line noise would scale down by factor 2 when we get 15V from 230V instead of 110V when using "analog" transformers.

Paul S wrote:
Yes, John, US residential service is center tapped, with a "true ground", ideally 0 Volts, a legacy of Thomas Edison's old DC system. And obviously, with split single phase, we also need to allow for load imbalances. It is not until just now that I realized that the European standard "neutral" was so lax, and put in the context of lightning, it would indeed make one think twice about creating the best ground path for one's apartment block!

So, why is it that Europeans on this board claim that their electricity is seldom (or never) a problem for them?

And what of "galvanic" issues in the "absence of ground"?

Best regards,
Paul

Whenever I feel blue, I start breathing again.

11-24-2015

Post does not mapped to Knowledge Tree

Paul S
San Diego, California, USA
Posts 2,656
Joined on 10-12-2006

Post #:	1484
Post ID:	22261
Reply to:	22260

Loop Gain vs. "Clogs"

Of course, some of the noise heard is due to system gain, including loops. But it seems like some of my electricity problems are a sort of "clogging up" of the system, like a hotrod running on cheap gasoline. Basically, the system is "sluggish", and it "under-performs", and it's not just a matter of "power", it's what the power can or can't do. Maybe this is a kind of "interstitial" noise that taxes the system. I don't know what it is, but I know it when I hear it. Some claim fewer issues when they use 240V power conditioners that drop the user voltage to 120. I've been in studios that used these, but I've never used one.

Paul S

11-25-2015

Post does not mapped to Knowledge Tree

N-set
Gdansk, Poland
Posts 617
Joined on 01-07-2006

Post #:	1485
Post ID:	22263
Reply to:	22255

NG bond

Paul S wrote:
N-set, the "neutral" wire is certainly referenced (connected) to ground. The usual "ground" wire is redundant, and this is why it often causes problems with hi-fi, also why a dedicated bleeder that does not go to the mains ground will sometimes do the trick for "residual" hum.

Definitely is, if I understand correctly somewhere at the power plant. This must be taken into account when designing
equipment IMHO-->R. Morrison's classics on the grounding.
We tried to find the NG bound with my electrician. It's neither in my panel, nor in the main power entrance to the building
(I have a luck of a "dedicated" run to my apartment directly from the main house panel, ommiting all other apts).
I'm not sure what the practise is to get the ground in my region, but John seems to shed some light:

JOHNR wrote:
In the case of Europe, this DB ground comes from a tap taken from the incoming cable armored sheath. This is designed to handle fault currents only.
It used to be grounded only at the feeder transformer but this has been relaxed but you are not allowed to fit your own ground. Besides, it is incredibly noisy.

As for CMBR I intentionally resigned from it after an approval of the electrician. I want just a simple old-style fuse (costing me a fortune since I wanted Siemens,...)

Cheers,
N-set

Cheers,
Jarek
STACORE

11-27-2015

Post does not mapped to Knowledge Tree

Romy the Cat

Boston, MA
Posts 10,155
Joined on 05-28-2004

Post #:	1486
Post ID:	22265
Reply to:	2931

Two or more cross-taking switching power supplies?

As I reported before I have problem with two PP in my home to work at the same time. I have my main audio system all plugged in PP3000+ and I have my video/audio system in opera room running off PP3000. The use of PP3000 in my Opera Room is very essential and I did make a number of efforts to stop using PP3000 in Opera Room but each time the PP was drifting back as the difference is too huge to ignore in both audio and video. The problem that I have with this configuration that I have two PP3000 running in the same home and I create very recognizable by me sonic problem. So, basically when I listen more or less critical my main audio system I always forced to shut down my Opera Room regenerator. I have to do it in reveres as well but I seldom do and to power down the audio PP3000 is a big deal: DAW and few other all time on devised are plugged into it.

So, I wonder if it is possible to find some solution that would prevent the PPs to inflict each other. To hook them up on the different sides of 220V is not a good idea. I was trying to decouple the Video PP with 5kW isolation transformer. The impact to Audio PP was substantially minimized but the sound on Video PP was not as good as without the transformer. Perhaps I need a different isolation transformer? I do not want to get re engage into the games again when I was looking for something and I do not know what I need to look for. Did anybody has any practice of dealing with two or more cross-taking switching power supplies in the same system?

"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche

11-28-2015

Post does not mapped to Knowledge Tree

Paul S
San Diego, California, USA
Posts 2,656
Joined on 10-12-2006

Post #:	1487
Post ID:	22266
Reply to:	22265

Less Than Zero

"Theoretically", it should work to precede all hi-fi gear with a balanced type 120V power supply, where the ground is defined as the difference between two 60V legs that are opposite phases (ie., plus and minus 60V), with a perfectly centered tap that functions as the neutral (ground reference). This "should" block or cancel any noise stemming from "incidental" capacitive discharging along with the reactive noise caused by component leakage, and it "should" also keep the AC supply at its practical maximum, since none is wasted on this sort of noise. Of course, the devil is in the details, and this sort of balanced system has to remain perfectly balanced into any load. One benefit from a "properly functioning" balanced AC system might be that one no longer needs a regenerator.

For fun reading, start at Equi=Tech, also check out similar Furman units.

Paul S

11-29-2015

Post does not mapped to Knowledge Tree

JOHNR
Posts 24
Joined on 10-12-2015

Post #:	1488
Post ID:	22267
Reply to:	22265

Dual PP3000 units

This is classical common mode noise interference on a common supply. This type of noise will pass right through any power conversion technology and in both directions. Due to the timing variations, I would expect it to be a little bit of mains interference and a lot of cellular RFI.

Question -
Do these units have front end power factor correction? If they do not, then you need to search for products that do this.

I note from earlier posts that suggestions to fit dedicated filament supplies to tubes will help. I can help with that but you will have to wait a while.
One aspect of both small signal and large DHT tubes is that RFI can pass through the filament and be amplified by a factor of 1.

The supply should be a.c. only since it is the diodes that create the common mode noise. You can of course invest in a CMChoke on every tube if you wish starting with the high gain tubes.

I saw your post showing the hand drawn curly "snakes" through the main waveform but never saw any answers to curing what is a 5 meg RF waveform mixed in with the 60 Hertz. Cell traffic is peaky at certain time during the day explaining why it varies so much. If the interference is in fact RFI and not mains borne, it will give your sound the appearance of greater brightness and having a gritty characteristic.

Suggestions:
Suggest you try inserting a 5 amp rated common mode choke firstly on the Opera network. Look for a 3 to 5 milli-Henry choke

Fit a 1 uF/440 volt capacitor each side of this choke and see what happens. I use MKP1848 rated at 900 volts for this application.
If that works but needs more, add a pair of 33nF "X" type capacitors at the center of the filter; one on the 220 rail volt and the second on the 0 volt rail. Bring the legs together and connect to a ground point using thin, single conductor wire.

On the low power units e.g. DAC, install a 50uF oil filled motor run cap in series with the line side of the power feed to the DAC. This removes most broad band 5 to 8 meg of noise without spoiling the sound. This will help if you have lots of cellular traffic in the area. Does not require any ground path, it absorbs the RF.

Another question - what filtration does the PP3000 have on its output rails? As an example, does it have a common mode choke?

I saw you comment about manufacturers fitting too much filtration on their products. My experience is they fit only what Regulators such as TUV and the UL underwriters see fit. These standards are FAR TOO LOW for ultra fi.
John

11-29-2015

Post does not mapped to Knowledge Tree

Romy the Cat

Boston, MA
Posts 10,155
Joined on 05-28-2004

Post #:	1489
Post ID:	22268
Reply to:	22267

Ah, the common mode noise!

Thanks, John, I sink it might be call common mode noise. My concern about calling it common mode noise derived from the fact that it is not normal megahertz common mode noise but artificially induced common mode. The PP3000 operate at around 30kHz sampling rate , it means that common mode noise will be injected at roughly the same frequency. My presumption was that the other PP3000 shall be able to filter out the own sampling rate by venture that the noise samples are not shorter then own samples, I am not sure if I am write but this is how I see it. Anyhow, if it is the common mode problem then indeed to put a good common mode choke would be effective way to deal with the problem. I would not go for shunting the side with caps and balding a differential filter as according to my experience a presence of any shunting cap on AC side damage more than helps.

"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche

11-30-2015

Post does not mapped to Knowledge Tree

JOHNR
Posts 24
Joined on 10-12-2015

Post #:	1490
Post ID:	22269
Reply to:	22268

PP3000 and 30KHz

Romy,
The common mode noise will lie between 5 meg and 8meg or 8 meg to 12 meg depending upon whether it is forward or backward mode working.
It does not have much to do with the 30Kilohertz.
It is possibly true that one unit may have filters to keep out such noise and that they work in both directions thus should also reject such emissions from a second unit. In real life, I doubt that - refer to the following text.

The regulatory system publishes a "mask" that a machine must comply with before being given an approval. The problem is that the mask is not adequate for high level audio. As an example powering a second dac whilst listening to music will affect the music and that in itself proves that they do interact.

The mains filters are basically inadequate.

If you are concerned that placing a 1uF cap across the mains spoils the sound then that surprises me because the cap by itself will alter the lagging power factor in a system by a miniscule amount. You should be able to add 100uF to bring the phase angle up to 95% lagging without altering the sound.
It looks to me as if your power factor is being corrected to a leading PF somewhere else in your system.

BUT, if your PP3000 unit operates with an internal power factor correction front end, the front end filtering will be minimal because the PF will be 95% or better at all times, but still lagging.

The least you should do at this stage is to try the "Y" network of 2 caps. These will sum the opposing voltages thus will self cancel. If this reduces the problem then the nature of the interference is essentially proven. The cap value is not in the audio passband therefore should not affect the audio.

A second suggestion is to fit a clamp on ferrite to the power cable on each unit. If no difference, then RFI is not your problem.

Question: are you by chance connecting each PP3000 to a single external ground rod?

John

11-30-2015

Post does not mapped to Knowledge Tree

Romy the Cat

Boston, MA
Posts 10,155
Joined on 05-28-2004

Post #:	1491
Post ID:	22270
Reply to:	22269

The grounds on loading side might be the key.

JOHNR wrote:
The common mode noise will lie between 5 meg and 8meg or 8 meg to 12 meg depending upon whether it is forward or backward mode working.
It does not have much to do with the 30Kilohertz.

The normal common mode noise is megahertz level noise and it of cause will go through any switching filtration like hot knife through butter. Since I observe the ELEVATED common mode noise 9if we presume that it is common mode noise) then it would make scene to recognize PP3000 as course of the common mode noise. It is need to be pointed out that there is hardly anything in PP3000 fasters then PP3000. That was all that I was saying.

JOHNR wrote:
If you are concerned that placing a 1uF cap across the mains spoils the sound then that surprises me because the cap by itself will alter the lagging power factor in a system by a miniscule amount. You should be able to add 100uF to bring the phase angle up to 95% lagging without altering the sound. It looks to me as if your power factor is being corrected to a leading PF somewhere else in your system.

I played a LOT of filtration on AC side and although I was able to get much auditable noise does it has very little impact to what I was hearing and it ALWAYS get worse sound (particularly in very lower bass) in case I get any capacitance on AC side. Regarding the power factor. I am sure that PP have power factor correction circuit at input.

JOHNR wrote:
The least you should do at this stage is to try the "Y" network of 2 caps. These will sum the opposing voltages thus will self cancel. If this reduces the problem then the nature of the interference is essentially proven. The cap value is not in the audio passband therefore should not affect the audio.

Do you means the "Y" network of 2 caps to the ground? I would not run it in real time but I think as proof of consent it will work wonderfully.

JOHNR wrote:
A second suggestion is to fit a clamp on ferrite to the power cable on each unit. If no difference, then RFI is not your problem.

I have it for years.

JOHNR wrote:
Question: are you by chance connecting each PP3000 to a single external ground rod?

The loads are isolated from source ground and use own grounds. I did try to kill grounds on loading side and make the system to float and it had no impact.

"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche

12-01-2015

Post does not mapped to Knowledge Tree

JOHNR
Posts 24
Joined on 10-12-2015

Post #:	1493
Post ID:	22272
Reply to:	22271

PFC correction

PFC is a function of adding Xc using capacitors to bring the current in the circuit downwards.
What you have is an inductive load giving a lagging PF eg a simple motor gives a PF of 0.85.
Thus the inductive element can be manipulated to bring it from lagging to leading. The economic target is 95%.

If you step back from the needs of an audio system, if applied properly, the current in a large supply network can be reduced quite considerably. In fact above certain loads, your power company will insist on implementing PFC.

The best way to observe this at home is to connect a clamp on meter to the line side of the supply and read off the current. For the purposes of this test, assume 10 amps.
Now add an 8uF capacitor, motor run type anywhere convenient, oil filled for cool running and watch the 10 amps go down to 8 amps.
Add a second cap and watch it go down to 6 amps. By adding more caps a point will be reached when the current increases. The PF will now be leading which is not good. Back off and aim for a 10% current reduction.

An example of this is if you have push pull amp, any technology will do.
Equip the socket with an 8uF cap and the sound will improve. Adding a second one may also help but be careful as the cap draws current.

It is not trickery, it is engineering. It is not done on small loads because the capacitive bank needs to be connected according to the load. You cannot leave it in because you will have to pay for the energy usage. In the case of the PP example, you turn on the amp then turn on the cap.

John

12-01-2015

Post does not mapped to Knowledge Tree

JOHNR
Posts 24
Joined on 10-12-2015

Post #:	1494
Post ID:	22273
Reply to:	22271

Ferrites

A clue!
You are not using the ground on the PP3000 unit and have ferrite packages on the incoming mains supply.

I am guessing that you have fitted a clamp directly around the power cord.

If this is the case, then you are using the ferrite on the ground conductor as well which is NOT good.

First off, a ferrite should never be used on the PE system. Why? Because it will act either as a ferrite that increases the impedance of the PE conductor which is NOT good OR, if the Q is correct, will absorb noise from the PE lead as intended BUT it may well transfer PE noise to the L and N conductors.

When using ferrites, one must consider the fact that if it manipulates noise, where does this noise go to?

Consider breaking into the power cord and retrieving the PE conductor to create an external loop for the L and N. Fit a clamp to the L and N only.

Never fit ferrites to ground wires unless you know the noise frequencies and also know the suitability of the ferrite device and how it responds to those frequencies. Peak up or absorb. They are not a broadband device as such.

Going back to the quoted Y network. Use 2000 volt rated Y capacitors, value 33nF, one per rail, join and using a single solid core wire e.g. Tefzel silver over copper to a convenient ground point.

John

12-01-2015

Post does not mapped to Knowledge Tree

Romy the Cat

Boston, MA
Posts 10,155
Joined on 05-28-2004

Post #:	1495
Post ID:	22274
Reply to:	22273

Sometimes patient dies from a cure.

John, what you say about power factor correction is completely accurate and very much makes sense. There is a many ways to do it BTW. You can do it with caps compensating the inductive load, you can do it with filtration or you can add active SS circuit that would engage the caps at the very specific time and kind of fake activeness of the load. They all work very well and for sure will do benefit the efficiency of your home grid and make you to draw less electricity. None of them work perfect (unless you are spinning an idle synchronous motor purely for the sake of power factor correction) as load go up and down and amount of current is not stable, the hot tubs pumps and heaters goes on and off, the freezers go on and off, the thermostat turn ether AC or boilers, lights and everything else goes on and off changing in power factor landscape. (BTW, I do not think light do as they are pure resistive load). In case you have a SET amp then you for sure can heat it up as it thermos-stabilized and begin to draw very constant current then you can implement a perfect power factor correction for this amp only. Here however we face a dilemma: adding a cap to AC line always worsen sound, primary the very lower bass and LF’s dynamic. 0.2-2uF was very auditable. So I do not know if I welcoming the idea of power factor correction for audio. You might argue that I already have plenty of caps siting on my power lines and I would not disagree with you but I only presume that less would it be the better sound I would get. Saying all of it, I do think that it might makes sense to play with PPC on the PP3000 source side.

Regarding the ferrite packages on the incoming mains supply. That was a brilliant observation. I do float ground on the supply side, only on load side. Still, the idea of differentiating of ferrite rings between the power cord’s hot, neutral and ground is something that very much worth attention.

Rgs, Romy the Cat

"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche

12-04-2015

Post does not mapped to Knowledge Tree

JOHNR
Posts 24
Joined on 10-12-2015

Post #:	1497
Post ID:	22285
Reply to:	22274

Fitting Ferrites for noise control

Romy and Clarke

A couple of tips here:
To be truly effective, you need at least 3 turns of L+N through the ferrite; more if you can.

The "Y" network can be placed either side depending upon where the PP3000 to PP3000 interference is coming from.
You may need one each side.

In really bad situations, you may need to use two different ferrite types.
A ferrite with "Y" cap will give you 12dB whilst a second pair will give 18dB after that, it is a waste of money.

The discussions to date have been about finding and treating high frequency noise and not about PFC correction as such.

One interesting fact that has been discovered about commercial SMPS unit is that the regulations require a 1 meg ohm resistor to discharge the "X" caps when the plug is pulled out. The cap has no direct discharge path. Changing this resistor can change the sound. How? I don't know.

Regards
John

12-04-2015

Post does not mapped to Knowledge Tree

Romy the Cat

Boston, MA
Posts 10,155
Joined on 05-28-2004

Post #:	1498
Post ID:	22286
Reply to:	22285

Will be looking into it.

John, thank you very much. You did point out the area when I did not explored deep enough and I will certainly will experiment with it. I need to educate myself about ferrite types and about proper ways to form the ferrite coils. Up to now I used ferrite whatever I was able to get to fit the cable. Indeed it might be with to experiment with it deeper.

"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche

	Target	Threads for related reading	Most recent post in related threads	Forum	Replies	Views	Started
»	New	What lives in Symmetric Sound?..	The beginning of our journey is ALWAYS symmetrical...	Audio Discussions Forum	19	175213	05-28-2004
»	New	Always check power-line polarity...	The Cost of Knowing...	Audio For Dummies ™ Forum	11	112297	07-10-2005
»	New	RAAL “Water Drop” tweeter for Macondo...	Your comment takes me by surprise...	Horn-Loaded Speakers Forum	77	924584	02-16-2007
»	New	My feelings about new exciting audio products..	Vacuumstate...	Audio Discussions Forum	25	264743	04-30-2007
»	New	Musique Concrete horns..	These are now sold as Kornhent products...	Horn-Loaded Speakers Forum	6	108258	06-12-2007
»	New	Compression drivers and the “clean signal”...	The NEW “Compression drivers and the clean signal”....	Horn-Loaded Speakers Forum	14	163515	07-12-2007
»	New	Digi Redux; Drive 1 transport and iDAT-44+ DAC..	Moray James SPDIF!...	Didital Things Forum	27	230576	09-28-2007
»	New	Metal domes..	Try the one Lansche is using...	Audio Discussions Forum	6	78946	11-08-2007
»	New	The power AC Outlets?..	Where to Pick Up the Gong?...	Audio Discussions Forum	2	43192	10-31-2008
»	New	The Avicenna's failure is the great Avicenna success!..	New life for Avicenna...	Audio Discussions Forum	8	83926	02-03-2009
»	New	Internet and electricity..	Suboptimal. . ....	Didital Things Forum	1	29349	01-07-2010
»	New	Electricity... power strips and ac improvements..	Electricity... power strips and ac improvements...	Audio Discussions Forum	0	16687	03-30-2010
»	New	Another example of energy..	Tehran 230v...	Audio Discussions Forum	1916	9941625	01-29-2011
»	New	I good spot-light for a turntable?..	Reply...	Analog Playback Forum	15	154661	10-24-2010

	Target	Threads for related reading	Most recent post in related threads	Forum	Replies	Views	Started
»	New	What lives in Symmetric Sound?..	The beginning of our journey is ALWAYS symmetrical...	Audio Discussions Forum	19	175213	05-28-2004
»	New	Always check power-line polarity...	The Cost of Knowing...	Audio For Dummies ™ Forum	11	112297	07-10-2005
»	New	RAAL “Water Drop” tweeter for Macondo...	Your comment takes me by surprise...	Horn-Loaded Speakers Forum	77	924584	02-16-2007
»	New	My feelings about new exciting audio products..	Vacuumstate...	Audio Discussions Forum	25	264743	04-30-2007
»	New	Musique Concrete horns..	These are now sold as Kornhent products...	Horn-Loaded Speakers Forum	6	108258	06-12-2007
»	New	Compression drivers and the “clean signal”...	The NEW “Compression drivers and the clean signal”....	Horn-Loaded Speakers Forum	14	163515	07-12-2007
»	New	Digi Redux; Drive 1 transport and iDAT-44+ DAC..	Moray James SPDIF!...	Didital Things Forum	27	230576	09-28-2007
»	New	Metal domes..	Try the one Lansche is using...	Audio Discussions Forum	6	78946	11-08-2007
»	New	The power AC Outlets?..	Where to Pick Up the Gong?...	Audio Discussions Forum	2	43192	10-31-2008
»	New	The Avicenna's failure is the great Avicenna success!..	New life for Avicenna...	Audio Discussions Forum	8	83926	02-03-2009
»	New	Internet and electricity..	Suboptimal. . ....	Didital Things Forum	1	29349	01-07-2010
»	New	Electricity... power strips and ac improvements..	Electricity... power strips and ac improvements...	Audio Discussions Forum	0	16687	03-30-2010
»	New	Another example of energy..	Tehran 230v...	Audio Discussions Forum	1916	9941625	01-29-2011
»	New	I good spot-light for a turntable?..	Reply...	Analog Playback Forum	15	154661	10-24-2010