Power grid interference: What does it look like measured with different tools?

[DavidPiska]

If you're interested to see what power grid interference looks like, take a look at the videos below, or a comprehensive article about it here: https://rdacoustic.cz/en/blog/2021/04/30/f4sx-audio-power-filter/ (explains in detail what I did here and includes background info + why is it important in audio). A few months ago, I posted measurements of power grid interference using a spectrum analyzer. Other measuring techniques include EMI line meters and oscilloscopes.

Measurements using an EMI meter:

Using an oscilloscope:

A closer look:

And using a spectrum analyzer:

This is Europe. I would be interested to see what it looks like on other continents if anybody has any measurements or experience.

 

[Addicted to hifi]

Very interesting.

 

[Peguinpower]

This guy posted his own version of AC line noise test, and says he didnt see a difference in the actual output of a Topping D90 DAC. Im not impressed by this fella, the test if very narrowly focused, but its just as valid as your experiment.

 

[Alrainbow]

Without a reference how does this matter, ?
The filter shown is most likly loading the freq if the line noise centered on the measuring device shown . What other noise is there ? A person like ralph can explain better . Also if you cant see any change on a cheap toppings dac then it matters even less. Ps i own a toppings d90se lol

 

[Republicoftexas69]

One cannot take anything Amir at ASR posts as serious. Science left his forum long ago. His tests are limited in product scope and test detail.

 

[MaxwellsEq]

I would expect any competent electronic designer should be able to design solutions that reject electrical line noise. Perhaps some domestic audio kit has inadequate mains noise rejection, but professional studio kit would be thrown out if it couldn't handle mains-bourne noise properly.

 

[Alrainbow]

Amir has his own thoughts and is done very well. Now conclusions is of course to each there own
not all noise effects over all sound much be Measured. I’ve read many of his detailed articles he is spot on in what he does. It’s the effects that things get off the road for some.
I own the toppings it’s a very good dac in many ways
Now it’s no Lampi H or many others above it’s level of musicality
it’s a twinkle high detail sound. but if we dial it in to make it have more body it’s pretty good.
his view in sound I don’t always agree but no one does here.

 

[Alrainbow]

I would expect any competent electronic designer should be able to design solutions that reject electrical line noise. Perhaps some domestic audio kit has inadequate mains noise rejection, but professional studio kit would be thrown out if it couldn't handle mains-bourne noise properly.

Regulators remove noise but also add there own st times
Line noise reduction can be needed but most don’t
where line noise matters is any device that has a
Large gain phono pre for one
A pre amp not so much and an amp
Less still.
a dac is very complex and has many of its noises to deal with from all directions.
just sayhg

 

[MaxwellsEq]

A mixing desk has high gain microphone stages dealing with low signals. Each channel has sequences of gain, attenuation, switching, gain, attenuation, switching etc.

Each stage has optimised power noise rejection.

If mixing desks were susceptible to mains-bourne noise, all recordings and broadcasts would have mains-bourne noise in them. But they don't. So this proves good designers can deal with power grid interference.

 

[Kingrex]

If you know me, I have been digging around the idea of battery inverter power. Most all these inverters operate wirh a 20KHz sine wave generator. As well as a GHz dc to dc voltage booster on high frequency inverters. All the people using these inverters use a Puritan 156 on the load of the inverter. They say it alters the sound in a positive way.

I watched ASR do a study of the Puritan. Sure enough, the puritan did an excellent job of filtering the 20KHz noise. But using the puritan had 0 impact with the simplistic study Amir did on the output of 2 devices. As in, test 1 direct to wall and test 2 from the puritan show the same output signal from the audio device no matter what it's plugged into it.

This makes me wonder, what noise is coming from the inverter the Puritan is filtering that does get through the power supply of a piece of equipment.

Is Amir a failed engineer that does not know what tools to measure with? What is he missing? Why do people find filters to help. Where are they impacting the AC.

 

[Onepoint5]

If you know me, I have been digging around the idea of battery inverter power. Most all these inverters operate wirh a 20KHz sine wave generator. As well as a GHz dc to dc voltage booster on high frequency inverters. All the people using these inverters use a Puritan 156 on the load of the inverter. They say it alters the sound in a positive way.

I watched ASR do a study of the Puritan. Sure enough, the puritan did an excellent job of filtering the 20KHz noise. But using the puritan had 0 impact with the simplistic study Amir did on the output of 2 devices. As in, test 1 direct to wall and test 2 from the puritan show the same output signal from the audio device no matter what it's plugged into it.

This makes me wonder, what noise is coming from the inverter the Puritan is filtering that does get through the power supply of a piece of equipment.

Is Amir a failed engineer that does not know what tools to measure with? What is he missing? Why do people find filters to help. Where are they impacting the AC.

There's a little more to consider with inverters. My specialty is with HV motor drives, and in the past LV motor drives and a taste of UPS, while manufacturers make up their own topology, there are certain criteria that can't be bypassed

a) The higher the output transistor, IGBT, the lower the efficiency by lots, think Class A amp losses. Typical efficiency of inverters is 98% to 95%.

b) Efficiency of UPS is about 85%, no more really than 5% above that. The conversion from battery to AC output is in two stages: Inverter itself, output filters.

c) The inverters are 99% PWM, use 6 switching devices for three phase, four devices for single phase. That output is square wave essentially. To create a sine wave, the PWM waveform is filtered by inductance and small amount of capacitance. That waveform is not bad, not 100% sinewave best 2% THD.

The filter can reduce/remove, depending on marketing terms, common mode noise.

Where does the common mode noise come from? inherently, the output switching devices interior connections form a capacitive coupling with the heatsink. As the device switches common mode noise is developed that eventually refers to the output, that's why an isolation transformer is used on the output of UPS to knock off common mode noise. Switching frequency is in the range of 800Hz to 2000-5000Hz, trading off efficiency as the switching frequency increases.

Ideally, the heatsink is floating above earth (ground) so other detection circuitry needs to be used to detect ground leakage/faults.

The output impedance of that inverter system is not much more than the output, say 10% tops. For a 3000VA inverter, is 4.3 ohms. Compared to the wall impedance depends on the network to the residence, but minimum would be several milliohms. For amplifiers, the high impedance will kill dynamics, so a good use for battery inverters are sources where the demand power is less than 10-30W. It's possible that people who say their amplifier sounds better on battery inverter could have high freq noise creeping into signal spaces, similar to untreated USB connections. Using the Purifier can mitigate that noise, but best to avoid in the first place.

I'm not in favour of a complex array of L & C filters as found in many line conditioners, chance of ringing and reducing source impedance.. Simple stuff, like shunt caps works out as a best compromise to filter out common mode noise, especially from SMPS wall warts which should not be connected to an audio system at all.

 

[Kingrex]

@Onepoint5

Whats you understanding of output impedance of low frequency vs high frequency inverters.
I can't find a spec on Victron or others.

So, your saying, with an amp, you would need to seriously oversize the system. Like a 12,000 to 15 000 watt inverter for a pair of 250 watt amps. This also increases the battery bank size. There is a ratio to how much output the inverter has to how much current you need the battery to flow. Wire gets larger too. All of which increasing in size will lower the impedance.

 

[Onepoint5]

@Onepoint5

Whats you understanding of output impedance of low frequency vs high frequency inverters.
I can't find a spec on Victron or others.

So, your saying, with an amp, you would need to seriously oversize the system. Like a 12,000 to 15 000 watt inverter for a pair of 250 watt amps. This also increases the battery bank size. There is a ratio to how much output the inverter has to how much current you need the battery to flow. Wire gets larger too. All of which increasing in size will lower the impedance.

The output impedance of inverters is the same regardless of switching frequency. Ultimately, the semiconductors are the fuse or the limiting factor, and hopefully overcurrent protection circuits are quick enough (they usually are).

For amps, if I plug 1 x 100W/ch and 1 x 250W channel amp (or solo) into an 1.5kVA Equitech, or a Topaz, the dynamics are lost. A lot better is plug into the wall for amps, and shunt filter noise. I'm away from my reference material, but have in mind the current draw from amps is Current (Watts/Volts) / 0.6 for the min size cable and breaker (you may have more mileage on this). So that factor on a 'high' impedance load is going to heat up something and not do any work.
The Equitech buzzes like crazy (even with the amps off circuit), so any largish demands will make this worse.

 

[Kingrex]

I am not moving forward with Exeltech. Too small of inverters in the XP line. Rack mounts get to any size but very very expensive to build a larger models. 10s of thousands.

I'm pursuing a certification with Victron and hopefully a reseller license. Victron is a hybrid inverter that uses a Fet bridge that operates at 20khz. With a 48 volt battery the sine wave comes out at about 42 volts, 60 hertz. Filters smooth the wave and a large toroid stepup brings it to 120 volts with 2% thd under load. I would use 5000 watt inverters. You can parallel up to 6 for higher wattage. 6 @ 120 volts is 250 amps.

lets say I start with 2. Thats 84 amps on the distribution panel bus. With 20 amp distribution breakers and 10 awg branch wire, I don't think I will have impedance issues. But I don't know the math of inverters. I'm just not hearing anyone say they have issues driving amps with an appropriately sized Inverter. One person pushes Parasound 250 watt monos.

I have heard of people frying CH Precision amps using a popular 1 box battery solution. The 2 x 20A power cords per amp CH model. They are plugging into a $32,000 one box solution that has a 750 volt amp inverter. Let say thats a 1000 watt inverter. I'm guessing, but lets say that CH amp has a standing load of 800 watts no music. It has 2 x 20 amp cords. At 80% thats 1600 watts per cord. I'm saying its drawing 400 per cord. Thats 800 watts static draw on a 1000 watt inverter. Now you play music. I can see how your easily maxing out the inverter and severly clipping the tops of the sine wave. Your creating massive heat in the amplifiers as its tries to draw even more current since the voltage is failing due to the sine wave not reaching its crests as it clips. And then the Genie comes out.

I keep hearing people come forward with properly set up battery inverter systems and they say they beat the wall. I need to investigate further. I will compare it to well designed wall power using a load center and 10 awg branch wire. I will also compare it to a all copper NQ panel. I think there is more to it than impedance. Utility power has many pollutant. Your house being one. Battery power has non of that.

I share a transformer with one business across the street that is a one man shop making slab tables. I have pretty clean power. I have a good place to test the system

FWIW Topaz sucks. Throw it out. I don't know how Equitech works, but I have never heard a Torus RM20 buzz. And I see people all the time under power their isolation transformer, then cry foul. No isolation transformer fixes small gauge wire in the wall. You have to feed them with 10 awg or larger. They fix common mode noise and RF over 50khz.

My 4500 watt Torus is feed with 240 volts of #8 wire. Thats 80 amps of wire for a 37 amp core. Always oversize wire feeding a isolation transformer. My Torus WM45 transformer never compressed any amp I put on it. I have put Audio Research Ref 160 monos on the 2000 watt Torus RM20 and it was all around better. I have a krell KSA250 on the RM20 and its no contest better than the wall. And the wall is my 10 awg grain oriented wire feed by my all copper panel. Same panel and wire at a diferent house and the RM20 beats the wall with a CH precision amp. The single 20A power cord version. The preamp and all digital gear is through it too. Another job same as above, 2 x 250 watt SS monos and all front end. Better than the wall. Impedance isn't the end all determining factor in good power.

IMO it's worth it to try as lots of people can't get wire to their room. The biggest issue is size constraints. A good inverter is not small. Batteries are not small. You need a distribution panel for higher wattage units. And making it drop in ready is probably the most difficult challenge to overcome.

 

[Kingrex]

One cannot take anything Amir at ASR posts as serious. Science left his forum long ago. His tests are limited in product scope and test detail.

Yes. I was mulling my too hot reaponse and getting kicked off his site reciently as I poked at his review of a Puritan 156. All he did was measure a preamp output from the wall and from the Puritan, saw no difference and claimed he had done a through analysis. I did say, why not use a BACCH for in room reaponse measurements. BACCH is high enough reaolution it would tell you if something changed. I would also think he should complete and audio chain. Tie a amp to the preamp and test the output of the amp when the pre is plugged into the wall or the Puritan. He should also put a load on the preamp and send a sonic blat through the preamp and use a specrograph to see if it changes. Problem may be he may not be educated enough to use such tools.
I did say he should perform a blind test. His minions scream bloody hell that no one performs blind test so its expectation bias. So why not perform a blind test. He won't because the results may come back conclusive and that would 100% validate his scientific method is flawed. And if he could not then figure a way to explain it, he would have to admit he wasn't intellignent enough to know how to use Science to measure everything.
FWIW I believe we can measure everything. We many times lack the knowledge of how to set up the experiment. But the tools are there.

There are a handful of measurements that would be useful. Amir is just a Hack. Hopefully his trolls will report this and Amir can defend himself. I wager he won't as it immediately illustrates the flaws in his techniques.

 

[Republicoftexas69]

Yes. I was mulling my too hot reaponse and getting kicked off his site reciently as I poked at his review of a Puritan 156. All he did was measure a preamp output from the wall and from the Puritan, saw no difference and claimed he had done a through analysis. I did say, why not use a BACCH for in room reaponse measurements. BACCH is high enough reaolution it would tell you if something changed. I would also think he should complete and audio chain. Tie a amp to the preamp and test the output of the amp when the pre is plugged into the wall or the Puritan. He should also put a load on the preamp and send a sonic blat through the preamp and use a specrograph to see if it changes. Problem may be he may not be educated enough to use such tools.
I did say he should perform a blind test. His minions scream bloody hell that no one performs blind test so its expectation bias. So why not perform a blind test. He won't because the results may come back conclusive and that would 100% validate his scientific method is flawed. And if he could not then figure a way to explain it, he would have to admit he wasn't intellignent enough to know how to use Science to measure everything.
FWIW I believe we can measure everything. We many times lack the knowledge of how to set up the experiment. But the tools are there.

There are a handful of measurements that would be useful. Amir is just a Hack. Hopefully his trolls will report this and Amir can defend himself. I wager he won't as it immediately illustrates the flaws in his techniques.

Rex. I am willing to wager that Amir will not present himself to us. He tends to gravitate to that other forum when his precious ASR and his testing are questioned.

All the best
Jeff.

 

[Kingrex]

I know he won't. I like technical expertise. That is why I pay attention when Ralph and some others comment. And why I ask questions at times. But it does piss me off when Amir presents a very basic, incomplete analysis as a definitive conclusion. He has had interesting things to say about switches and such and how they work. But when he falls on his face like he did with the Puritan analysis, it calls into question everything he did. Too bad. He could be a usefull resource. But he has shown himself to be limited in his knowledge and scope of electrical fundamentals and the tools used to analize equipment.