Safety And Design of DIY originated Audio Products

[Mivera]

It is not enough to teach a man a specialty. Through it he may become a kind of useful machine but not a harmoniously developed personality. It is essential that the student acquire an understanding of and a lively feeling for values. He must acquire a vivid sense of the beautiful and of the morally good. Otherwise he – with his specialized knowledge – more closely resembles a well-trained dog than a harmoniously developed person. He must learn to understand the motives of human beings, their illusions and their sufferings, in order to acquire a proper relationship to individual fellow men and to the community.

These precious things are conveyed to the younger generation through personal contact with those who teach, not – or at least not in the main – through textbooks. It is this that primarily constitutes and preserves culture. This is what I have in mind when I recommend the ‘humanities’ as important, not just dry specialized knowledge in the fields of history and philosophy.

Overemphasis on the competitive system and premature specialization on the ground of immediate usefulness kill the spirit on which all cultural life depends, specialized knowledge included.

It is also vital to a valuable education that independent critical thinking be developed in the young human being, a development that is greatly jeopardized by overburdening him with too much and with too varied subjects (point system). Overburdening necessarily leads to superficiality. Teaching should be such that what is offered is perceived as a valuable gift and not as a hard duty.

—Albert Einstein, "Education for Independent Thought"
New York Times, Oct. 5, 1952


Unfortunately the scientific method goes against many of these ideals. But what did Albert Einstein know?

 

[Cosmik]

The benefits of choke line filters for idiots:

"SCHURTER’s popular power entry module, KFA series, is now equipped with an additional ground line choke to suppress high frequency interference arising from ground loops on the ground line. In addition, the newly added ground line choke improves asymmetric attenuation in higher frequencies in medical filter versions. The KFA series with existing common-mode choke on the phase and neutral conductors, together with an additional ground line choke, is well enhanced to address the growing EMC challenges facing compliance engineers today.

Ground loops arise through different impedances of different ground line connections in one or multiple pieces of equipment. While all ground connections have the same potential level for the 50 Hz load current, these can have different potential levels for high-frequency interference. The ground line choke, which is wired between the power-inlet connector and the filter output, smartly addresses this phenomenon. Because of its placement directly at the power line input, high-frequency interference currents on the ground line are prevented from bypassing the filter through the equipment housing. The ground line choke isolates the equipment chassis against high frequencies and the equipment ground from the ground on the line power connection. This approach effectively breaks any possible ground loops.

Audio equipment for instance is known for being prone to ground loop problems. Often referred to as hum or buzz, the KFA, with ground line choke, is well suited for suppressing interferences that lead to diminished sound quality. The KFA is also well suited for medical equipment required to use low leakage filters for increased patient safety. The reduced Y capacitance to ground, or absence thereof, leads to reduced asymmetric attenuation in higher frequencies. The KFA with ground line choke helps to compensate for this reduced performance."



https://www.schurter.com/en/Newsroo...Module-with-filter-now-with-Ground-Line-Choke

But yeah I know you guys prefer this approach.

http://www.entreq.com/products/ground-boxes-17667704

Disregarding the abuse, that is really very interesting and if I understand it, appears to be a magic bullet that would eliminate the need for this probably-illegal device in amateur non-double-insulated equipment:

But it appears to contradict what Rod Elliott is saying by suggesting that a hum loop is a high frequency phenomenon. He puts a bypass cap on his earth lifter which must be doing the opposite of what Schurter are recommending..?

The capacitor in parallel ensures that the electronics are connected to the chassis for radio frequency signals, and helps to prevent radio frequency interference.

 

[Mivera]

Disregarding the abuse, that is really very interesting and if I understand it, appears to be a magic bullet that would eliminate the need for this probably-illegal device in amateur non-double-insulated equipment:

But it appears to contradict what Rod Elliott is saying, suggesting that a hum loop is a high frequency phenomenon. He puts a bypass cap on his earth lifter which must be doing the opposite of what Schurter are recommending..?

I'm not sure. I'm more results oriented than belief oriented.

 

[Mivera]

"Give us a break"
-alfe
Audiosciencereview, Aug.26,2017

Yes I know reality can be hard on the brain

 

[Mivera]

As members of the scientific community, I really think it's very important to understand the difference between belief's (often fantasy) and reality.

https://en.wikipedia.org/wiki/Belief

Clear knowledge of the difference between belief and reality, will help some members better understand the difference between belief and reality. In the scientific community, this is very important.

 

[amirm]

The benefits of choke line filters for idiots:

"SCHURTER’s popular power entry module, KFA series, is now equipped with an additional ground line choke to suppress high frequency interference arising from ground loops on the ground line. In addition, the newly added ground line choke improves asymmetric attenuation in higher frequencies in medical filter versions. The KFA series with existing common-mode choke on the phase and neutral conductors, together with an additional ground line choke, is well enhanced to address the growing EMC challenges facing compliance engineers today.

Ground loops arise through different impedances of different ground line connections in one or multiple pieces of equipment. While all ground connections have the same potential level for the 50 Hz load current, these can have different potential levels for high-frequency interference. The ground line choke, which is wired between the power-inlet connector and the filter output, smartly addresses this phenomenon. Because of its placement directly at the power line input, high-frequency interference currents on the ground line are prevented from bypassing the filter through the equipment housing. The ground line choke isolates the equipment chassis against high frequencies and the equipment ground from the ground on the line power connection. This approach effectively breaks any possible ground loops.

Audio equipment for instance is known for being prone to ground loop problems. Often referred to as hum or buzz, the KFA, with ground line choke, is well suited for suppressing interferences that lead to diminished sound quality. The KFA is also well suited for medical equipment required to use low leakage filters for increased patient safety. The reduced Y capacitance to ground, or absence thereof, leads to reduced asymmetric attenuation in higher frequencies. The KFA with ground line choke helps to compensate for this reduced performance."

https://www.schurter.com/en/Newsroo...Module-with-filter-now-with-Ground-Line-Choke

The only idiot is one who doesn't understand the most basics of electronics, ohm's law and misreads what he wants into what they have written. And of course, whoever wrote the above intended for inexperienced people to fall in the trap.

Let's see how they write this when they provide a write up to independent site: http://www.engineerlive.com/content/why-it-makes-sense-add-ground-choke

Herbert Blum [of Schurter] presents a simple and safe solution for ground connection interferences

Switched mode circuits are increasingly being used in electronic devices. Due to their rapid switching speeds, the probability of high-frequency ground connection interferences is also becoming more prevalent. A correctly specified and placed ground choke offers a simple and safe solution.​

You see what I highlighted and the application? Switchmode power supplies that run at high frequencies which can ride on top of the ground wires which act like antennas at high frequencies. Not mains frequencies. High frequencies where the wavelength of the noise matches 1/4 wavelength of the wire. For 60 Hz that wave length is 4 million feet! For 2.5 Ghz, that is just 1 inch! So what works for high frequencies does not at all related to 60 Hz mains.

They go on to say:

The topic of overlooked interferences concerns that of high-frequency interferences on the ground as well. The ground receives little attention (compared to interferences generated on the line) during the EMC-compliant design of electronic devices. Yet this can quickly change if a ground loop develops.​

So this is an EMC compliance which is not measured at 60 Hz. We are talking Megahertz and Gigahertz. But importantly, this type of filtering already exists in your switchmode power supply if it comes with proper certification. So there is no problem to solve, unless you screw up the grounding by using long lengths of wire (which you had done by not placing the supply near the IEC connector).

And the punchline:

When ground loops occur

Ground loops develop due to the different impedances of various ground connections from one or more devices. Although all ground connections have the same potential for the 50Hz load current, they can have different potentials due to the different line impedances for the high-frequency interferences.​

So right there it tells you this has nothing to do with 50/60 Hz. In the marketing write-up you posted, they are using audio hum as an analogy to make people understand same kind of ground loops can exist at high frequencies which cannot be experienced.

Back to ohm's law, what does a choke do anyway? Let's go to Wiki: https://en.wikipedia.org/wiki/Choke_(electronics)

In electronics, a choke is an inductor used to block higher-frequency alternating current (AC) in an electrical circuit, while passing lower-frequency or direct current (DC). A choke usually consists of a coil of insulated wire often wound on a magnetic core, although some consist of a doughnut-shaped "bead" of ferrite material strung on a wire. The choke's impedance increases with frequency. Its low electrical resistance passes both AC and DC with little power loss, but it can limit the amount of AC due to its reactance.​

It is very clear then that the choke is designed to limit high frequencies, not mains. Think about it. The choke is also used inline with the hot and neutral. If it blocked mains AC, how would that ever pass power???

An inductor is a device that changes its resistance with frequency. For an inductor, its impedance or resistance it provides to current flow is X = 2 * pi() * Frequence * L

L is inductance of the inductor or filter choke. We see that our total impedance rises with frequency. Schurter specs the ground choke inductance at .6 and .15 millihenries (https://www.schurter.ch/bundles/snceschurter/epim/_ProdPool_/newDS/en/typ_KFA.pdf). Plugging that into above formula we get:

60 Hz: X = 0.2 ohms!!!

That is a dead short. Nothing will dissipate inside that choke at 60 Hz.

Now at 60 Mhz, we get X = 220,000 ohms. Now current going through such a large "resistor" will drop a lot of voltage due V = I * R. And with it, attenuate those high frequencies from getting through.

Yes, you can make a much larger inductor to block 60 Hz but it would be massive in size due to such low frequency. It would be orders of magnitude larger than any piece of audio equipment in size and weight.

Summary:
A ground line choke is a simple inductor meant for filtering very high frequencies. The only ground currents/loops it eliminates is at those frequencies (many megahertz). It acts like a dead-short/piece of wire at mains frequencies and harmonics. And those are the frequencies we care about because they are in audible range.

And even the high frequency filtering that is there is done for compliance with emissions standards. In the case of Mike's DAC, all the required filtering is already inside the switchmode power supply. Yes, he may have created new sources of high frequency interference by using long wires and unneeded modules like that soft-start. So maybe if he applied for certification he would have to add additional filtering but good design would have avoided all of that. And at any rate, none of this has to do with mains leakage/hum in audio.

 

[amirm]

Disregarding the abuse, that is really very interesting and if I understand it, appears to be a magic bullet that would eliminate the need for this probably-illegal device in amateur non-double-insulated equipment:

No, there is nothing here but an extra inductor:

All they have added is L3. No magic there. That provides a high impedance path for high frequencies. It does nothing for mains. If it blocked mains, it would also stop the ground wire from operating in the case of a short! It would have to absorb all the power itself which can nearly 2000 watts in US. No way that little box could do that, or would want to. This is just a little can on the back of the IEC mains socket:

 

[Mivera]

The only idiot is one who doesn't understand the most basics of electronics, ohm's law and misreads what he wants into what they have written. And of course, whoever wrote the above intended for inexperienced people to fall in the trap.

Let's see how they write this when they provide a write up to independent site: http://www.engineerlive.com/content/why-it-makes-sense-add-ground-choke

Herbert Blum [of Schurter] presents a simple and safe solution for ground connection interferences

Switched mode circuits are increasingly being used in electronic devices. Due to their rapid switching speeds, the probability of high-frequency ground connection interferences is also becoming more prevalent. A correctly specified and placed ground choke offers a simple and safe solution.​

You see what I highlighted and the application? Switchmode power supplies that run at high frequencies which can ride on top of the ground wires which act like antennas at high frequencies. Not mains frequencies. High frequencies where the wavelength of the noise matches 1/4 wavelength of the wire. For 60 Hz that wave length is 4 million feet! For 2.5 Ghz, that is just 1 inch! So what works for high frequencies does not at all related to 60 Hz mains.

They go on to say:

The topic of overlooked interferences concerns that of high-frequency interferences on the ground as well. The ground receives little attention (compared to interferences generated on the line) during the EMC-compliant design of electronic devices. Yet this can quickly change if a ground loop develops.​

So this is an EMC compliance which is not measured at 60 Hz. We are talking Megahertz and Gigahertz. But importantly, this type of filtering already exists in your switchmode power supply if it comes with proper certification. So there is no problem to solve, unless you screw up the grounding by using long lengths of wire (which you had done by not placing the supply near the IEC connector).

And the punchline:

When ground loops occur

Ground loops develop due to the different impedances of various ground connections from one or more devices. Although all ground connections have the same potential for the 50Hz load current, they can have different potentials due to the different line impedances for the high-frequency interferences.​

So right there it tells you this has nothing to do with 50/60 Hz. In the marketing write-up you posted, they are using audio hum as an analogy to make people understand same kind of ground loops can exist at high frequencies which cannot be experienced.

Back to ohm's law, what does a choke do anyway? Let's go to Wiki: https://en.wikipedia.org/wiki/Choke_(electronics)

In electronics, a choke is an inductor used to block higher-frequency alternating current (AC) in an electrical circuit, while passing lower-frequency or direct current (DC). A choke usually consists of a coil of insulated wire often wound on a magnetic core, although some consist of a doughnut-shaped "bead" of ferrite material strung on a wire. The choke's impedance increases with frequency. Its low electrical resistance passes both AC and DC with little power loss, but it can limit the amount of AC due to its reactance.​

It is very clear then that the choke is designed to limit high frequencies, not mains. Think about it. The choke is also used inline with the hot and neutral. If it blocked mains AC, how would that ever pass power???

An inductor is a device that changes its resistance with frequency. For an inductor, its impedance or resistance it provides to current flow is X = 2 * pi() * Frequence * L

L is inductance of the inductor or filter choke. We see that our total impedance rises with frequency. Schurter specs the ground choke inductance at .6 and .15 millihenries (https://www.schurter.ch/bundles/snceschurter/epim/_ProdPool_/newDS/en/typ_KFA.pdf). Plugging that into above formula we get:

60 Hz: X = 0.2 ohms!!!

That is a dead short. Nothing will dissipate inside that choke at 60 Hz.

Now at 60 Mhz, we get X = 220,000 ohms. Now current going through such a large "resistor" will drop a lot of voltage due V = I * R. And with it, attenuate those high frequencies from getting through.

Yes, you can make a much larger inductor to block 60 Hz but it would be massive in size due to such low frequency. It would be orders of magnitude larger than any piece of audio equipment in size and weight.

Summary:
A ground line choke is a simple inductor meant for filtering very high frequencies. The only ground currents/loops it eliminates is at those frequencies (many megahertz). It acts like a dead-short/piece of wire at mains frequencies and harmonics. And those are the frequencies we care about because they are in audible range.

And even the high frequency filtering that is there is done for compliance with emissions standards. In the case of Mike's DAC, all the required filtering is already inside the switchmode power supply. Yes, he may have created new sources of high frequency interference by using long wires and unneeded modules like that soft-start. So maybe if he applied for certification he would have to add additional filtering but good design would have avoided all of that. And at any rate, none of this has to do with mains leakage/hum in audio.

Wrong again. We also care about the high frequencies as well. This is where textbook reading numbers mongers have a real hard time understanding things. Second order effects. But when you don't listen to things, or measure during the design phase, and you can't find it in a textbook, I can understand how it may be hard to understand.

 

[amirm]

I think I have rambled on enough, time for dinner. I leave with this question, "Does anyone want to know how it sounds?"

My warm welcome also and thanks for a good first post! And yes, let's hear about the sound.

 

[amirm]

Unfortunately the scientific method goes against many of these ideals. But what did Albert Einstein know?

He also said: “Two things are infinite: the universe and human stupidity; and I'm not sure about the universe.”

 

[Mivera]

We seen a very good example of second order effects at a clients place the other day. This guy happens to have a tube amp that's sensitive to ultrasonic frequencies. And when used together with an AVC based attenuator, DSD 64's ultrasonic noise that's far out of the range of human hearing caused noise that was audible across the entire audible range. This guy just couldn't understand how this ultrasonic noise could be heard since it's so far out of the audible range. Well just a great example of second order effects. Even noise above 5Mhz can cause second order effects in audio gear. But the folks who know this are experienced audio gear designers, not textbook copy and pasters.

 

[amirm]

Wrong again. We also care about the high frequencies as well.

Putting aside fairy tales reasons you may have for that worry, the only way you would know you have a high frequency problem is to measure it which you have not done.

What you have done is create problems in high frequencies by dragging long wires between mains and power supply input:

You see all those red lines? Those are nice antennas at high frequencies for the very reason I mentioned in my post.

And it is not me telling you this, the very manufacturer of the IEC with filter choke tells you: http://www.engineerlive.com/content/why-it-makes-sense-add-ground-choke

See? You stuck that filter in the middle of your box with wires connecting it, causing a different impedance for the case. Speaking of the case, we still have no idea how that is grounded.

This (high frequency issues) is a highly specialized domain that even most engineers don't understand (or care to understand) and rely on expert consultants (Henry Ott, etc.). To think that a hobbyist like you would have any remote understanding it would be the same as believing unicorns exist.

So no, you don't just slap filters in line without understanding the principles of physics and engineering. As I have noted, you most likely are doing harm than any good.

To clean up your act, you should turn around that switchmode power supply and connect it right at the terminals of the IEC mains. Don't put an EMC filter in there unless it is needed because as is, and I explained in my first post, it is adding AC mains leakage in order to filter high frequencies ("Y" cap problem). Get rid of soft-start module. Use a simple, mechanical switch with shortest wiring. Put a fuse in there just in case you screw up any of this when you wire it up.

And get a preliminary safety and emissions review done by a lab. It will cost a few thousand dollars but it will be educational for you. And would get you out of willfully exporting non-compliant equipment.

 

[Mivera]

Putting aside fairy tales reasons you may have for that worry, the only way you would know you have a high frequency problem is to measure it which you have not done.

What you have done is create problems in high frequencies by dragging long wires between mains and power supply input:

View attachment 8373

You see all those red lines? Those are nice antennas at high frequencies for the very reason I mentioned in my post.

And it is not me telling you this, the very manufacturer of the IEC with filter choke tells you: http://www.engineerlive.com/content/why-it-makes-sense-add-ground-choke

View attachment 8374

See? You stuck that filter in the middle of your box with wires connecting it, causing a different impedance for the case. Speaking of the case, we still have no idea how that is grounded.

This (high frequency issues) is a highly specialized domain that even most engineers don't understand (or care to understand) and rely on expert consultants (Henry Ott, etc.). To think that a hobbyist like you would have any remote understanding it would be the same as believing unicorns exist.

So no, you don't just slap filters in line without understanding the principles of physics and engineering. As I have noted, you most likely are doing harm than any good.

To clean up your act, you should turn around that switchmode power supply and connect it right at the terminals of the IEC mains. Don't put an EMC filter in there unless it is needed because as is, and I explained in my first post, it is adding AC mains leakage in order to filter high frequencies ("Y" cap problem). Get rid of soft-start module. Use a simple, mechanical switch with shortest wiring. Put a fuse in there just in case you screw up any of this when you wire it up.

And get a preliminary safety and emissions review done by a lab. It will cost a few thousand dollars but it will be educational for you. And would get you out of willfully exporting non-compliant equipment.

Another load of bs. For 1 all the mains wire is shielded. For 2, once the lid is on, that power supply is completely sealed and isolated from the rest of the gear.

And no it's not incorrectly installed. The ground is passing through the choke as intended.

 

[amirm]

We seen a very good example of second order effects at a clients place the other day. This guy happens to have a tube amp that's sensitive to ultrasonic frequencies.

And ultrasonic noise may very well be what you are spitting out of your DAC. See this picture again:

Your switchmode power supply is running at 30 Khz to 100 Khz most likely and harmonics going well into hundreds of Megahertz. Whatever noise doesn't get filtered in its mains filtering, will be running on those long AC wires you have.

You put those white signal wires close to those AC wires allowing capacitive coupling to occur and as such, allowing high frequency noise to bleed into your DAC output!

THere is a reason any properly designed DAC would put the signal output way the heck away from mains input. But you have not.

Now, you could show us measurements that demonstrate high frequency noise does not exist there. But you are fresh out of that.

What you do have is listening tests which thankfully due our ears not hearing above 20 Khz, means we are not hearing problems that you may have very well created.

 

[Mivera]

And ultrasonic noise may very well be what you are spitting out of your DAC. See this picture again:

View attachment 8375

Your switchmode power supply is running at 30 Khz to 100 Khz most likely and harmonics going well into hundreds of Megahertz. Whatever noise doesn't get filtered in its mains filtering, will be running on those long AC wires you have.

You put those white signal wires close to those AC wires allowing capacitive coupling to occur and as such, allowing high frequency noise to bleed into your DAC output!

THere is a reason any properly designed DAC would put the signal output way the heck away from mains input. But you have not.

Now, you could show us measurements that demonstrate high frequency noise does not exist there. But you are fresh out of that.

What you do have is listening tests which thankfully due our ears not hearing above 20 Khz, means we are not hearing problems that you may have very well created.

Another fantasy fabrication.

 

[amirm]

And no it's not incorrectly installed. The ground is passing through the choke as intended.

Once again here is the picture from the manufacturer:

Isn't your EMC filter external to IEC mains just like top scenario?

Do you even understand what they are showing above with high frequency ground loops inside the equipment?

 

[Mivera]

Once again here is the picture from the manufacturer:

Isn't your EMC filter external to IEC mains just like top scenario?

Do you even understand what they are showing above with high frequency ground loops inside the equipment?

Yes it's connected exactly like the correct installation. However I'm not using the same unit. The idea is to ensure the ground passes through the filter and doesn't bypass the filter. This is why I put the chassis ground after the filter.

 

[amirm]

For 1 all the mains wire is shielded.

What is the shield connected to?

 

[Mivera]

And these long mains wires you keep talking about are only just over 6" long. And they are shielded. Far less than the unshielded mains cables running to this big noisy transformer in the DAC you used as an example of engineering excellence:

And that big noisy transformer isn't even shielded from the rest of the gear! No wonder he went to $10 Chinese SMPS's in his latest models.

 

[Mivera]

What is the shield connected to?

Nothing.