Beyond Bit-perfect: EMC - EMI/RFI and the hidden gray areas of analogue in our digital audio.

[kumakuma]

Your contributions are totally relevant.

 

I do understand firedog's remark about how your posts could be perceived by some others, and in fact initially I had the same kind of perceptions too but after a while found some plausible explanations as to what could be the cause (these very same gray areas!), but he has a lot of great things to contribute too, so please, both of you, do stick around and continue to discuss as friends sharing the same hobby.

 

Have a great week-end!

 

YashN,

 

How did your testing of Alex's theories go? Were you or your girlfriend able to hear differences between different rips having identical checksums?

 

Tom

[YashN]

Has anybody tried a Linear PSU with a Broadband Cable Modem to see if it helps with RF/EMI?

I have been tempted to try one with a changeover to a small battery for when there is a blackout, as the supplied Modem doesn't have battery backup and the Telephone is then off as well.

 

Good idea. I know I tried a few rips and up-conversions by turning off the cable modem and router.

 

I still want to enable a few more filter lines and then hook the Cable Modem and the Wi-Fi router into it soon.

 

Another thing I thought of is to re-commission a little DIY circuit I built to have oscilloscope readings on the computer so as to troubleshoot my Korg Polysix originally.

 

I am planning to hook this up to the USB cable or perhaps the DAC output or something like that and see what I can detect there first, and then do a step-by-step disconnection of each of the brick power supplies and look for any significant noise profile change. I am assuming here that most of if not all of my wall-wart power supplies must be switching ones.

 

Here are a few I have:

- Cable Modem

- Wi-Fi Router (Linksys)

- Soundcraft Mixer

- M-Audio Fast-TracK Pro

- Two external HDD (one of which holds my main Audio partition, so I may not be able to turn this one off permanently, but I can still do it as a test).

 

Of course, sometimes some synths are plugged in as well, and the odd one here is the Alesis Micron, which has an AC-AC brick (yep!, not AC-DC!).

 

In addition the A-V gear is connected too: TV, Blu-Ray, Apple TV, Centre channel (active studio monitor) and Subwoofer (active).

Fun?

[YashN]

Untangled the little wire mess of my DIY Oscilloscope probe circuit that I used to troubleshoot the Polysix to check if it still works.

 

Nothing fancy here, and all twisted pairs (not coaxial), so that thing can also show its own interference from around. That's all I have for now.

 

I am using Soundcard Scope for Win under WINE on Mac OS X.

 

Hooked it up to my DIY USB cable on the VBus, then GND, then one Signal line.

 

Found everything quite noisy!

 

It looks like there are peaks at 8.KHz, then 16KHz. I thought I saw something at around 500Hz as well, and 60Hz (mains) and below 10Hz is quite a mess as well, at least for VBus.

 

Has to be taken with a grain of salt because of the guerilla-style equipment, but I think I did see some people mentioning 8KHz as inherent to USB.

 

Scope doesn't give me above audio-band readings, so that's quite limited for what I would like to do. I had thought that the upper limit would be half the max sample capture rate of the internal Mac OS X input, i.e. (half of 96kHz to go up to 48kHz), but I can't change that in the soft.

 

A few screenshots:

 

 

 

[wgscott]

Yes, I do.

This is a lack of education, not some advanced technique that only sandyk has managed to figure out.

What he suggests can't happen.

 

Tolerating posts such as that, is the reason that people point and laugh at "audiophiles".

This is the equivalent of comparing homeopathy to medicine and saying that everyone's opinion on the matter is equally valid, whether they are a layman or a doctor/researcher with decades of experience.

 

Because I care about audio quality, not mysticism.

 

Because you are not playing back audio from the drive in real-time.

You do multi-pass secure rips to verify that the data is extracted correctly, and can verify that against an external database.

If the rip is good, it doesn't matter what created it. The resulting audio does not contain anything related to the drive which created it.

 

The burden of proof lies with those people making extraordinary claims.

Repeating something often enough and loud enough is not creating proof.

 

If you want to claim that noise from an idle optical drive in the system is somehow being transmitted over the USB ground during playback and that is somehow affecting how things sound I'd think it's a bit far-fetched, but not entirely implausible.

But you're claiming that this is somehow stored in the file itself in a manner which cannot be detected or verified by computers.

 

+ 6.02 x 10²³

[Jud]

+ 6.02 x 10²³

 

I think we have a mole!

[sandyk]

I think we have a mole!

 

A bottomole ?

[YashN]

A bottomole ?

 

Hehe!

[YashN]

A chart showing the ambient frequencies we're subjected to.

 

 

 

Since every cable is also an antenna (can receive or radiate), this gives a small idea of what we could up against relative to EMC. In particular, how does vintage gear behave in today's environment?

 

Also, cheap ferrite beads can filter out RFI. These beads can be bought from electronics stores or cut from the weighted cylinders found on many cheap USB cables and then wrapped around the wires *running from an audio source to the amplifier or, in rare cases, around the speaker wires themselves.

 

 

Electromagnetic fields drop off quickly, so keeping a cellphone just 4 or 5 feet from vulnerable speakers or amplifiers will usually dull interference sounds beyond perception. A Wi-Fi-attacking microwave can be thwarted by keeping routers and wireless devices at least 10 to 15 feet away.

[YashN]

Tested another 'Scope' software yesterday through WINE. This one does allow for setting a higher capture mode, so I do get up to 48kHz as bandwidth.

 

Apart from the previously mentioned test where each potentially troublesome power supply is removed in turn, I could also use clamp-on ferrite beads on the USB cable and monitor the effects from there.

 

I am also planning to implement filtering in my USB cable.

[YashN]

Did you try turning off fridge/CFL-LED lighting/anything at all to do with Internet, wifi or computers/unplugging all SMPSs throughout the house and seeing whether any of those made any difference?

 

Edit: Oh yeah, and toasters/coffeemakers/any other small appliances too, as they're very likely these days to have microprocessors built in. (These were mentioned as possible sources of harmonic distortion and phase problems in a paper done for the power industry I read during my solar/battery power research.)

 

Two lists from 'A smarter approach to resolving power line noise':

 

Common Household Items that Cause Interference

 

• Door bell transformers
  
• Electric blankets
  
• Heating pads (of all kinds)
  
• Recessed ceiling light fixtures
  
• Furnace control circuits
  
• Refrigerators (becoming a frequent problem)
  
• TV top and stereo, amplified antennas
  
• Light dimmers
  
• Aquarium heaters
  
• Screw-in photocells
  
• Low-energy compact (screw-in) fluorescent lights
  
• Touch control lamps
  
• Clean air machines (table top and furnace type)
  

 

 

Common Power-Line Noise Sources

 

(Listed in order from most common to least common)

 

• Loose staples on ground conductor
  
• Loose pole top pin
  
• Ground conductor touching nearby hardware
  
• Corroded slack span insulators
  
• Guy touching neutral
  
• Loose hardware
  
• Bare tie wire used with insulated conductor
  
• Insulated tie wire on bare conductor
  
• Loose crossarm braces
  
• Lightning arrestors
  

[Jarrett]

Is there anything that doesn't have an impact on sound quality?

[Fitzcaraldo215]

Is there anything that doesn't have an impact on sound quality?

 

There are several possible answers to this.

 

1. Yes, some things have no effect on SQ.

 

Other things have one of two possible effects on SQ:

 

- 2. clear, demonstrable effects via unbiased testing or measurement.

 

- 3. hypothetical effects based on anecdotal listening perceptions that may upon further review and experimentation be reclassified into 1.or 2., above, or which may never reach beyond the anecdotal speculation. Some of these might be:

 

- 4. True, but unprovable.

 

- 5. Untrue, in spite of continued speculation.

[Tony Lauck]

After doing a little research into battery and solar power and thereby learning just enough to be dangerous, I wonder whether the junk on the power line isn't due at least as much, if not more, to what's inside the home as what's outside. Did you try turning off fridge/CFL-LED lighting/anything at all to do with Internet, wifi or computers/unplugging all SMPSs throughout the house and seeing whether any of those made any difference?

 

Edit: Oh yeah, and toasters/coffeemakers/any other small appliances too, as they're very likely these days to have microprocessors built in. (These were mentioned as possible sources of harmonic distortion and phase problems in a paper done for the power industry I read during my solar/battery power research.)

 

The power down process does not have to stop with the circuit breakers in one's home. There are the neighbors' main breakers as well. Just saying, ....

[Tony Lauck]

Homeopathy: really not a good analogy. People can and do die from using homeopathy instead of proven therapies. No one dies b/c of their audio beliefs.

 

Indeed, not a good analogy. People can and do die from using "proven" therapies.

[jabbr]

Yes, I do.

This is a lack of education, not some advanced technique that only sandyk has managed to figure out.

What he suggests can't happen.

 

Tolerating posts such as that, is the reason that people point and laugh at "audiophiles".

This is the equivalent of comparing homeopathy to medicine and saying that everyone's opinion on the matter is equally valid, whether they are a layman or a doctor/researcher with decades of experience.

 

Because I care about audio quality, not mysticism.

 

Because you are not playing back audio from the drive in real-time.

You do multi-pass secure rips to verify that the data is extracted correctly, and can verify that against an external database.

If the rip is good, it doesn't matter what created it. The resulting audio does not contain anything related to the drive which created it.

 

The burden of proof lies with those people making extraordinary claims.

Repeating something often enough and loud enough is not creating proof.

 

If you want to claim that noise from an idle optical drive in the system is somehow being transmitted over the USB ground during playback and that is somehow affecting how things sound I'd think it's a bit far-fetched, but not entirely implausible.

But you're claiming that this is somehow stored in the file itself in a manner which cannot be detected or verified by computers.

 

The answer is rather simple, and entirely consistent with the most modern physics: there are branes where this happens.

[elcorso]

Indeed, not a good analogy. People can and do die from using "proven" therapies.

 

+1

 

More or less 75:1

 

Roch

[elcorso]

-----

Because I care about audio quality, not mysticism.

-----

 

Then you care only about "listening to music science", if you really love play and listen to music you need some "mysticism".

 

Life is an art essay and not a proven equation...!

 

Roch

[YashN]

Is there anything that doesn't have an impact on sound quality?

 

Perhaps a more pragmatic approach is 'where do I stop?'.

 

I think every time we work on the weakest link in the chain, we hear other things in the next weakest in the line.

 

In turn, solving one issue somewhere may make other things previously masked appear more readily as well.

 

The increasing high frequency range of our electronic equipment is a worry as what could be good for today may not be as good a few years down the line.

 

The major issue with digital is that some are content with the abstraction and leave it at that, but in reality, we are dealing with complex issues in the analogue domain.

 

A lot of this is an electrical problem.

 

Sometimes, I am tempted to leave the system in the state where it gives great attack transients and soundstage as well as presence and realness and enjoy it like that.

[Tony Lauck]

+1

 

More or less 75:1

 

Roch

 

Unfortunately, my wife was one of them.

[Superdad]

After doing a little research into battery and solar power and thereby learning just enough to be dangerous, I wonder whether the junk on the power line isn't due at least as much, if not more, to what's inside the home as what's outside. Did you try turning off fridge/CFL-LED lighting/anything at all to do with Internet, wifi or computers/unplugging all SMPSs throughout the house and seeing whether any of those made any difference?

 

Edit: Oh yeah, and toasters/coffeemakers/any other small appliances too, as they're very likely these days to have microprocessors built in. (These were mentioned as possible sources of harmonic distortion and phase problems in a paper done for the power industry I read during my solar/battery power research.)

 

Jud: John does turn off a lot of stuff in his lab when looking at the AC line. One of the key things he mentioned to me in talking about the degradation of supplies house mains (and he has a nice grid-tied solar system BTW) is that the utility companies (and government in some cases) are using the grid for communication and monitoring of all sorts of things. Those smart meters us high frequency carriers to send usage data all day long. I seem to recall John saying he was seeing all sorts of stuff on his mains at 19KHz.

 

He really wants to do some serious measuring of power supplies and what they kick back into the wall but the mains itself is far noisier than any of the SMPS units (at the AC side) which he has attempted to compare! I bought him one of those CyberPower power factor corrected "sine-wave" UPS units, but it's not clean enough. Next step may be for use to invest in a real AC regenerator (basically a 60hz power amp). The cheapest proper unit I could find is this 100W one from Monarchy Audio . He really does not need a high wattage unit like the PS Audio Power Plants, though maybe Mr. McGowan could hook us up with an older unit for a similar price.

[mayhem13]

Then you care only about "listening to music science", if you really love play and listen to music you need some "mysticism".

 

Life is an art essay and not a proven equation...!

 

Roch

 

Nicely put! One has to ask themselves when obsession overrides passion.

[YashN]

Jud: John does turn off a lot of stuff in his lab when looking at the AC line. One of the key things he mentioned to me in talking about the degradation of supplies house mains

 

I was going to ask him about that since yesterday this came to my thoughts again - after researching power supplies, I came across his posts on the SB PSU enhancements he brought again and this reminded me he mentioned the noise he found on the mains early in this thread.

 

(and he has a nice grid-tied solar system BTW)

 

I am planning some radical off-grid solar and non-solar powering equipment whose research has hooked me even more than audiophilia (I am beyond salvation now!). The reason is the huge bills we incur come winter-time, even after sealing the front door leak issues. Either Hydro QC is the culprit or the owner is stealing. Don't know yet how I can prove either.

 

is that the utility companies (and government in some cases) are using the grid for communication and monitoring of all sorts of things. Those smart meters us high frequency carriers to send usage data all day long. I seem to recall John saying he was seeing all sorts of stuff on his mains at 19KHz.

 

I have large sliding mirrors in the dining room. People ask if I have a 'Walk-in' closet there, but no, the unit I live in actually houses all the meters for all other units in the building too. One day (probably 4 years ago or so), they came in and installed smart meters.

 

This means that behind these mirrors are now 7 smart meters for all units of the building running simultaneously. This is not counting the huge Power Box near the listening area on which "600 amps" is written.

 

This probably gives you an idea now of why thinking about EMI/RFI at my place is preponderant. I was even thinking of installing some metal meshes behind the mirrors since some people talk about health hazards of a single smart meter. Now, imagine 7 of these working simultaneously!

 

He really wants to do some serious measuring of power supplies and what they kick back into the wall but the mains itself is far noisier than any of the SMPS units (at the AC side) which he has attempted to compare! I bought him one of those CyberPower power factor corrected "sine-wave" UPS units, but it's not clean enough. Next step may be for use to invest in a real AC regenerator (basically a 60hz power amp). The cheapest proper unit I could find is this 100W one from Monarchy Audio .

 

When I post some pics of what we have to deal with, he'll probably consider himself lucky! God knows what's lurking in my audio system and I'm not properly equipped to diagnose this.

 

It will be interesting to see what he finds out concerning the mains. I remember that for the SB PSU, he was essentially solving for 40kHz to 100MHz kick-back and resonance. There must be much more to deal with in situations of differing equipment as well as the new AC mains pollution.

[YashN]

Here's an extract of John's post, post #2 in this thread, concerning what he found on the AC mains recently. More details would be great.

 

On AC power lines, things have gotten MUCH worse in recent years. 5 years ago I did a fairly detailed analysis of what was on my AC line in my house. A few weeks ago I re-did the analysis and was horrified at what I found. There is so much more stuff on our AC line today. I have a feeling that approaches like "line filters" etc are just going to be swamped by this much "noise". I'm really thinking that brute force approaches like motor generator sets are getting to be appropriate. Or how about two tanks of water, one down low and one up high, the water flows from the upper one through a turbine which drives a generator. The external power runs a pump which pumps water from the lower tank to the upper tank. That should block pretty much anything. Or you can do it with air pressure in tanks.

 

I have been thinking batteries and solar and perhaps some non-solar recharging circuits which are quite unconventional if not controversial, but I know how they work, they make sense to me.

[JohnSwenson]

I'm in the process of redoing all the lighting in my house. I used to have a lot of CFLs, but they have really crappy SMPS in every "bulb", after two of them caught on fire I decided it was high time to get rid of them.

 

What I am using now are 12V LEDs (either spotlights or ribbons), that way I can choose a much better PS than what you get in the individual "bulbs". I tried a bunch of supplies and wound up with some MeanWell units that work well, seem to be very reliable and don't emit as much stuff as others I have tried.

 

Dimming is a bit complicated, there are some nice 12V LED dimming modules available, but they all use PWM to turn the LEDs on and off at high frequencies, unfortunately that emits a fair amount of EMI.

 

I've come up with a solution that works, but is a little out of the ordinary. The 12V LEDs work by having three LEDs in series, each of which has a 3V drop, so the string has a 9V drop. They have a resistor in series that sets the current to the required amount across the 3V (12-9=3). So I'm working on a system that provides between 9 and 12V in several steps using good old linear regulators. At 9V you get no light, at 12V you get full light. In this situation it turns out the regulators are not dissipating much heat, as the voltage goes down the current goes way down. If your input is not too far above 12V the drop down to 12V does not produce very much heat either.

 

In the way over the top audiophile way I'm building units that are essentially JS-2s with the voltages adjustable from 9-12V. Each can run 5 12W LED fixtures or strings, with diming, without emitting EMI or crud back into the power line.

 

And to top it all off I found some REALLY good looking LEDs from a company called Soraa that will make this into a really good looking, efficient and electrically completely clean system.

 

Now I just need some time to get it all done!

 

John S.

[YashN]

Very cool plans, John.

 

Now I just need some time to get it all done!

 

My girlfriend says we'd need clones of myself as I seem to be too taken up often!