• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

RF Interference in Speaker Cables??? (video)

Joined
Sep 12, 2019
Messages
85
Likes
42
Connectors with silver contacts can be designed to be self-cleaning (i.e. clean themselves on connect/disconnect). Silver can handle higher currents than gold, which is important if you want the connector to survive hot-swapping. Gold also tends to wear off with repeated connect/disconnect cycles.

Tom
I suppose for pro applications the silver would be usefully more durable, but in the home I suspect a gold plated speaker connector would outlast the electronics it hooks up. The wear is very, very slow, especially when lubricated (who doesn't have DeoxIT lying around?)

Do you have any resources re: gold and current capability? I see lots of gold-plated connectors rated for 20 A or more.

From my link above:
There is an absorption layer built on top of the hard gold surface, which can be broken very easily with very low normal forces. Similarly, the coefficients of friction and, thus, the insertion and withdrawal forces of gold surfaces are rather low. This leads into very low wear and, therefore, to a high number of mating cycles. Thus, gold is also suitable for low voltages, corrosive environments and ideal for coatings of connectors in the data and signal range. The main disadvantage of hard gold is its high cost and its price-dependence on the stock market.
 

Francis Vaughan

Addicted to Fun and Learning
Forum Donor
Joined
Dec 6, 2018
Messages
810
Likes
3,777
Location
Adelaide Australia
I was going to mention the issue with silver contacts as well. I have seen mention that silver sulphide will not break down at six volts, and is a cause of failure in low voltage switching systems of poor design. When one sees otherwise unquantified mention of low voltages and high voltages, splitting them along the line of what will kill you and what won't is a good first approximation. Nearly everything in domestic audio is low voltages.

Tarnished silver in the modern world is going to be sulphides as much as anything else.

Don't underestimate nickel as a contact material.
 

tomchr

Addicted to Fun and Learning
Technical Expert
Manufacturer
Forum Donor
Joined
Nov 5, 2018
Messages
654
Likes
1,872
Location
Calgary, Canada
That's why relays intended for switching low-voltage, low-current signals, such as audio, have gold contacts.

Tom
 

amm

New Member
Joined
Feb 23, 2021
Messages
2
Likes
4
I have seen this happen firsthand at RF, but I'm not sure this is what's going on at audio frequencies.
PIM can also occur at lower frequencies (like in corroded contacts that can act like a diode). At higher frequencies due to current crowding, its effect is more pronounced. It is also heavily investigated in RF systems because of their wide dynamic range and high sensitivities. Receivers can have a noise floor of -150 dBm (10 to the power of -18 W) or lower while transmitters can produce very high power levels. I am not an audio expert, but given that people don't worry about it at audio frequencies tells me it is not audible.
 
Joined
Sep 12, 2019
Messages
85
Likes
42
That's why relays intended for switching low-voltage, low-current signals, such as audio, have gold contacts.

Tom
Yes. My point still stands that gold-plated contacts should be good for speaker-level connections as well, and better for the home than basically all the non-noble or semi-noble alternatives.
 

DonH56

Major Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
5,111
Likes
8,758
Location
Monument, CO
Wow, this has headed off... Current-carrying capacity gets a bit complicated since it depends upon cross-sectional area (resistance), cooling (thermal) capacity (usually better for shorter wires, for exposed vs. insulated wires, what insulation etc.), melting point (Cu>Au>Si but they are all around 1000 degC/2000 degF IIRC), etc. I do not know off-hand (not my area), but since they are not terribly far off in resistance or melting point, I would not expect a great deal of difference among them. Things like nickel and chromium (often used as a protective plating over copper and under gold) are much harder and have higher melting points but also much higher resistance.

IMO -- this is not my area of expertise! - Don
 

jbrown

Member
Forum Donor
Joined
Oct 21, 2019
Messages
5
Likes
2
Location
Florida
Did you even read the article? Granted, you can certainly criticize it for the low number of participants, but the methodology is solid, the data are there, and the statistical analysis chosen to analyze the data is appropriate for the sample size. Is that not empirical enough?

Tom
Yes I read it ... in 2018 and again a few days ago. I am sure you read it too. The question should be "did either of us understand it?" And if you are flying it's flag as a reference then I believe that perhaps you did not understand it, or at least didn’t see its flaws.

Here is my take:
1. The results were mixed. Plain and simple. They achieved "psychoacoustical significance" on two tests with non-level-matched pink noise. Nothing with music. That may make you think there is no difference, but ...
2. The samples weren’t all level-matched and to no one’s surprise the non-level matched tests were the ones most people got right.
3. They did not qualify the "listeners" hearing abilities or if they could even detect proper pitch. Nine of the listeners were self-proclaimed audiophiles. I know plenty of audiophiles who can’t hear a darn thing. The other two were high school kids. Applying scientific methodology to non-scientifically attained results does not equal science. When your "meter" is not calibrated, you cannot trust the results.
4. The listeners knew they were comparing speaker cables and were even allowed to view them all first and develop opinions. If you want to do a blind test you shouldn’t even tell the subjects what they are comparing, much less allow open inspection and listening before the blind test begins.
5. The "high end cable" was nothing special. The 2018 re-hash states "Monster’s early speaker cable was essentially also twisted-filament copper like the other test subjects, with nothing particularly advanced in the cable design". So they could have just doubled or tripled up on the 16 AWG lamp cord and been pretty close to the "high-end" cable.
6. What is a “laboratory grade audio comparator” (ABX)? Sounds like a relay to me. They don’t mention whether they tested it for transparency. They don’t mention what cable was used to go from the amp to the input of the ABX. They don’t even say they tested the fidelity of the source amplifier or speakers (I am declining to recognize their Audiophile "street cred" since so many supposed audiophile devices have been debunked on this very site). I wonder where the Perreaux, Spendor, and Kef control devices would score in Amir’s tests? Were they even able to resolve enough info as a system to allow the cables to be a difference-maker?
7. The author states that the original editors set out with an agenda to debunk high-end cables. That’s not journalism, it’s an opinion piece and it appears to be propped up by the same types of shaky science the cable manufacturers use to justify their products.

So yeah. Not good in my opinion.
 

tomchr

Addicted to Fun and Learning
Technical Expert
Manufacturer
Forum Donor
Joined
Nov 5, 2018
Messages
654
Likes
1,872
Location
Calgary, Canada
I agree that the methodology could be better. I would have liked to see more participants. N ≥ 20 would have allowed them to use the various statistical analyses that assume normality (says the Central Limits Theorem). Had they done so, I bet they would have found no statistically significant difference between the cables, at least not for any reasonable P value (such as the P ≤ 0.05 normally used in humanistic research).

Level match: They ran the experiment in two conditions: 1) Level matched and 2) constant volume setting. In the constant volume setting, they measured an 1- to 2-dB difference in level between the 24 AWG lamp cord and the 16 AWG cable.

Still, the level of scientific rigour in the article – by far – exceeds the common "me and my friends..." types of "experiments" typically run by audiophiles or reviewers.

Audiophiles vs others: Sean Olive demonstrated years ago that groups of various listeners (trained, untrained, college students, audio reviewers, etc.) show no statistically significant difference in their relative rank of a particular hifi product relative to other products in a blind test. The only significant difference he found between groups was that the trained listeners consistently ranked the equipment under test lower than did the rest of the groups.

I'll throw a couple of keywords at you for further exploration: Better-than-Average Effect; Confirmation Bias; Expectancy Bias. Also, I think you may be making the assumption that some cables are 'special' (per you #5 above). Gee. I wonder where that assumption comes from. :)

Tom
 

jbrown

Member
Forum Donor
Joined
Oct 21, 2019
Messages
5
Likes
2
Location
Florida
Level match: They ran the experiment in two conditions: 1) Level matched and 2) constant volume setting. In the constant volume setting, they measured an 1- to 2-dB difference in level between the 24 AWG lamp cord and the 16 AWG cable
It has been proven 1000 times over ... louder = better when being reviewed in blind tests. As such, every non-level-matched listening test is immediately invalidated in my opinion.

Still, the level of scientific rigour in the article – by far – exceeds the common "me and my friends..." types of "experiments" typically run by audiophiles or reviewers.
Better than garbage? Yes. Good enough to use as a reference? Nope.

Audiophiles vs others: Sean Olive demonstrated years ago that groups of various listeners (trained, untrained, college students, audio reviewers, etc.) show no statistically significant difference in their relative rank of a particular hifi product relative to other products in a blind test. The only significant difference he found between groups was that the trained listeners consistently ranked the equipment under test lower than did the rest of the groups.
I didn't say audiophiles were better or worse, I said the authors did nothing to qualify the ability of the "experts" to actually be able to hear, as in a basic hearing test like you would give a child at school or an elderly person who needs a hearing aid or even a quick pitch test. Nothing. Uncalibrated instruments deliver inconsistent results, plain and simple.

I'll throw a couple of keywords at you for further exploration: Better-than-Average Effect; Confirmation Bias; Expectancy Bias. Also, I think you may be making the assumption that some cables are 'special' (per you #5 above). Gee. I wonder where that assumption comes from. :)
You can take your keywords and keep them to yourself as I will keep mine to myself. This thread is about RF interference, so if a cable in the SR test were "designed" to mitigate RF interference, then it would be relevant to this thread. BUT in the context of the SR article, my point was they did not use a "special" cable. They used a 3X lamp cord to try to debunk the "special" cables, which further invalidates the test.
 

tomchr

Addicted to Fun and Learning
Technical Expert
Manufacturer
Forum Donor
Joined
Nov 5, 2018
Messages
654
Likes
1,872
Location
Calgary, Canada
It has been proven 1000 times over ... louder = better when being reviewed in blind tests. As such, every non-level-matched listening test is immediately invalidated in my opinion.
I agree. But, as I said above, they ran the test level matched as well. It's indicated clearly with a "levels matched" note in the data tables. It's mentioned in the text as well.

Anyway. Welcome to ASR.

Tom
 
Joined
Sep 12, 2019
Messages
85
Likes
42
Interestingly, speakON connectors tend to use silver contact platings instead of gold for some reason.
Since it's too late to edit my earlier post, I just wanted to comment that you can in fact get speakON connectors with gold-plated contacts instead of their standard silver-plated. Anyway, that's the end of this off-topic-ness...
 

pma

Major Contributor
Forum Donor
Joined
Feb 23, 2019
Messages
1,604
Likes
2,572
Location
Prague
I get this level of RFI pick-up on Magnat Monitor 102 speaker + about 2m speaker cable laying freely on the ground near common mains cables etc. The notebook works with large LCD monitor which is grounded to mains PE. This makes some difference above 1MHz noise voltages.

First, the setup photo
RF_speaker_spectrum_setup.jpg



Time record, 1Mohm input as per photo
RF_speaker.png


Spectrum record, 1Mohm input
RF_speaker_spectrum.png

The first 3 spectral lines (600 - 1300kHz) are local AM transmitters

Spectrum record, 50 ohm throughput terminator at scope input
RF_speaker_spectrum_50R.png

Again dominated by local AM transmitters, with decay of industrial VHF mess above 3MHz.

This is then perfectly cleaned with a 2xRC 30kHz passive lowpass filter.
RF_speaker_spectrum_50R+LP.png



setup with 50ohm terminator + filter
RF_speaker_spectrum_setup_filter.jpg
 
Last edited:

tomchr

Addicted to Fun and Learning
Technical Expert
Manufacturer
Forum Donor
Joined
Nov 5, 2018
Messages
654
Likes
1,872
Location
Calgary, Canada
Your o'scope shot shows a roughly periodic signal with a period of ~25 us (-> f = 40 kHz). I bet that's a switching power supply. Try turning your LED lights off. Turn your external monitor off as well.

I use a monitor with an external power brick for the PC that drives my APx525. I've tried with a monitor that has a built-in supply as well, but the switching frequency of the SMPS in the monitor gets into the measurement when I do that.

Thanks for posting the bandwidth limited measurement as well. It just goes to show that just because you can measure a disturbance doesn't mean you can hear the disturbance. :)

Tom
 

pma

Major Contributor
Forum Donor
Joined
Feb 23, 2019
Messages
1,604
Likes
2,572
Location
Prague
Yes the 40kHz repetition comes from numerous SMPS that are on, however the AM transmitters are there at exact frequencies of their own. I agree it is a non issue in most cases, however cannot be disregarded completely. We know about audio components prone to demodulation. So it is a non issue in case that the component is properly designed and built. Shielded speaker cables reduce induced EMI level. Both the speaker crossover elements and the unshielded cable contribute to the captured EMI. Again re SMPS, almost everyone has them in his setup, so no reason to switch them off.
 

tomchr

Addicted to Fun and Learning
Technical Expert
Manufacturer
Forum Donor
Joined
Nov 5, 2018
Messages
654
Likes
1,872
Location
Calgary, Canada
Fair enough.

A shielded speaker cable might not work, actually. The speaker ground connection on an amp is typically not connected to the chassis, so the cable shield would just be yet another antenna. And all bets are off with BTL amps. :) You'd need to ground the shield to an RF ground (assuming such a thing existed).

Tom
 

pma

Major Contributor
Forum Donor
Joined
Feb 23, 2019
Messages
1,604
Likes
2,572
Location
Prague
And all bets are off with BTL amps. :) You'd need to ground the shield to an RF ground (assuming such a thing existed).
Thank you, however I know this (designing 40+ years measuring instruments for places like HV testing labs, short circuit tests of power circuit breakers, plasma guns or measuring tiny signals at 400kV voltage potential) . With BTL, you will use a shielded twist or quad, similar as for inverter power. The same for SE output and again the shield connected to a proper pint. And it works - the rest I will leave to you ;).

And the last comment - I almost never see an audio component designed with respect of VHF EMI suppression. This is almost always ignored. When you DIY, you may do it right.
 
Last edited:

pma

Major Contributor
Forum Donor
Joined
Feb 23, 2019
Messages
1,604
Likes
2,572
Location
Prague
I get this level of RFI pick-up on Magnat Monitor 102 speaker + about 2m speaker cable laying freely on the ground near common mains cables etc.
Now the same measurement, but the unshielded speaker Litz wire is replaced by a properly shielded cable. 50 ohm throughput terminated at the scope to eliminate reflections. No LPF filter. Same geometrical arrangement as yesterday. This is the result

RF_speaker_shielded.png


and spectrum
RF_speaker_spectrum_shielded.png



SMPS induced impulse bursts are almost gone, AM transmitters are gone, some HF noise modulation remaining in a negligible level. So it is up to the reader to evaluate what is better. Yes one may speculate on audibility. However the question if to do it right or wrong remains.
 

tomchr

Addicted to Fun and Learning
Technical Expert
Manufacturer
Forum Donor
Joined
Nov 5, 2018
Messages
654
Likes
1,872
Location
Calgary, Canada
With BTL, you will use a shielded twist or quad, similar as for inverter power. The same for SE output and again the shield connected to a proper pint. And it works - the rest I will leave to you ;).
Right. "With the shield connected to a proper point". You're absolutely correct. Would you show me the rear panel of a BTL audio amp that offers this "proper point"?

I know how star quad works too. :)

And the last comment - I almost never see an audio component designed with respect of VHF EMI suppression. This is almost always ignored. When you DIY, you may do it right.
I've seen several. You won't pass the FCC or CE requirements for EMI tolerance without some sort of filter.

Tom
 
Joined
Sep 12, 2019
Messages
85
Likes
42
Right. "With the shield connected to a proper point". You're absolutely correct. Would you show me the rear panel of a BTL audio amp that offers this "proper point"?

Tom
A small hole drilled & tapped to allow connection of a ring terminal? :)
 
Top Bottom