-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
RCA cables make no difference, that's what they say
- Thread starter pma
- Start date
- Thread Starter
- #22
This is questionable, because the stray current induced voltages interfere with signal harmonic components depending on signal.
Though the RG-59 follows the generator distortion
the Generic #2 cable creates quite different spectral content even at signal harmonic frequencies. Shall we make audibility test?
This is a relative Scale.
You can’t take away from this measurements that the Noise is always 100dB below the signal.
You can Only see the relative difference of about ~13dB between the cheapest and the best cable.
So if you have Hum or HF noise from your PC you maybe get ~13dB more or less with better cables.
Can be more of a difference can be less of a difference depending on many parameters like your ground loop impedance and loop area
Why use a generator at all? your not testing Caleb distortion or loss but Noise so just terminate the and with a typical output source resistor.
@pma can you pleases make a test with a bunch of ferrites on the cable and without
- Thread Starter
- #25
For those interested, attached is a 100Hz generator signal with setup as described in post #1 and recorded through 50cm of RG-59 and 1m of Generic #2 cable.
Below is my ABX result (through Vivanco SR-750 headphones, FR already shown in another thread). Attached is also pkmetrics by @pkane DeltaWave software.
And I will go quite far - RCA cables in this test make bigger difference than an exchange of two good A/D converters. Same will apply for D/A converters.
Below is my ABX result (through Vivanco SR-750 headphones, FR already shown in another thread). Attached is also pkmetrics by @pkane DeltaWave software.
And I will go quite far - RCA cables in this test make bigger difference than an exchange of two good A/D converters. Same will apply for D/A converters.
foo_abx 2.0.2 report
foobar2000 v1.4.8
2021-11-08 15:11:40
File A: Generic_100.wav
SHA1: 25fe578aa217ff0881de6e149b48e30fe49d384a
File B: RG59_100.wav
SHA1: a32a02c10a7dd9f84f2a781e859d436ce9e2d6e3
Output:
ASIO : Focusrite USB ASIO
Crossfading: NO
15:11:40 : Test started.
15:11:54 : 01/01
15:12:01 : 02/02
15:12:07 : 03/03
15:12:14 : 04/04
15:12:25 : 04/05
15:12:46 : 05/06
15:12:54 : 06/07
15:13:04 : 07/08
15:13:15 : 08/09
15:13:30 : 08/10
15:13:38 : 08/11
15:13:48 : 09/12
15:13:57 : 09/13
15:14:03 : 10/14
15:14:12 : 11/15
15:14:23 : 12/16
15:14:23 : Test finished.
----------
Total: 12/16
Probability that you were guessing: 3.8%
-- signature --
14021b06a5de70343dd28ac0d96883abdcb05c69
Last edited:
Can I request for a formal abstract/summary of the experiment and its results?
Something like this?
"We have successfully rejected the hypothesis that different RCA cables of similar lengths (1 m and 0.5 m), with different constructions, can't affect the transmission of a sinwave of 100 Hz to a degree, that once amplified, the difference between both signals becomes audible by a human."
I feel there are a lot of "articles" in this site that just need a bit of work to become serious articles. I believe this site needs a space for publishing papers.
- Thread Starter
- #27
The case is that there are too many scenarios when you make an SE interconnection between 2 or more components
class I - class I same mains outlet
class I - class I different mains outlets (worst case)
class I - class II
class II - class II
Any generalization is difficult, the only "general advice" I would give is to use balanced interconnects everywhere.
Or you are good enough in instrumentation and localize the issue.
1:1 isolation mains transformers make a good job for individual components and are not too big up to some 250VA, so they can supply DACs or preamps well.
I think that this is one of the most underestimated and misunderstood part of the consumer audio.
tw99
Senior Member
@pma do you have a view on what difference would be heard in normal music between these cables ?
I think that what you're trying to show here is that there are differences in noise immunity between various cables and that this may be audible (e.g. in your test above, one cable clearly has a lot more 50hz than the other which perhaps is what you're listening for in the ABX test).
Whereas audiophiles will say that cable X has "more musical expression" or "deeper bass", or "amazing female voices" or something. I'm not sure that your measurements support evidence for this type of difference, do you agree ?
I think that what you're trying to show here is that there are differences in noise immunity between various cables and that this may be audible (e.g. in your test above, one cable clearly has a lot more 50hz than the other which perhaps is what you're listening for in the ABX test).
Whereas audiophiles will say that cable X has "more musical expression" or "deeper bass", or "amazing female voices" or something. I'm not sure that your measurements support evidence for this type of difference, do you agree ?
- Thread Starter
- #29
1) You know it would be like forecasting from oracle crystal ball, as I tried to explain, technical conditions like ground balancing currents, spectra of SMPS leaking currents etc. are numerous without just one standard condition.
2) You might be surprised, but 50Hz line as seen here is quite unimportant. It is the forest of almost equivalent level non-harmonic higher frequency components that are audible under critical listening conditions. They are seemingly low in level below -80dBr, but they are so numerous to become audible.
3) No my measurements are not about and not to support audiophile claims that you have mentioned. They are here to point out some technical issues that are not often mentioned and that are not well understood not only by common users. Look, if we are nit-picking -120dBr harmonic distortions, shall we overlook much higher level of disturbing signals that are very often in common RCA link based installations? And shall we say that cables do not matter if the difference as big as 20dB is quite common?
KSTR
Major Contributor
Most of the error of RCA cables is the balancing/leakage currents producing a voltage drop along the shield ( + connector) resistance.
Error Voltage = this interface resistance * shield current.
A whopping 1mV (-60dBV) is quickly developed along ohm of interface resistance (long RCA cable) and 1mA of injected current via direct GND (Class-I to Class-I) or capacitive Mains (Class-II to Class-I/II) coupling.
No way you could achieve this signal integrity in a normal user setup, using the dongle as a line-level DAC... and sometimes the induced error matters (enough ground loop issue threads everywhere) and sometimes it doesn't.
==> A way to objectively rate RCA cables is to measure DC gnd conductor resistance (to 10mOhms precision). The lower the better.
-----------
And the general rules of thumb for system wiring:
Error Voltage = this interface resistance * shield current.
A whopping 1mV (-60dBV) is quickly developed along ohm of interface resistance (long RCA cable) and 1mA of injected current via direct GND (Class-I to Class-I) or capacitive Mains (Class-II to Class-I/II) coupling.
Exactly. For example, I just measured the Apple USB-C dongle (seems to be a bit different than the one Amir test some years ago) and the only way to eliminate hum/buzz was to use fully galvanically isolated (USB and power) ADC measurement setup plus diverting the remaining balancing current away from the dongle:
No way you could achieve this signal integrity in a normal user setup, using the dongle as a line-level DAC... and sometimes the induced error matters (enough ground loop issue threads everywhere) and sometimes it doesn't.
==> A way to objectively rate RCA cables is to measure DC gnd conductor resistance (to 10mOhms precision). The lower the better.
-----------
And the general rules of thumb for system wiring:
- Make the GND wire + connector resistance of your RCA interconnects as low as possible. Short cables in general!
- If you have balanced inputs, use them, even with unbalanced senders (with the proper adapter cable which is a RCA 3-wire to XLR where GND (Pin1) and -IN (Pin3) are connected together at the sending end -- if you know the sender's source impedance, typically like 100Ohms, insert another 100 Ohms in the -IN (Pin3) feed).
- Make the Mains Supply impedance as high as possible (notably with all Class-I gear), using cable drums (10m or more), and btw, generously put clamp-on ferrites on every(!) cable in your system (sans speaker cables). We want the primary GND connections via our interconnects (and helper connections) to dominate the grounding scenario.
- Use the highest possible send level. If the input is overloaded, use a passive voltage divider right at the output end of the cable.
- Connect all unused RCA shells (which carry GND potential, that is) together with the shortest and thickest cables (or best: use a copper ground plane/grid to bond them down).
- If all of that still fails, use Mains isolation transformers and USB isolators etc. to break some of balancing current paths.
Last edited:
- Joined
- Aug 14, 2018
- Messages
- 2,731
- Likes
- 7,995
I don't think you should leave, and I really appreciate the test results you conducted and shared here.
The difficulty I have with the comments you've made is that the assertions you make are not supported by the measurements you have taken and posted here. There might be audible differences between RCA interconnects, but the measurements you have provided do not prove that. And to the extent those graphs suggest that any differences might or might not be audible, they pretty clearly seem to suggest the latter - not likely to be audible.
I have nothing but respect for the time and effort you have put into making those measurements. But if you think that running those cables with an amplifier would boost the noise and differences to a level that wold easily be audible, then why not take such measurements?
With respect, I have found this to be a pretty consistent pattern with your comments in many threads here: you cite technical data or principles, and then you make claims based on that data or principles - but your claims go beyond what those data or principles actually prove or show. And then you get exasperated when people don't simply take your word for it that you are right.
I hope I am not causing offense (and happy to edit or delete this comment of the mods wish). I am just observing what appears to be a pattern of discussion that might be leading to some dissatisfaction on your part and perhaps some miscommunication in threads like these.
- Thread Starter
- #32
I think you have seen the post #25 in this thread and I suppose it is no proof to you
RCA cables make no difference, that's what they say
Hi @xaviescacs , thank you. I have such cables, but only XLR-XLR and TRS-XLR. As you have already stated, they are intended for professional use. I normally run 10m of XLR-XLR and 3m of TRS-XLR without smallest issues and interferences, but yes, in balanced links. I would need to cut some of...
www.audiosciencereview.com
I agree I do not go with the main stream, should it be either the subjective audiophile stream or the so called scientific stream which is based on secondary school textbook simplifications and limited technical knowledge. So your assessment is definitely right.
Last, I do not think I do not bring proofs to my claims. I think I am almost always trying to support my claims with measurements or test files on audibility. Contrary to most members here, who only come with opinions or personality assessments, which are both OT in a fact oriented forum
.
Last edited:
What I see is the following: you can perform an experiment in which a change of one RCA cable by another of similar length produces an audible difference. If this can be peer reviewed by someone and arranged properly, this is a valid and valuable scientific research. Not the most important discovery in the world, but this is how science gets done, step by step. Let's not underrate or underestimate the work being done here by people like you.Last, I do not think I do not bring proofs to my claims. I think I am almost always trying to support my claims with measurements or test files on audibility. Contrary to most members here, who only come with opinions or personality assessments, which are both OT in a fact oriented forum
KSTR
Major Contributor
If you make some back of an napkin type of calculations for the error level of @pma's graphs you'll immediately see he actually presented very benign cases, so probably inaudible (most of which depends on the downstream stuff).
As I said, it's not uncommon to see 100uV..1mV error voltages (-80dBV..-60dBV) and if that hits the 26dB gain of a power amp it will be audible in allmost all cases (2mV..20mV), and the root cause is the general design issue with RCA cables: Do NOT use when you have any mains balancing/leakage currents, but if you do, make sure those currents find another, better path than the cable shield.
- Thread Starter
- #36
Yes because they were only 2 class I components supplied from the same wall outlet and interconnected with only 35cm - 100cm cable length. The root of the problem however remains the same and I was hoping this thread would make it clear to the readers. The problem is much bigger in case of more components interconnected, more SMPS working on the same mains at the same moment, etc. The outcome should have been that the voltage drop across the screen of SE link cable may easily result in audible effect. And if something like non-coaxial topology is used, like twisted pair for SE link (see the 35cm Monster cable bad result and multiply it by usual length), the impact is much worse as there is a combination with poor EMI immunity, not only a mere voltage drop on the screen impedance. However it is difficult to explain to people who have no personal experience with the issue or not much understanding in the technical background. Of course I am not talking about you, Klaus.
- Joined
- Aug 14, 2018
- Messages
- 2,731
- Likes
- 7,995
I think you have seen the post #25 in this thread and I suppose it is no proof to you
RCA cables make no difference, that's what they say
Hi @xaviescacs , thank you. I have such cables, but only XLR-XLR and TRS-XLR. As you have already stated, they are intended for professional use. I normally run 10m of XLR-XLR and 3m of TRS-XLR without smallest issues and interferences, but yes, in balanced links. I would need to cut some of...
I agree I do not go with the main stream, should it be either the subjective audiophile stream or the so called scientific stream which is based on secondary school textbook simplifications and limited technical knowledge. So your assessment is definitely right.
Last, I do not think I do not bring proofs to my claims. I think I am almost always trying to support my claims with measurements or test files on audibility. Contrary to most members here, who only come with opinions or personality assessments, which are both OT in a fact oriented forum
I agree that you bring evidence and data to most of your comments here. My point, again with respect, was that the evidence you supply does not always support the claims you make, this thread being a great example.
As for not going with the mainstream subjective or objective viewpoint, I would say that one's self-positioning is the very definition of a subjective claim. We all find it appealing to claim that we're not at the extremes of a spectrum, that we're independent thinkers, that we're different than others in some positive way, and so on. But we're rarely as exceptional in our thinking as we like to believe - we all have some confirmation bias about ourselves.
KSTR
Major Contributor
As for audibility, we have to consider the RMS noise (20Hz...20kHz) when looking at hum/buzz spectra, and other artifacts as normal distortion, idle tones etc.
An example, this is the Apple USB-C Dongle (EU version) in two setups differing only in the balancing current flowing through 0.5m of 3.5mm to 3.55mm cable before it enters the measuring ADC. One time it is zero, the other time the injected mains leakage was that of an Intel NUC PC with its class-II SMPS supply flowing through the cable into the class-I Notebook.
The RMS noise floor was measured at -108dBr ref 0dBFS at the input, again displayed as dBV, sorry). A single sine at that level would have the same energy, so it is certainly audible when embedded in that noise, assumed we listen loud enough to actually hear the the tone and the noise floor. So we have some chances that it's audible occasionally.
The shaded area is the level below which a single tone within that noise fades being audible even at highest playback levels downstream, this happens at around -20dB. This is the more or less guaranteed ignore area for any noise and distortion even under most stringent circumstances (maximum gain after the DAC).
In this case most of the dirt is below that and so can be safely discarded as definitely inaudible.
Given that the actual level of that noise usually is at or even below the hearing threshold most of the time (otherwise normal playback levels would deafen you in a split-second), the chances of audibility are even less, in this specific example. Equal Loudness Contour Curves at work here, too.
But in a more complex setup hum/buzz etc levels may quickly reach audible levels, though. Typical is a PE-grounded PC with a soundcard feeding another class-I device like an amp or mixing desk etc. With separate mains outlets and several meters of RCA cable this more often than not an unusable connection, IME.
An example, this is the Apple USB-C Dongle (EU version) in two setups differing only in the balancing current flowing through 0.5m of 3.5mm to 3.55mm cable before it enters the measuring ADC. One time it is zero, the other time the injected mains leakage was that of an Intel NUC PC with its class-II SMPS supply flowing through the cable into the class-I Notebook.
The RMS noise floor was measured at -108dBr ref 0dBFS at the input, again displayed as dBV, sorry). A single sine at that level would have the same energy, so it is certainly audible when embedded in that noise, assumed we listen loud enough to actually hear the the tone and the noise floor. So we have some chances that it's audible occasionally.
The shaded area is the level below which a single tone within that noise fades being audible even at highest playback levels downstream, this happens at around -20dB. This is the more or less guaranteed ignore area for any noise and distortion even under most stringent circumstances (maximum gain after the DAC).
In this case most of the dirt is below that and so can be safely discarded as definitely inaudible.
Given that the actual level of that noise usually is at or even below the hearing threshold most of the time (otherwise normal playback levels would deafen you in a split-second), the chances of audibility are even less, in this specific example. Equal Loudness Contour Curves at work here, too.
But in a more complex setup hum/buzz etc levels may quickly reach audible levels, though. Typical is a PE-grounded PC with a soundcard feeding another class-I device like an amp or mixing desk etc. With separate mains outlets and several meters of RCA cable this more often than not an unusable connection, IME.
KSTR
Major Contributor
For completeness, find attached a recording of the dirty noise floor with 70dB of boost applied. The hum is audible within the noise, as is the HF dirt, though only by means of the high magnification factor with the noise now being elevated to -40dBFS rather than -110dBFS
Similar threads
