Grizaudio Posted April 17 Posted April 17 On 16/04/2024 at 9:11 AM, frednork said: Thanks for posting this Torben. Very interesting. Are you able to describe what your setup was before and after the change and also any audible changes you perceived. Mark, why would you encourage this? Honestly. 1
Grizaudio Posted April 17 Posted April 17 (edited) Ok, so I tried to keep an open mind. I thought why not try this for my self.... but take the idea to the next level. Whitepaper to follow. I took my highest performing generic CAT cable, and literally stretched the cr^p out of it around a Teflon tape roll. To get the cable to stay in place, I just stretched the cable even more as hard as I could. I then wrapped my bass guitar strap around the middle section of the CAT cable, and then proceeded to pass the cable through my high quality vinyl label cleaning protector. ......connecting the modified cable between my Roon server and Streamer...... a veil has been lifted. My wife instantly heard the difference from the kitchen, its clear the mix of vinyl label protector and teflon tape provides a very detailed, smooth and nuanced presentation. I might try moving the label protector to the opposite cable end, and even incorporate a few additional metal objects in the design. No doubt its a keeper. Edited April 17 by Grizaudio 2 5
frednork Posted April 17 Posted April 17 26 minutes ago, Grizaudio said: Mark, why would you encourage this? Honestly. I guess I am just more curious and encouraging and interested in experimentation than some others. No one on this thread except @TRHH has tried this cable yet lots of very "firm" opinions . At least the config above has been shown to make a measurable difference. Most eth related gear comes with no measurements whatsoever so I would have thought you guys would at least be supportive of that. Whilst I appreciate the information regarding the minimum diameter spec for an eth cable (as I didnt know that) I may still choose to try it and potentially wreck one of my cables. I am allowed to if I want! If the thread (or me being interested in it) upsets people they can just ignore it (or me) of course.
frednork Posted April 17 Posted April 17 8 minutes ago, Grizaudio said: Ok, so I tried to keep an open mind. I thought why not try this for my self.... but take the idea to the next level. Whitepaper to follow. I took my highest performing generic CAT cable, and literally stretched the cr^p out of it around a Teflon tape roll. To get the cable to stay in place, I just stretched the cable even more as hard as I could. I then wrapped my bass guitar strap around the middle section of the CAT cable, and then proceeded to pass the cable through my high quality vinyl label cleaning protector. ......connecting the modified cable between my Roon server and Streamer...... a veil has been lifted. My wife instantly heard the difference from the kitchen, its clear the mix of vinyl label protector and teflon tape provides a very detailed, smooth and nuanced presentation. I might try moving the label protector to the opposite cable end, and even incorporate a few additional metal objects in the design. No doubt its a keeper. If you can show measurements that show an improvement as much as the op's cable I might give it a shot
March Audio Posted April 17 Posted April 17 (edited) 52 minutes ago, frednork said: No one on this thread except @TRHH has tried this cable yet lots of very "firm" opinions . At least the config above has been shown to make a measurable difference. The measurements show nothing useful and are inappropriate. They appear to be measuring RF emmissions using a coil wrapped in close quarters around the cable. This is the wrong method. It's not telling you the field strength, which BTW falls off rapidly with distance and becomes irrelevant. If the field strength from ethernet cable was, or considered a problem, it wouldn't pass any regulations which properly test for this sort of thing. Secondly, it doesn't consider the conducted emissions which are flowing straight into end device through the conductors themselves. Thirdly, why is it assumed that these RF emissions are causing any kind of problem in the audio output? Due to some vague and unsubstantiated comments in a Rob Watts video? It's a bit like the view that femto clocks in network switches are beneficial. It's an inappropriate and misinformed extrapolation based on the genuine fact that stable word clocks are beneficial inside DACs. It has no relevance in the switch. The measurements should have been at the DAC audio output to establish if there was any problem, or if the cables use made any change. Finally there is the issue of expectation bias. Of course an improvement is heard. 52 minutes ago, frednork said: If the thread (or me being interested in it) upsets people they can just ignore it (or me) of course I think a balanced set of views is more informative and useful. Edited April 17 by March Audio
March Audio Posted April 17 Posted April 17 (edited) I'm still sort of reeling from Watts claims that artifacts and noise floor modulation are audible at -350dB. I'm not sure if people understand just how extraordinary this claim is. Some context. The thermal noise generated by a simple resistor, sat there at 20 deg C is about 0.05uV rms (20Hz to 20kHz). Thos is about -150dB lower than the 2 volt RCA output of a DAC. The best dacs acheive a noise floor around -130dB. Domestic gear is really quite close to what is acheivable at room temperature. If you have a can of liquid nitrogen handy you could pour it over your dac and get maybe another 20dB. So Watts reckons you can hear things going on another 200dB below the noise level in the quietest electronics. Oh and don't forget that the acoustic noise levels in a domestic environment are way, way above this anyway OK..... Edited April 19 by March Audio 2
frednork Posted April 17 Posted April 17 37 minutes ago, March Audio said: The measurements show nothing useful and are inappropriate. They appear to be measuring RF emmissions using a coil wrapped in close quarters around the cable. This is the wrong method. It's not telling you the field strength, which BTW falls off rapidly with distance and becomes irrelevant. If the field strength from ethernet cable was, or considered a problem, it wouldn't pass any regulations which properly test for this sort of thing. Secondly, it doesn't consider the conducted emissions which are flowing straight into end device through the conductors themselves. Thirdly, why is it assumed that these RF emissions are causing any kind of problem in the audio output? Due to some vague and unsubstantiated comments in a Rob Watts video? It's a bit like the view that femto clocks in network switches are beneficial. It's an inappropriate and misinformed extrapolation based on the genuine fact that stable word clocks are beneficial inside DACs. It has no relevance in the switch. The measurements should have been at the DAC audio output to establish if there was any problem, or if the cables use made any change. Finally there is the issue of expectation bias. Of course an improvement is heard. I think a balanced set of views is more informative and useful. Hmm, well I might just try making this cable anyway. and (horror of horrors) LISTEN to it and then me myself and I will decide its benefit (or not) to me. Who knows , I may agree with you! 2
March Audio Posted April 17 Posted April 17 1 minute ago, frednork said: Hmm, well I might just try making this cable anyway. and (horror of horrors) LISTEN to it and then me myself and I will decide its benefit (or not) to me. Who knows , I may agree with you! It's your time and if you enjoy the exercise that's great. I mean it. Just don't fall into the trap of fooling yourself with sighted bias. 1
Guest Posted April 17 Posted April 17 28 minutes ago, March Audio said: I'm still sort of reading from Watts claims that artifacts and noise floor modulation are audible at -350dB. I'm not sure if people understand just how fantastic and far fetched this is. Some context. The thermal noise generated by a simple resistor, sat there at 20 deg C is about 0.05uV rms (20Hz to 20kHz). Thos is about -150dB lower than the 2 volt RCA output of a DAC. The best dacs acheive a noise floor around -130dB. Domestic gear is really quite close to what is acheivable at room temperature. If you have a can of liquid nitrogen handy you could pour it over your dac and get maybe another 25dB. So Watts reckons you can hear things going on another 200dB below the noise level in the quietest electronics. Oh and don't forget that the acoustic noise levels in a domestic environment are way, way above this anyway OK..... When is your DAC being released?
March Audio Posted April 17 Posted April 17 Just now, Marc said: When is your DAC being released? We don't have a dac in design.
Guest Posted April 17 Posted April 17 Then respectfully, as per the Commercial Interests Policy of the Website Guidelines: Quote You may not denigrate the product of a competitor nor use StereoNET to further business disputes. This includes questioning design choices, pricing, or making slurs. It does not matter if you think your answer is factual, e.g. “brand XXX has a very poor reliability record”. Whilst it may be true, such statements are not allowed by members with commercial interests. Rob Watts is a celebrated DAC designer with many successful products. He specialises in DACs. You specialise in Amplifiers and Speakers, so perhaps show some professional courtesy. 1
March Audio Posted April 17 Posted April 17 (edited) 29 minutes ago, Marc said: Then respectfully, as per the Commercial Interests Policy of the Website Guidelines: Rob Watts is a celebrated DAC designer with many successful products. He specialises in DACs. You specialise in Amplifiers and Speakers, so perhaps show some professional courtesy. Marc, I totally agree he has designed some excellent products. I used a Mojo as my daily driver for a long time. I have not denigrated his products or questioned his design choices or competence etc. It is not in any way professionally discourteous to disagree with his views in this particular area. The audibility of sounds that of such low level they can't be physically created. Views that are not supported by known physics, or psychoacoustic science. I am certainly not alone in my disagreement, it's been discussed elsewhere extensively. The audibility of noise floor modulation is not a subject that is exclusive to dac design. So to imply I need to be a DAC designer to comment is erroneous. There are some designers that think class d amps are no good. Some that disagree with my design rationale with speakers. That disagreement is not in any way discourteous to me. It's just a difference of opinion. Edited April 17 by March Audio 8 1
frednork Posted April 17 Posted April 17 @TRHH by coincidence I happen to have an TL-WR902AC lying around. I notice you use a fibre isolation step instead. Did you try the WR902AC as recommended in the blog? l
Guest Posted April 17 Posted April 17 10 hours ago, March Audio said: Marc, I totally agree he has designed some excellent products. I used a Mojo as my daily driver for a long time. I have not denigrated his products or questioned his design choices or competence etc. It is not in any way professionally discourteous to disagree with his views in this particular area. The audibility of sounds that of such low level they can't be physically created. Views that are not supported by known physics, or psychoacoustic science. I am certainly not alone in my disagreement, it's been discussed elsewhere extensively. The audibility of noise floor modulation is not a subject that is exclusive to dac design. So to imply I need to be a DAC designer to comment is erroneous. There are some designers that think class d amps are no good. Some that disagree with my design rationale with speakers. That disagreement is not in any way discourteous to me. It's just a difference of opinion. Fair call Alan. As you were. 3
Addicted to music Posted April 19 Posted April 19 Heres a measurement of the analog output of a $250US Topping D50III DAC. Note here that all the claimed "digital noise" "jitter" "SMPS Noise" "non bit perfect zeros and ones " "Digital induced RF noise or what ever catergory RF noise you want to labelled it" has literally no effect. What makes you think that physically putting a ethernet cable out of spec is going to resolved what its claimed to resolved? one of the worst thing you can do to any ethernet cable is to add a low pass filter unless its to reduced noise from the device to satisfy FCC rules. what counts is whats measured at the analog output of the dac, what happens from the source to the dac is irrelevant. 5
March Audio Posted April 19 Posted April 19 (edited) 41 minutes ago, Addicted to music said: what counts is whats measured at the analog output of the dac, what happens from the source to the dac is irrelevant. This. Unless the modifications you are making up front of the DAC have a discernible effect on the analogue audio output of the DAC, then those modifications are ineffectual and of no audible consequence. Edited April 19 by March Audio 4
Addicted to music Posted April 19 Posted April 19 (edited) 4 hours ago, Addicted to music said: Heres a measurement of the analog output of a $250US Topping D50III DAC. Note here that all the claimed "digital noise" "jitter" "SMPS Noise" "non bit perfect zeros and ones " "Digital induced RF noise or what ever catergory RF noise you want to labelled it" has literally no effect. What makes you think that physically putting a ethernet cable out of spec is going to resolved what its claimed to resolved? one of the worst thing you can do to any ethernet cable is to add a low pass filter unless its to reduced noise from the device to satisfy FCC rules. what counts is whats measured at the analog output of the dac, what happens from the source to the dac is irrelevant. Just to cover the grounds that it doesn't cover ethernet inputs, well heres a streamer that does. Even at this level, its a non issue. Edited: the reason digital is so good, is that its immune to what audiophile mantra brings to the table. Unlike analog, digital is not effected by the electrical interference that effects analog but its not fully immune. Receiving interface only responds to a transition from high to low or a low to high, anything in between is disregard, it doesnt care! Lets say, the transmission has a checksum that indicate A PACKET IS erroneous, with ethernet it resends the packet, if errors continue, it slows the transmission down till the packet is error free. Well you then think this will effect timing that comes into mind! Relax, most applications will have hardware RAM built in and the software ensures that it buffers the data to cater for interrupted transmission. Dont believe me, then look at the reliability of digital TV transmission to pass era of analog TV transmission. With digital TV, the picture quality is flawless from corner to corner; no snowing effect, bleeding or ghosting that plague analog TV transmission ever! it either works or it doesnt, and when digital doesnt work its usually a weak signal that makes it pixelated and studded and you get the blue screen or a frozen still picture. Edited April 19 by Addicted to music 2 1
bob_m_54 Posted April 19 Posted April 19 (edited) On 17/04/2024 at 9:07 PM, frednork said: If you can show measurements that show an improvement as much as the op's cable I might give it a shot I'll bet the crosstalk figures are lower on @Grizaudio's cable, due to the bend radius being not as severe as the OP's distorted cable. Given the two, and if they were the only cables available, I'd pick Griz's if I absolutely needed something to work. Edit: Oh, and the OP has yet to demonstrate any improvement at all in measurements. Edited April 19 by bob_m_54
frednork Posted April 20 Posted April 20 12 hours ago, bob_m_54 said: I'll bet the crosstalk figures are lower on @Grizaudio's cable, due to the bend radius being not as severe as the OP's distorted cable. Given the two, and if they were the only cables available, I'd pick Griz's if I absolutely needed something to work. Edit: Oh, and the OP has yet to demonstrate any improvement at all in measurements. All good, you do you and I'll do me. 1
Grizaudio Posted April 21 Posted April 21 What do you think Nvidia would think of this cable? A company that has exponentially increased compute and data throughput. Its ok to have an opinion, knock yourself out..... but it doesn't trump science, physics and or appropriate measurements. Worth a watch by the way... 1 1
Addicted to music Posted April 25 Posted April 25 On 21/04/2024 at 12:08 PM, Grizaudio said: What do you think Nvidia would think of this cable? A company that has exponentially increased compute and data throughput. Its ok to have an opinion, knock yourself out..... but it doesn't trump science, physics and or appropriate measurements. Worth a watch by the way... There's a lot of misconceptions especially when i got onto this site and it still exists as per the OP. I lot of this still goes on here on this site. When i look at where we were and where we are going its huge jump in advances. When i 1st started out, Telecom, now Telstra could only guarantee 9600 baud down a standard copper 600ohm telephone line, yet Fax manufacturers were pushing G3 Fax with a spec of 15,000 baud. At the end on the fax life cycle it was pushing 30,000 baud, down a standard copper telephone wires, and then they required ISDN lines when G4 was introduced. By the end of 2019, the same copper network with probable interface modifications was pushing 20mb/s down the same copper wire to the same exchange at the same distance from your house. A lot of basic electronics was used to justify "jitter" etc and the way impedance mismatching can cause "jitter" was the normal justification on this sites by certain members. Yes, impedance mismatch does present a huge drawback or restrictions, but....now here is the but, engineers are finding ways in overcoming these inherent restrictions and the speed down facing impedance mismatch in the line is no longer a problem as you can see from your link.
Addicted to music Posted April 29 Posted April 29 Here are 2 real and active examples of where ferrite cores are properly used. One is on the supply line of 2 identical motors, it’s used to suppressed the noise that’s comes from the arcing of the motor brushes. The other is on a 230VAC mains supply to a SMPS, it’s to suppress the noise coming of the SMPS back onto the mains so it complies with FCC rules.
Recommended Posts