- Joined
- 6 Sep 2016
- Posts
- 10,483
I guess his measurement equipment can't detect those magical frequencies, and thereby your ears are far more sensitive than something which can measure again and again in the megahurtz range.
Lan noise reducer
- Thread starter hornetstinger
- Start date
Scope 200MHz (software option), Sig get 60MHz (software option).
test lead response..
Mains cable response..
You'll note the pk-pk in mV at the end as I slipped the connection.
The key is if the bandwidth of the audio components actually even registers GHz.
Rod Elliot: frequency-vs-gain.htm
As most amp designers look at square wave production being a way to test an amp, then that requires supporting a large number of harmonics (multiples) .. 110MHz would therefore only support 1/10 of that as a square wave for example.. You'll see the graph further down for one of the AD op amps.. it has a decreasing gain (amplification) for MHz up to 256MHz. However by the time it gets to above GHz that's more like serious attenuation.. so all depends on your circuit.. and the parts used.
Is there a response graph for the GHz range vs a standard power lead?
I agree that RF filtering is good - that's why I add filtering to my amp and decouple too with small ceramic caps. However the question here is what is the result of comparing with a standard power cord?
test lead response..
Mains cable response..
You'll note the pk-pk in mV at the end as I slipped the connection.
The key is if the bandwidth of the audio components actually even registers GHz.
Rod Elliot: frequency-vs-gain.htm
As most amp designers look at square wave production being a way to test an amp, then that requires supporting a large number of harmonics (multiples) .. 110MHz would therefore only support 1/10 of that as a square wave for example.. You'll see the graph further down for one of the AD op amps.. it has a decreasing gain (amplification) for MHz up to 256MHz. However by the time it gets to above GHz that's more like serious attenuation.. so all depends on your circuit.. and the parts used.
Is there a response graph for the GHz range vs a standard power lead?
I agree that RF filtering is good - that's why I add filtering to my amp and decouple too with small ceramic caps. However the question here is what is the result of comparing with a standard power cord?
Last edited:
I've never once seen the response graph for a Kimber mains cable. I would be very interested to see one however.
I am repeating myself, but the RA Kimber Kables have an effect on the audio. The effect is less harshness in the sound, instruments are more separated, it's just better sound quality.
If we ever do that group HiFi meet, I'm happy for you to test my cables on your scope.
EDIT. I did find these measurements on twisted cables. So the Kimber Kables should be doing something very similar if not better due to the multiple twisted cores.
Effect of Cable Twisting on Radiated Emissions
It is common knowledge in electronic cable design that twisting wire pairs together is usually a good thing. However, the benefits of twisting and the degree to which twisting is valuable is less w…
emianalyst.wordpress.com
Last edited:
Soldato
Instruments more separated
Hilarious that you think a cable can substantially alter the recording..
Hilarious that you think a cable can substantially alter the recording..
- Joined
- 6 Sep 2016
- Posts
- 10,483
In the megahurtz range lol.
Why doesn't jason send nickk a couple of cables and compare a £3 IEC cable to RA. Compare the speaker outputs from a sine wave 10-20khz etc.
Of course the recording does not change.
What does change is reduced noise into audio components. When you lower the noise you increase the detail.
Last edited:
- Joined
- 6 Sep 2016
- Posts
- 10,483
Do you understand how power supply and electricity storage for capacitors work?
I've never once seen the response graph for a Kimber mains cable. I would be very interested to see one however.
I am repeating myself, but the RA Kimber Kables have an effect on the audio. The effect is less harshness in the sound, instruments are more separated, it's just better sound quality.
If we ever do that group HiFi meet, I'm happy for you to test my cables on your scope.
EDIT. I did find these measurements on twisted cables. So the Kimber Kables should be doing something very similar if not better due to the multiple twisted cores.
Effect of Cable Twisting on Radiated Emissions
It is common knowledge in electronic cable design that twisting wire pairs together is usually a good thing. However, the benefits of twisting and the degree to which twisting is valuable is less w…
You’ll ne pleased to know that tube amps have prior art on twisting of wires to reduce radiated EMI. Most tube amps twist the AC heater wires and route them in the corner of the chassis to minimise radiated 50Hz hum.
I do have a problem that Russ Andrews states before the ASA he didn’t have the proof and after the ASA he did.. so basically he seems to be suggesting he was blocked as he was selling without evidence for many years to back up the claims then only when he had evidence did they unblock.. it seems to me yo suggest he was happy claiming prior to the ASA without having evidence, have a watch and make your own opinion: https://youtu.be/iGWu5bxX0Bw?si=MBoBds-Jtj_Zfkcu
Personally I would have evidence then sell it.. but that’s just me..
I would also point out:
* RF on the power line should be filtered before it gets into the chassis but at a point there’a no RF ingress (ie a shielded point on the chassis wall).
* RF radiated to your speaker cables would be relatively small compared to the speaker signal peaks travelling through the speaker cable. Granted a speaker sensitivity can buzz but then ask yourself - if that’s a GHz signal then now does the speaker radiate it (physically) and you hear it? So the issue is more todo with the packet being transmitted with a timing between comms packets that causes say 50-500Hz impulses.. your phone does create several dB impulse (front of the packet)…
So if the RF packet has induced a 50-500Hz hum then no filter can magically differentiate between RF-hum and audio.. except if it’s common mode noise (same noise current on both conductors at the same time) vs audio being differential mode (ie the ac current travelling if different directions on each conductor). Not sure I’d risk that on a bridged amp where the signal is differential to both sides of the speaker (or fancy designs like a circlotron where the power supply is floating on the audio signal
)
Last edited:
Seems that you can get different versions from 120 to 320Vac. It’s designed for surge (ie lightning etc) rather than noise.
Last edited:
Soldato
This is all snake oil because it's so badly misrepresented..
Networks are incredibly noisy, for starters they naturally create a massive amount of noise as they carry 250mhz (CAT 6) of digital signals, so have a very messy spectrum due to the complex harmonics.. The ethernet transformer/phy and all the filtering should take care of that, and certainly any modern power topology within a device would ensure that sensitive circuits are on a separate adequately filtered supply..
We make medical devices, some have sensitive measurements down to picoamps and ranging up to 5mhz, these are on the same PCBA as a computer module (x86, incredibly noisy) as well as 2 network switch ICs and 8 PHYs, not to mention all the RMII and SERDES lines for the highspeed switch ICs.. you'd imagine it would be impossible to measure complex signals that are a few picoamps on the same PCBA.. but it's not that hard.. You ensure as much isolation between power supply rails for each subsystem, you don't feed the same 3v3 to the network switch ICs as you do the ADCs for the measurement system and modern EMC best practices can be applied such that radiated emissions with housing do not become conducted noise within the measurement circuit.. we even use 'precision power supplies' for the most sensitive ICs...
The problem I have is that the claims are made to look like a product should universally be effective, but in reality it's really only going to apply if you have a badly designed susceptible device and ultimately that will be measurable..
And further to that, the biggest factor in the dynamics and overall presentation is the DAC and amplification sections, those are well downstream of any conducted immunity issues in any meaningfully modern design, and that's where the real quality starts and ends.. and 99% of these products can't and measurable don't affect those unless as mentioned someone has dropped the ball when designing them.
Networks are incredibly noisy, for starters they naturally create a massive amount of noise as they carry 250mhz (CAT 6) of digital signals, so have a very messy spectrum due to the complex harmonics.. The ethernet transformer/phy and all the filtering should take care of that, and certainly any modern power topology within a device would ensure that sensitive circuits are on a separate adequately filtered supply..
We make medical devices, some have sensitive measurements down to picoamps and ranging up to 5mhz, these are on the same PCBA as a computer module (x86, incredibly noisy) as well as 2 network switch ICs and 8 PHYs, not to mention all the RMII and SERDES lines for the highspeed switch ICs.. you'd imagine it would be impossible to measure complex signals that are a few picoamps on the same PCBA.. but it's not that hard.. You ensure as much isolation between power supply rails for each subsystem, you don't feed the same 3v3 to the network switch ICs as you do the ADCs for the measurement system and modern EMC best practices can be applied such that radiated emissions with housing do not become conducted noise within the measurement circuit.. we even use 'precision power supplies' for the most sensitive ICs...
The problem I have is that the claims are made to look like a product should universally be effective, but in reality it's really only going to apply if you have a badly designed susceptible device and ultimately that will be measurable..
And further to that, the biggest factor in the dynamics and overall presentation is the DAC and amplification sections, those are well downstream of any conducted immunity issues in any meaningfully modern design, and that's where the real quality starts and ends.. and 99% of these products can't and measurable don't affect those unless as mentioned someone has dropped the ball when designing them.
See, this sounds just wishy washy nonsense.. sure you could argue the higher the SNR the more 'detail' you get, but that's just far too vague, the ability to resolve detail from a DAC is by far more a function of the design of the DAC vs the SNR.. I'd take a 3db increase in the noise floor (therefore lower SNR) with a more precise DAC.. it's very very very difficult to distinguish between a 120db SNR and a 117db SNR.. which is why I don't wholly subscribe to the holy grail of SNR.. I'd rather more precision and more importantly how dynamic a DAC is which is all measurable.
Last edited:
Indeed add to that the random impedance matching and reflections, in addition to noise. This is why it's vitally important to isolate for digital signals. After that the digital signal (as it's powered by a quiet side supply) is far far cleaner.
I don't think filters on are snake oil, but they're being used to patch the obvious flaw in the design and being sold as a solution to everything. Key for me is removing the issues of RF and oscillations caused by bad design (using an RF capable opamp without taking the precautions - leading to a harsh sound). In the end the average audio geek is simply going to follow the herd, attribute the fix to the device but not actually be able to measure the impact other than what is perceived via the speakers..
Having designed and built, refined down to -150dB.. after -120-130dB it doesn't really make a difference however what does make a difference is ensuring it's quite across the spectrum.. no point loosing power. For digital that is not just the power supply inlet but the power supply noise caused by the internal components you're using (ie clocks etc) on the power rail. So I agree.
Last edited:
OK, instead of sitting on the sidelines throwing stones, I ordered and have just had delivered a Delock 62619, ie
www.reichelt.com
It’s up and running and the phrases that instantly sprung to mind were:
- wow
- veiled lifted
- punching above it’s weight
- far more “musical” and involving
Naw, only joking. I’m going to give it a little while with it in system before attempting to make any judgements.
More to the point, I don’t want to setup any potential future bias based upon solo testing.
So, I’m going to make the offer once again for those on either side of the bench to come and have an unsighted test. Happy to run it however you wish, but with the following conditions:
- I’m hosting and we’ll use my system. Details of my system are on page 2 of the user systems thread, aa attached to this forum
- not more than 6 attendees, it is after my home
- happy to run it in week if that works
I’m based in Surrey, within the M25.
Here’s you chance guys. The snake oil people say this thing works.
I’ve no idea either way. It wasn’t expensive enough to really be bothered.
If you feel that you want to genuinely contribute, do so from experience, not just because of current audio theory.
DELOCK 62619 - Network over-voltage protection 6 kV
Network isolator 6 kV RJ45The Delock RJ45 isolator protects devices from power surges on a connected network connection. The compact adapter supports Ethernet with 10/100/1000 MBit/s. Technical [...]
It’s up and running and the phrases that instantly sprung to mind were:
- wow
- veiled lifted
- punching above it’s weight
- far more “musical” and involving
Naw, only joking. I’m going to give it a little while with it in system before attempting to make any judgements.
More to the point, I don’t want to setup any potential future bias based upon solo testing.
So, I’m going to make the offer once again for those on either side of the bench to come and have an unsighted test. Happy to run it however you wish, but with the following conditions:
- I’m hosting and we’ll use my system. Details of my system are on page 2 of the user systems thread, aa attached to this forum
- not more than 6 attendees, it is after my home
- happy to run it in week if that works
I’m based in Surrey, within the M25.
Here’s you chance guys. The snake oil people say this thing works.
I’ve no idea either way. It wasn’t expensive enough to really be bothered.
If you feel that you want to genuinely contribute, do so from experience, not just because of current audio theory.
Soldato
It's snake oil, they claim "zero-jitter memory buffer and galvanically-isolated inputs", not only is the 'isolation' part laughable since their PCBA look to have just slapped down a bog standard Ethernet (pulse) transformer which any equipment that wanted isolation would already have, but worse, they are then mis-representing it as a 'zero-jitter memory buffer'.. that's misleading and downright snake oil IMO, although I'm open to correction but I'll stick to my guns since 'jitter' is not a problem on ethernet, it's all buffered and transmission rates are far higher than the consumption rate, which means the presentation is absolutely nothing to do with any form of jitter..
It's bad and needs to be called out, that £90 or several of those if you went mad sticking them in every orifice they fit in could fund a much better amp, dac, speakers, source, etc.. all things that measurably can improve the experience.
Last edited:
It's snake oil, they claim "zero-jitter memory buffer and galvanically-isolated inputs", not only is the 'isolation' part laughable since their PCBA look to have just slapped down a bog standard Ethernet (pulse) transformer which any equipment that wanted isolation would already have, but worse, they are then mis-representing it as a 'zero-jitter memory buffer'.. that's misleading and downright snake oil IMO, although I'm open to correction but I'll stick to my guns since 'jitter' is not a problem on ethernet, it's all buffered and transmission rates are far higher than the consumption rate, which means the presentation is absolutely nothing to do with any form of jitter..
It's bad and needs to be called out, that £90 or several of those if you went mad sticking them in every orifice they fit in could fund a much better amp, dac, speakers, source, etc.. all things that measurably can improve the experience.
Ahh I was talking in general - isolation using a digital isolator (where you have Schmitt triggers etc) reduces noise.. but the signal going across it will still be subject to jitter still as you're quantifying it with the Schmitt triggers. Can't see a buffer in those images.
This was mine - clock + 1:3 fan board and i2c + i2s digital isolators PCB, my first ever PCB design. The additional circuitry is for a couple of ways to add a shunt supply:
I think I could have done a better job than their board
Last edited:
Soldato
Awesome, good work!Ahh I was talking in general - isolation using a digital isolator (where you have Schmitt triggers etc) reduces noise.. but the signal going across it will still be subject to jitter still as you're quantifying it with the Schmitt triggers. Can't see a buffer in those images.
This was mine - clock + 1:3 fan board and i2c + i2s digital isolators PCB, my first ever PCB design. The additional circuitry is for a couple of ways to add a shunt supply:
I think I could have done a better job than their board
When recruiting for Electronics engineers we have a small example of some layout with varying errors and some more minor items, right down to aesthetic issues. We only care if candidates get the obvious functionally related errors, then best practices, and they do get bonus points if they care about the aesthetis because it’s not a requirement, but it shows pride and that is a very hidden and imo valuable attribute..
Soldato
LAN iSilencer by iFi audio - LAN iSilencer
Connect the LAN iSilencer between your broadband router and network audio streamer to remove electrical noise that distorts the audio signal.ifi-audio.com
Hmmmm.
I believe there are scam routers and switches as well. People charge hundreds of pounds more for gear that has been sealed in resin, so you can't see that nothing has been done to it. But amazingly, some people have bought these things.
Then the same applied to cables. People will pay thousands of pounds for audio cables.
You even see that sort of behaviour for network cables. On Amazon, 24000 people have bought ugreen Cat8 cables and, reading the comments, it's clear that most of them think it will perform better than a Cat5e with their home router.
HOW AND WHY we test network switches...
We love streaming audio at Alpha Audio. But it's not because it's plug and play... because it isn't. There is a lot involved when you want to get the best po...
youtu.be
Watch video where he measures noise from the switches.
This thread is on 16 pages, yet things can be demonstrated quickly if actually trying them. The passive Russ Andrews mains conditioners for example, the difference in audio can be heard during playback as you plug in or out a conditioner from the mains. Strong views are fine, but people should remain curious enought to try for themselves first.
Last edited:
The network cable is the least important cable. I use Amazon Basic Cat 7 as they have a very good outer braid on them, they can be purchased for around £5 on Amazon Warehouse. Using a good shielded network cable is good audio hygiene. That said I have never noticed an audio difference from a network cable