The null tester

[MLXXX]

2 hours ago, Addicted to music said:

Here’s the circuit from the video.

 

To get that null dial close to the center you would have to hand pick and measure and tested each opamp and resistors used.  It’s easy said than done.

Another approach would be to feed a high level balanced signal into the test cables, and then amplify any difference. The steps for the signal path would be as follows:

• test signal
• high level amplifier with closely matched 180 degree out of phase outputs:   A and A'
• interposed cable X for the A signal, interposed cable Y for the A' signal
• mixing of the signals emerging from the cables with a simple resistive network with a fine adjustment pad
• high gain amplifier

An advantage of the above is that it would probably be easier to implement a stable precision high level balanced amplifier stage, than a matched pair of stable high precision high gain amplifiers

 

Objections to testing that involves use of amplifiers to compare signals 

 

On the other hand, what if the "resolution" of the test signal was blocked by the high level amplifier with closely matched  signals A and A'?

 

To answer that objection, we could note that the "resolution" of the signal coming out of either cables X or Y could also be blocked using  "a matched pair of stable high precision high gain amplifiers". 

 

I don't feel I have a clear understanding of what "resolution" means in relation to a power amplifier.  A glossary of Audio terms referred to in 1993 (see http://www.integracoustics.com/MUG/MUG/bbs/stereophile_audio-glossary.html) provides this explanation:

 

definition (also resolution) That quality of sound reproduction which enables the listener to distinguish between, and follow the melodic lines of, the individual voices or instruments comprising a large performing group. See "focus."

 

 

I gather it isn't simply having low THD, a flat frequency response, and plenty of power. It's just a quality that some audiophile devices (such as power amplifiers) are reported to have that others are reported not to have, or not to have as much of. It typically enables some devices to let you hear more "air" and "detail", and a better "sound stage", than other devices.

 

A better "resolving" system may also give you a lower "noise floor". This specialised audiophile use of the term "noise floor" has emerged in relatively  recent years. It is not referred in the glossary  cited in 1993.  It isn't the meaning of "noise floor" as used by an engineer. So, it doesn't mean that there is necessarily a better measurable signal to noise ratio. It's just that the sound is "blacker". You can find out if an amplifier has a "low noise floor" by noting the listening comments of a reviewer. Also some people on this forum are able to assess the noise floor, and whether or not it has been improved with a change of cable. As far as I am aware, there is no method available to measure this type of noise floor with test instruments.

 

------------

 

With so much talk on this forum about very minor differences and whether they are audible, I yearn for positive reports about new technology that provides a demonstrable, marked, improvement. If only there could be a revolution in loudspeaker technology!

[davewantsmoore]

14 hours ago, Eggcup The Daft said:

Has the box he used been independently verified to contain the circuits shown in the diagrams?

Do we know what he is actually doing (as opposed to the claim) when he adjusts the controls?

No, but because we know of other null testing devices, the theory of how they work, and other result attained with them by a lot of people ...... then we know that what EW has presented is not an extraordinary result.    It doesn't require extraordinary evidence in order to accept it.

 

Indeed, if you wanted, you could build a null testing device, and try to invalidate his (and everyone else's) results.

 

On the other hand.... if EW had of made a video saying "hey guys, null testing doesn't work" ..... then he would need to have provided extraordinary evidence .... as the claim contradicts zillions of results which show it does, and the basic theory which predicts it would..

[davewantsmoore]

13 hours ago, MLXXX said:

I'm sure that wouldn't have been the case here

If they did have different attenuation, then the gain control needs to be adjusted for them to null.    That doesn't invalidate anything....  it only shows that different cables could have a teeny tiny differences in level (but that nothing else happens to the audio)

[davewantsmoore]

5 hours ago, MLXXX said:

I gather it isn't simply having low THD, a flat frequency response

That IS what it is.

 

Low linear distortion (ie. flat amplitude and phase response)

Low non-linear distortion (ie. no added frequency components which didn't exist in the input.... of which THD, and noise floor are examples)

 

That's it.  There is nothing else.

 

[acg]

17 hours ago, davewantsmoore said:

 

One scary thing about this pic, is that my cat Steve looks just like this.

 

Well, I've seen this now (thanks for that Dave) and I can't null Steve out of my mind...and who calls a cat Steve?  

[davewantsmoore]

9 minutes ago, acg said:

and who calls a cat Steve?  

Pets should have 'real' names.... people names.   You don't call your dog Spot, or Bluey .... You call them Craig, or Simon.    

 

Our other cats are named Alan, and Freya.

 

https://www.youtube.com/watch?v=XgvR3y5JCXg

 

Edited November 17, 2018 by davewantsmoore

[MLXXX]

1 hour ago, davewantsmoore said:

That IS what it is.

 

Low linear distortion (ie. flat amplitude and phase response)

Low non-linear distortion (ie. no added frequency components which didn't exist in the input.... of which THD, and noise floor are examples)

 

That's it.  There is nothing else.

 

With respect, that appears to be only one usage of the term, a technical usage.

I often read on this forum references to a highly resolving system and that appears to be purely a subjective evaluation. 

[acg]

35 minutes ago, davewantsmoore said:

Pets should have 'real' names.... people names.   You don't call your dog Spot, or Bluey .... You call them Craig, or Simon.    

 

Our other cats are named Alan, and Freya.

 

https://www.youtube.com/watch?v=XgvR3y5JCXg

 

 

But I've known plenty of dogs called Bluey...you may be surprised, we humans are not always particularly inventive when naming cattle dogs.  Also plenty of blokes called Bluey as well.

 

So, back to the topic, being a practical kind of guy I always prefer to see things tested in the way that they are actually used i.e. tested in use if possible.  For example, I might be looking for a new plough and find one that is very heavy and uses large girths of steel with plenty of wall thickness, over-specified bearings etc. and think that it will last me my lifetime if I were to buy it.  Get it home and stick it behind the tractor and I'm busting welds and bending this heavy steel and although it met my original specifications when shopping for the plough in real life, when I used the thing, those spec's were not suitable for telling me if the damn thing would last my lifetime or not.  That's an extreme example, but testing in real-world situations is where things really matter and this null test machine does not meet that primary criteria for me because the cable is being tested in isolation from the audio system. 

 

Not that I am in any way disputing EW's null test results or his honesty, but I just think that if he were to test the wires/cables in the audio system the result would be more meaningful.  Right near the start of the video he says... 

Quote

"audio engineers will tell you that it is simple to measure the wires to learn how they affect the audio passing through them."

 

Here is an audio engineer making a somewhat different claim that transmission lines effects are significant at audio frequencies, and he is doing a real time null test in an actual representative audio system with a cd player, amplifier and speakers...

 

 

 

 

 

 

 

 

EW also claims...

Quote

"the only thing wire can affect is overall volume level and frequency response due to its resistance, capacitance and inductance"

 

I am disputing this, because when these wires are used in the real world for their real purpose other things may become important such as shielding.

 

Anyway, I'm not presenting the Townshend videos above as gospel, but I do find them quite interesting nonetheless.

 

 

 

  

 

[MLXXX]

16 minutes ago, acg said:

Anyway, I'm not presenting the Townshend videos above as gospel, but I do find them quite interesting nonetheless.

 

Unfortunatey the following material from their website does not use the language of engineering. It uses "audiophile-speak":-

 

Townshend Audio strives to make the most transparent, revealing and uncoloured audio products possible. The importance of scientifically designed and fully engineered cables is hard to overestimate. 

Townshend Audio employ a number of unique concepts when manufacturing the near perfect ideal cable. We pioneered Enhanced™ Deep Cryogenic Treatment or EDCT for audio cables,

a technique that is so effective it has been copied across the cable industry and beyond. 

We have now gone a step further and developed Fractal-Wire™. This gives copper wire a characteristic of absolute silence and unbelievable purity. Initially, we incorporated Fractal-Wire™ into the signal connections and transformers of our components, but now we are also offering F1 Fractal-Wire™ interconnects.

Most cables merely change the balance of a system, where as the entire range of Townshend cables has been developed to give the greatest transparency with the least amount of character. Put these cables into a well-balanced system and it will reveal more, much more, guaranteed! 

Your system will deliver deeper bass, more open midrange and greater treble sparkle than anything else.

 

 

[bob_m_54]

1 hour ago, davewantsmoore said:

Pets should have 'real' names.... people names.   You don't call your dog Spot, or Bluey .... You call them Craig, or Simon.    

 

Our other cats are named Alan, and Freya.

 

https://www.youtube.com/watch?v=XgvR3y5JCXg

 

Our old across-the-road neighbour used to have a dog called Jeffrey (or it could have been Geoffrey). Little bugger used to pee on my front door mat.....

[MLXXX]

1 hour ago, acg said:

Here is an audio engineer making a somewhat different claim that transmission lines effects are significant at audio frequencies, and he is doing a real time null test in an actual representative audio system with a cd player, amplifier and speakers...

The demonstration shows that differences are visible on a dual trace oscilloscope. With long speaker leads this result is fully to be expected.

 

The question I would raise is whether the differences are "significant" for human hearing. For that you'd need to set up an actual test involving human subjects.  Just as you would prefer to see a plough in action, I would prefer to see human beings listening to music!

 

(I note that 2kHz square waves with very fast rise and fall times don't exist in nature. If you used a special ultrasonic loudspeaker you might be able to create an approximation to a 2kHz square wave in air, an approximation  with fast rise and fall times with frequency components extending well into the ultrasonic range. If you then placed a studio microphone near the loudspeaker and examined the analogue output of the microphone on a scope you'd see distinctly rounded edges. Typically, published response curves for studio microphone stop at around 20kHz, even though there would be some response beyond.

 

A modern audio power amplifier would do a much better job of reproducing an approximation to a square wave than a studio microphone would do attempting to capture the sound of an approximation to a square wave.

 

As for CD players and human ears they have an upper limit for frequency components of around 20kHz. You can't get CD players to give you perfect square waves. A square wave is a theoretical ideal.  CD players can give slightly rounded square waves. And human ears will respond to slightly rounded square waves. )

Edited November 18, 2018 by MLXXX

[LHC]

3 hours ago, davewantsmoore said:

Indeed, if you wanted, you could build a null testing device, and try to invalidate his (and everyone else's) results.

Of course people should do that always, that is what a science approach is all about, verifying the repeatability of results. The additional step required is to publish the results in peer reviewed scientific journals to complete the scientific process. Simply demonstrating results on youtube is not science. 

 

Actually done a quick search of Google Scholar and I couldn't find any peer reviewed scientific papers written by Ethan Winer. Maybe someone could cite some of them for references. He has written a book.

 

1 hour ago, acg said:

Not that I am in any way disputing EW's null test results or his honesty, but I just think that if he were to test the wires/cables in the audio system the result would be more meaningful. 

Agree that 'real world' testing is needed also to ensure the 'lab results' have taken account of all contributing factors, known or unknown. For example (but outside of audio) car safety ratings are lab-based using staged car crashes and impact on dummies; but this should also be correlated with real road accident results collected over time (this data can be available) to be complete. 

[davewantsmoore]

1 hour ago, acg said:

testing in real-world situations is where things really matter and this null test machine does not meet that primary criteria for me because the cable is being tested in isolation from the audio system. 

So in this case, you could build/buy a null test circuit, which you can hook up the whole system to....  but it becomes exponentially harder to draw conclusions from the result

 

31 minutes ago, MLXXX said:

I note that 2kHz square waves with very fast rise and fall times don't exist in nature.

That's true.... but they are still good test signals, because they tell you about the worst case scenario.

 

The key thing, is when you pass a square wave through a device ....  you need to compare the result with not the original square wave.   We know that is not a good comparison, because we know the device isn't going to do that.    We need to compare withe result with what we expect in theory the result to be.    Eg. we might know that the frequency response of the device is X .... and we need to look at a similarly band limited square wave.

 

That's the things with "tests".   If you do them wrong (either accidentally or on purpose), or interpret the results wrong .... then they can tell you anything you want them to, if you really squint   

[davewantsmoore]

33 minutes ago, LHC said:

Simply demonstrating results on youtube is not science. 

Rubbish.

 

FWIW....  "verifying the repeatability of a result" is boring, and is not published.    What people are really trying to do, is to present results which contradict previous results.

 

33 minutes ago, LHC said:

Actually done a quick search of Google Scholar and I couldn't find any peer reviewed scientific papers written by Ethan Winer.

So what?   What he presented in the video is boring basic stuff, that hasn't been controversial for nearly 200 years.

 

He made no new claims in the video - and what he did hardly requires "peer review".

 

33 minutes ago, LHC said:

For example (but outside of audio) car safety ratings are lab-based using staged car crashes and impact on dummies; but this should also be correlated with real road accident results collected over time (this data can be available) to be complete. 

I don't understand what wasn't "real world" about the experiment  (aside from not using an entire audio system - but the experiment was specifically about "a cable").

 

He used a real cable... a real audio device .... real (ly bad) music.     This was not a "staged car crash".

 

[LHC]

2 minutes ago, davewantsmoore said:

He made no new claims in the video - and what he did hardly requires "peer review".

If so, then what is the actual purpose of that video? If peer reviewed results already exist then why not simply point viewers to them? Of course explaining what the cited papers mean in layman's terms is worthwhile as a youtube video. But that is science communication, not the science itself. 

[LHC]

8 minutes ago, davewantsmoore said:

I don't understand what wasn't "real world" about the experiment  (aside from not using an entire audio system - but the experiment was specifically about "a cable").

He should lay his cables over carpet 

[LHC]

21 minutes ago, davewantsmoore said:

So in this case, you could build/buy a null test circuit, which you can hook up the whole system to....  but it becomes exponentially harder to draw conclusions from the result

Yes and no. In a whole system there may be other factors that could upset a null. But if a null is still achieved, then that would be a more convincing result. 

[davewantsmoore]

4 minutes ago, LHC said:

If so, then what is the actual purpose of that video?

To demonstrate to people who might be unaware of it.  Like did you even read the video description?!?

Quote

This video explains what a null test is and how it works, then uses a custom made Null Tester device to compare RCA wires

 

4 minutes ago, LHC said:

If peer reviewed results already exist then why not simply point viewers to them?

Cos normal people don't read papers ... they watch "demo videos".

 

I don't like to a "peer reviewed paper" every time I mention "gravity", or the observation that water boils when it's heated up.

 

[davewantsmoore]

4 minutes ago, LHC said:

But if a null is still achieved

A "null" will always be achieved .... it is just a question of how loud is the difference signal (and the content) .... and what you then conclude that result means.

[LHC]

On 18/11/2018 at 12:00 PM, davewantsmoore said:

To demonstrate to people who might be unaware of it.  Like did you even read the video description?!?

 

EDIT: deleted my response as it wouldn't get us anywhere. 

 

EDIT: For anyone who really wants to find out the full purpose of making this video should visit that page on YouTube and the video description in full. Read also the comments made in their comment sections and the responses written by Ethan Winer himself. Then make your own mind on its agenda. 

Edited November 21, 2018 by LHC

[MLXXX]

1 minute ago, LHC said:

Yes and no. In a whole system there may be other factors that could upset a null. But if a null is still achieved, then that would be a more convincing result. 

Well I guess if you want to test whether beans in can A taste different to beans in can B (from a different supplier) it could be more persuasive to do the testing as part of a meal with wine, meat, and other vegetables.

 

Most people though would I think be happy just to taste test the beans in isolation. If they seemed to taste the same in a direct comparison out of the two cans, that would be the end of the matter!

 

However in the audiophile world there seems to be a determination to keep testing and testing and testing, trying to find even the smallest audible difference.

 

I find this so surprising when loudspeaker systems really do sound very very very different from each other.  There really is a whole world of differences there. Unlike with interconnects, and power cables, it is not disputed that different loudspeaker systems sound different.

 

If I want to improve my listening experience am I more likely to achieve that by changing an interconnect cable, or by upgrading my speakers? 

 

I sometimes question what utility this forum provides to me. Going around in circles talking about jitter, and power cables, can be mildly entertaining, and I have participated in such threads in recent times. But this is doing nothing to improve the sound I am experiencing from my hi-fi system.

[LHC]

5 minutes ago, MLXXX said:

However in the audiophile world there seems to be a determination to keep testing and testing and testing, trying to find even the smallest audible difference.

I hear you. This may be due in part to the wide variety of experiences that audiophiles have. For a start we mostly have different systems, set up, room environment, music content etc. This may be a weak thing to say, but at least we should be glad that people can agree on the need for testing, so audio is not a faith-based pursue. 

 

10 minutes ago, MLXXX said:

If I want to improve my listening experience am I more likely to achieve that by changing an interconnect cable, or by upgrading my speakers? 

I think (and hope) all audiophiles would agree with you on that. I would be shock if anyone thinks cable upgrades are more important than speaker choices. 

 

13 minutes ago, MLXXX said:

I sometimes question what utility this forum provides to me. Going around in circles talking about jitter, and power cables, can be mildly entertaining, and I have participated in such threads in recent times. But this is doing nothing to improve the sound I am experiencing from my hi-fi system.

In the survey on another thread, it would seem the large majority of members here are already quite satisfied with their hi-fi system. Perhaps the utility of this forum may be to help out the newbie to this hobby to get up to speed. I do wonder if the quantity of posts to speakers threads and to other component/tweaks threads may be disproportional to their importance.  

 

For myself I openly admit I still have an enormous amount to learn. While I am aware of null tests in the past, just following the discussion in the threads allowed me to gain a greater understanding of them. I appreciate that. 

[acg]

5 hours ago, MLXXX said:

Unfortunatey the following material from their website does not use the language of engineering. It uses "audiophile-speak":-

 

Townshend Audio strives to make the most transparent, revealing and uncoloured audio products possible. The importance of scientifically designed and fully engineered cables is hard to overestimate. 

Townshend Audio employ a number of unique concepts when manufacturing the near perfect ideal cable. We pioneered Enhanced™ Deep Cryogenic Treatment or EDCT for audio cables,

a technique that is so effective it has been copied across the cable industry and beyond. 

We have now gone a step further and developed Fractal-Wire™. This gives copper wire a characteristic of absolute silence and unbelievable purity. Initially, we incorporated Fractal-Wire™ into the signal connections and transformers of our components, but now we are also offering F1 Fractal-Wire™ interconnects.

Most cables merely change the balance of a system, where as the entire range of Townshend cables has been developed to give the greatest transparency with the least amount of character. Put these cables into a well-balanced system and it will reveal more, much more, guaranteed! 

Your system will deliver deeper bass, more open midrange and greater treble sparkle than anything else.

 

 

 

Unfortunately that is true.  But the guy is still an engineer, from WA if I remember correctly.

[acg]

3 hours ago, davewantsmoore said:

So in this case, you could build/buy a null test circuit, which you can hook up the whole system to....  but it becomes exponentially harder to draw conclusions from the result

 

 

That may be true, but Max (?) Townshend managed it and was playing music from a cd player and nulling (or not) while playing on a speaker.  Is it audible?  Probably is if my experience with low inductance speaker cables is reliable.

[acg]

3 hours ago, davewantsmoore said:

Cos normal people don't read papers ... they watch "demo videos".

 

That's me too if possible: sum it up and get it over with quickly.  What does that say about the world?  And about me?