[rodthebod]

TBH I struggled with the word “hobbled” as I typed it. I couldn’t come up with a better term. What’s in my mind is that external clocks have a lot more challenges than internal ones which are “close to the action”. Surely the external clocks must first of all overcome the downside of being external. I’m sure I don’t understand enough of the technicalities, it just seems unfortunate that what could be an excellent internal clock can be improved on even in the face of the adversity that comes with being external…maybe being external isn’t all downside, ie power supply and RF shielding.

This is why wanting to be "close to the action" I thought placing the LB inside the R26 with, what eventually will be, a very short 50mm connection (with a proper solid RF copper screened link) made good sense to me.

I'm awaiting delivery of alternative PCB MCX connections for the LB PCB to achieve this, doing away with the clumsy BNC connector. RF shielding is being addressed for the LB board as well as a better PSU. As Jake discovered the LB benefits from a better LPS, so that's what I'll give it.

It just takes weeks to get parts! Copper tape and MCX connectors etc. I can try the LPS now though.

 

[Stellabagpuss]

No need to wonder unless someone changed digital audio to no longer be based on sampling and forgot to tell the world.
The science, math and implementation hasn’t changed either.

“So, although sonic differences may be perceived when using an external clock as compared to running on an internal clock, and those differences may even seem quite pleasant in some situations, this is entirely due to added intermodulation distortions and other clock‑recovery related artifacts rather than any real audio benefits, as the test plots illustrate.”

If you like the sound of your clock go buy one of these and save a ton of money!
https://diyaudiostore.com/products/h2-harmonic-generator

Yes l understand your view point, but l was talking about professionals in this field since 2010.

 

[Stellabagpuss]

This is why wanting to be "close to the action" I thought placing the LB inside the R26 with, what eventually will be, a very short 50mm connection (with a proper solid RF copper screened link) made good sense to me.

I'm awaiting delivery of alternative PCB MCX connections for the LB PCB to achieve this, doing away with the clumsy BNC connector. RF shielding is being addressed for the LB board as well as a better PSU. As Jake discovered the LB benefits from a better LPS, so that's what I'll give it.

It just takes weeks to get parts! Copper tape and MCX connectors etc. I can try the LPS now though.

lt's not much faster in the UK These days Rodders, still it gives you more time to listen, which in between painting, something l hope to do this weekend.

 

[piakoB]

Just a quick question to owners. Do R2R DACs in general have a latency between the music and the source? I was wondering if I could put this DAC into my office and also use it for multimedia (videos and stuff). Sometimes there is a lag between video and audio sync I heard on some DACs.

 

[sajunky]

“So, although sonic differences may be perceived when using an external clock as compared to running on an internal clock, and those differences may even seem quite pleasant in some situations, this is entirely due to added intermodulation distortions and other clock‑recovery related artifacts rather than any real audio benefits, as the test plots illustrate.”

It sounds like a mumble, fitting entirely in scientology vocabulary. On the other side there are attempts to define absolute levels of noise by scientologist.



However it prompts some creative thinking, i.e. I never thought of external clocks in terms of intermodulation characteristics. Do we have reliable methods of measuring correlation level between phase noise and a signal? The only method is visual. Periodic noise of a given frequency shows a spike on FFT plot. As a test signal is periodic, there is a suspicion that spikes around a lobe come in result of intermodulation with test signal.

The reviewer of a Gustard is right, it is all wrong. It shows reaction of a clock, changing frequency periodically acording to the amplitude, most likely it is the tests signal. A smooth lobe shows a random noise, not influenced by a test signal. A smooth, even a wider lobe is not so harmful to our ears as series of spikes.

I want to draw a parallel to distortions introduced in result of truncating word size or DSP mathematical errors. These errors are strictly correlated to the signal. Long time this problem was not recognised, milions CD's were produced just by truncating word lenght to 16-bits. They sounded very harsh.

A solution is simple: decorrelating noise by dithering. A noise revel is increasing in total, but if is random we can hear through the noise (recognise a sound below the noise level).

The same is (I suspect) with external clocks. If clock frequency periodically changes which we can see on a FFT plot as a narrow spike, in can be sourced by amplitude of the test signal, but it leads to intermodulation distortions which are strictly correlated with signal. Now, if we move a clock generation outside, we have a guarantee that any noise coming with a clock is not correlated to the sound. It can look not nice on the FFT plot, but sound is more refined.

 

[sajunky]

Just a quick question to owners. Do R2R DACs in general have a latency between the music and the source? I was wondering if I could put this DAC into my office and also use it for multimedia (videos and stuff). Sometimes there is a lag between video and audio sync I heard on some DACs.

Audio GD R2R-11 has no latency. Denafrips has a large latency, it is variable, cannot be compensated. For Gustard lets others talk.

 

[camrector]

Are you in the camp that say the stock R26 sounds the same as other DACs that measure as most modern DACs and are essentially perfect? I gather whilst the R26 may not measure the best, any difference will be inaudible.

I’m unsure of what your is question is. DACs sound different. I’m certainly NOT some ASR type. I wouldn’t have an R2R DACs if I was.

 

[rodthebod]

Just a little update, I was reading about damping material for capacitors and components via the Ali cable thread, Jake's trying out many tweaks. I used to do this back in the nineties!

Anyway, I thought what's the most reactive item in the chain as far as micro vibrations go........ the clock chip.

So I placed a tiny (2x2mm) pad made from the most dense neoprene rubber I had at hand (actually from a roll of double sided pads used for adhering interior car mirrors to glass - they were for sticking things to my old hi-fi stands -Stands Unique they are about a millimetre thick made by 3M you know...). You should see the pad (yellow PTFE backing strip) on the clock chip in the pic. I have posted a pic earlier of the whole pad used to protect the LB PCB from shorting with the R26 power caps.

I had made a crude RF shielded cap for the clock chip and strapped it to the earth on the LB PCB and have been evaluating this mod. This RF shield seems to work for improving almost everything in the sound. Maybe because the LB is inside the R26.

The addition of the damping pad, however, made an even bigger difference, I listened to Diana Krall 'No Moon at all' before and after. What a difference in clarity using the damping pad with the RF cap.
I can't see any problem with extra heat build up as you could argue the pad acts as an insulator, can anyone else?

Jake - try out your special fo.Q TA-102 damping pads maybe they are even more effective?

Onwards and upwards....

 

[camrector]

If the goal is to reproduce the analog signal precisely, then less jitter is better. Random jitter increases noise. Random jitter generates harmonics, some of that energy is inharmonic. Then there is jitter that is not random, such as noise pickup from radio transmission, power lines or other environmental disturbance in the clock cables. There is also a cable termination issues (tolerance of cable impedance and termination resistance) which causes reflections on the clock cable. A cable driver, a cable receiver Then there is the clock recovery circuitry (PLL) etc etc.
This is objective. If the idea is to keep as much of the music un altered, then jitter is your enemy. If you are for transparency, then less jitter is better.

 

[camrector]

intermodulation distortions which are strictly correlated with signal. Now, if we move a clock generation outside, we have a guarantee that any noise coming with a clock is not correlated to the sound.

The clock doesn’t process any signal. It generates a frequency that affects the signal. That’s is correlated.

 

[Stellabagpuss]

Just a little update, I was reading about damping material for capacitors and components via the Ali cable thread, Jake's trying out many tweaks. I used to do this back in the nineties!

Anyway, I thought what's the most reactive item in the chain as far as micro vibrations go........ the clock chip.

So I placed a tiny (2x2mm) pad made from the most dense neoprene rubber I had at hand (actually from a roll of double sided pads used for adhering interior car mirrors to glass - they were for sticking things to my old hi-fi stands -Stands Unique they are about a millimetre thick made by 3M you know...). You should see the pad (yellow PTFE backing strip) on the clock chip in the pic. I have posted a pic earlier of the whole pad used to protect the LB PCB from shorting with the R26 power caps.

I had made a crude RF shielded cap for the clock chip and strapped it to the earth on the LB PCB and have been evaluating this mod. This RF shield seems to work for improving almost everything in the sound. Maybe because the LB is inside the R26.

The addition of the damping pad, however, made an even bigger difference, I listened to Diana Krall 'No Moon at all' before and after. What a difference in clarity using the damping pad with the RF cap.
I can't see any problem with extra heat build up as you could argue the pad acts as an insulator, can anyone else?

Jake - try out your special fo.Q TA-102 damping pads maybe they are even more effective?

Onwards and upwards....

Now that sounds like a excellent Mod Rod
And one l will add to my ever growing list. l gave put copper shields on chips, but never considered a damping

 

[sajunky]

The clock doesn’t process any signal. It generates a frequency that affects the signal. That’s is correlated.

Be careful, as first sentence is simply not true. It does process a noise, unintentionally. Internal and external. A result is not ideal shape or timing.

A second one is not expressed clearly enough to bring for consideration. A third one bring a word 'correlated' in a different meaning and seems contradicting with the first sentence. I used a standard engineering term brought from theory of signal processing.

 

[camrector]

Be careful, as first sentence is simply not true. It does process a noise, unintentionally. Internal and external. A result is not ideal shape or timing.

A second one is not expressed clearly enough to bring for consideration. A third one bring a word 'correlated' in a different meaning and seems contradicting with the first sentence. I used a standard engineering term brought from theory of signal processing.

I think you are confusing what I’m saying.
The clock sends a signal to the dac. It does not receive and process a signal from the dac. It’s a one way street.

Here is the simplest way I can say this:

An external clock is just a distortion inducing device and many people seem to like this coloration and call it improvement. The issue is that this coloration isn’t predictable like an EQ.
It varies and changes.
All evidence ( FFT distortion plots) have always proved only that.

Personally, I prefer the D/A conversion to be transparent. I do not want my amplifier to do EQ, I do not want my speakers to do compression. I do not want my room nullifying or boosting certain frequencies. I do not want my D/A conversion to yield an "edge" that was not there.

 

[Clivem2]

Now that sounds like a excellent Mod Rod
And one l will add to my ever growing list. l gave put copper shields on chips, but never considered a damping

Is he Rod the Mod or Mod the Rod?

 

[sajunky]

I think you are confusing what I’m saying.
The clock sends a signal to the dac. It does not receive and process a signal from the dac. It’s a one way street.

Here is the simplest way I can say this:

An external clock is just a distortion inducing device and many people seem to like this coloration and call it improvement. The issue is that this coloration isn’t predictable like an EQ.
It varies and changes.
All evidence ( FFT distortion plots) have always proved only that.

Clear, no objection to this part. However you talk exclusively about adding environmental (uncorrelated) noise to the sound you hear. It cause harmonic distortions, responsive for colouration. Intermodulation products have far more nasty effect and they are present. However it is one way, no feedback.

Good you said one way street, as it should be easier to understand what happen when you place a clock inside DAC. It is exposed to a noise which is dependant on the sound currently played. Jitter shape is modified accordingly to the sound. Then a modified such way clock signal is used for D/A conversion. A noise from this conversion enter a clock again, this is not one way street anymore.

A different issue is a problem explained in a comment you were quoting, but now it is clear you do not understand. It is a sound dependent jitter, as I wrote before, 'correlated' to the sound, changing properties of the sound during D/A conversion in the way our hearing is more sensitive than the environmental noise in a clock. It is why moving the clock outside is beneficial.

For explanation of a difference I brought an example from early CD production. This was an example that a bigger amount of noise means very little when correlation to the original sound is removed.

A environmental noise is not equal to a different type of noise.