[flyte3333]

I recommend ferrites on DAVE's mains cable as DAVE is definitely not immune to RF noise on the mains.

Agreed. Ferrites are a fantastic treatment but not the cure I found after experimenting a lot with Hugo2.

The ideal solution seems to be blocking these naughty leakage currents, but with a USB source that is PERMANENTLY mains powered, it’s easier said than done.

We just have to trust our ears. And for you the 40 ferrites does the job and for me Toslink does the same job (audibly). There’s no right or wrong. It just comes down to finding the best PRACTICAL solution for your own system. As an example , my Frankenstein microrendu chain (battery powered and battery powered optically isolated ethernet) - fully blocks leakage current but a pain in the a$$ to maintain, recharging batteries etc.

If you have a spare USD1200 there is a company that optically isolates high speed USB2.0 but it still needs 5Vdc external power...

 

[marcmccalmont]

RF noise is a problem for me (solved with dozens of ferrites), due to my mains-powered PC as USB source and I only have headphones connected to DAVE.

I recommend ferrites on DAVE's mains cable as DAVE is definitely not immune to RF noise on the mains.

Now playing: Skott - Wolf

I ended up with 18 topnisus ferrites on my power cord into Dave, smoother, sweeter and quieter

 

[Jawed]

Agreed. Ferrites are a fantastic treatment but not the cure I found after experimenting a lot with Hugo2.

The ideal solution seems to be blocking these naughty leakage currents, but with a USB source that is PERMANENTLY mains powered, it’s easier said than done.

We just have to trust our ears. And for you the 40 ferrites does the job and for me Toslink does the same job (audibly). There’s no right or wrong.

I used optical to check that the filtering on USB works, so that I can't find a difference between them

It just comes down to finding the best PRACTICAL solution for your own system. As an example , my Frankenstein microrendu chain (battery powered and battery powered optically isolated ethernet) - fully blocks leakage current but a pain in the a$$ to maintain, recharging batteries etc.

There are reports that ferrites on the power cable plugged into LPS-1 make a difference to the sound.

Check to find out if you can hear a difference depending on whether you have power going into LPS-1 or not. Just pull the power cable out of the connection on LPS-1.

If you have a spare USD1200 there is a company that optically isolates high speed USB2.0 and it needs only 5Vdc external power....

Lots of ferrites work.

It's not the current that causes the problem, it's the fact that the current is the means for noise to travel through the cable and spoil DAVE's sound. If you filter the noise with the equivalent of thousands of ohms using lots of ferrites (4-5000 is a good target if you can find a specification for the ferrites), then the noise doesn't cause a problem inside DAVE, it's simply too weak. The current still flows, but there's not enough noise for the current to spoil the sound.

Now playing: JFDR - White Sun

 

[flyte3333]

There are reports that ferrites on the power cable plugged into LPS-1 make a difference to the sound.

When I say my Frankenstein microRendu is battery powered, I don’t mean Uptone LPS-1 powered (which tries to approximate a really low impedance output battery but some leakage currents still get through). I mean really battery powered

Which power cable with the LPS-1? Do you mean the AC power cable or DC power cable?

 

[flyte3333]

It's not the current that causes the problem, it's the fact that the current is the means for noise to travel through the cable and spoil DAVE's sound.

We are getting deep into the weeds of very complex stuff and unknowns but according to John S it’s both.

“Yes, even with a headphone amp DAC you CAN have a leakage loops forming from the computer's and the DAC's power supplies. This can affect the SQ of the DAC, it can affect the local clock in the DAC, the DAC chips and the analog stage of the DAC.”

So it appears that things are more complex than just reducing the effects of leakage causing radiated RF in the USB cable. Ideally you still want to block these leakage currents getting inside the DAC. Ferrites very effectively help to filter RF outside the USB cable shield but the leakage can still get through inside the cable shield and affects things inside the DAC (per John S):

https://www.computeraudiophile.com/...-and-grounding/?do=findComment&comment=799479

“Another way leakage can get into systems is through a DAC, the leakage current can go through the ground plane of the DAC PCB, that current creates a small voltage which modulates the oscillators(s) producing the clocks in the DAC, adding jitter to those clocks. Even if the leakage doesn't get to a preamp or power amp it can add jitter to the clock in the DAC, thus subtly distorting audio output.”

https://www.computeraudiophile.com/...-and-grounding/?do=findComment&comment=735311

 

[flyte3333]

I used optical to check that the filtering on USB works, so that I can't find a difference between them

Absolutely and this was my point earlier. Despite all the complexities involved, if we can't hear differences then just pick whatever is most convenient.

For me, I took Rob's advice that a battery powered USB source is best - no leakage currents (don't need any ferrites) and the DAC's master clock controlling things means no jitter (and no DPLL engaged of course).

So I compared Toslink with a battery powered USB source and heard no significant difference - both are fatigue-free listening and both tease me to turn up the volume louder and louder.

But my own Toslink chain is the most practical for a permanent solution (for me) - no recharging batteries etc. And there's the feel good factor of no leakage currents getting into the DAC.

 

[elviscaprice]

I agree, getting a little too ferrite happy. Okay, I get the BLU to DAVE, but other than this, my question would be what is wrong in my system design that I would need such a fix as ferrites, which obviously isn't the fix all.
For DAVE a new power chord with heavier gauge wire and coming from an isolation transformer would be a better solution, more encompassing. USB cables eliminated for neutral adapter such as the USPCB. Keep the system as short pathed and isolated as possible. Every inch of extra chord and extra/oversized component create potential noise reintroduction and increased impedance.
I don't find batteries the solution either. Not a fan of their influence on sound signature or as some here would say lack of influence. Either way, I don't like them.

 

[Jawed]

When I say my Frankenstein microRendu is battery powered, I don’t mean Uptone LPS-1 powered (which tries to approximate a really low impedance output battery but some leakage currents still get through). I mean really battery powered

Nice. Some people describe LPS-1 as a battery power supply, but it still has a connection to the mains which causes a problem

Which power cable with the LPS-1? Do you mean the AC power cable or DC power cable?

DC from what I remember.

We are getting deep into the weeds of very complex stuff and unknowns but according to John S it’s both.

I think you might want to check with him. A current with no noise will have no effect.

“Yes, even with a headphone amp DAC you CAN have a leakage loops forming from the computer's and the DAC's power supplies. This can affect the SQ of the DAC, it can affect the local clock in the DAC, the DAC chips and the analog stage of the DAC.”

Also we're talking about DAVE and clocking/jitter simply isn't an issue. The RF that gets into DAVE is causing a problem in the analogue part of the DAC, the pulse array, due to noise in the ground plane.

So it appear more than just the effects of leakage causing radiated RF in the USB cable.

Yes, all along we've been talking about common mode noise at the electrical sockets (USB, AES or BNC) that causes the problem. Not radiated RF.

Ideally you still want to block these leakage currents getting inside the DAC.

They don't matter with DAVE as long as there's no RF noise in the current. Rob has already talked about how lower (audio) frequency noise is dealt with by DAVE so that only RF noise is a problem.

Ferrites very effectively help to filter RF outside the USB cable shield but the leakage can still get through inside the cable shield and affects things inside the DAC (per John S):

https://www.computeraudiophile.com/...-and-grounding/?do=findComment&comment=799479

“Another way leakage can get into systems is through a DAC, the leakage current can go through the ground plane of the DAC PCB, that current creates a small voltage which modulates the oscillators(s) producing the clocks in the DAC, adding jitter to those clocks. Even if the leakage doesn't get to a preamp or power amp it can add jitter to the clock in the DAC, thus subtly distorting audio output.”

https://www.computeraudiophile.com/...-and-grounding/?do=findComment&comment=735311

I'm going to repeat myself: the current isn't a problem. It's the RF noise carried by the current that's the problem. Lots of ferrites filter this noise. The current, on its own, has no detrimental effect. Swenson has never tested 10s of ferrites as far as I can tell. He belongs to the "noise free clocks" school, which may well be relevant to other DACs (though the evidence points to simple noise modulation in analogue circuits, a side effect of RF noise in the ground plane). Jitter isn't relevant to Rob's pulse array.

For me, I took Rob's advice that a battery powered USB source is best - no leakage currents (don't need any ferrites) and the DAC's master clock controlling things means no jitter (and no DPLL engaged of course).

Rob recommends optical. Battery powered USB is next best. The DPLL is irrelevant.

Now playing: Adrian Knight - What Light (Never Goes Dim)

 

[flyte3333]

Rob recommends optical. Battery powered USB is next best. The DPLL is irrelevant.

https://www.head-fi.org/threads/chord-electronics-dave.766517/page-649#post-13970809

and

"While 100% of RF is eliminated with optical, as Rob has indicated, because his SPDIF inputs have to go through a buffer and then a DPLL before it is synchronized to the FPGA clock, USB is DAVE's best input."

https://www.head-fi.org/threads/chord-electronics-dave.766517/page-382#post-13007750

So combining the above, battery powered USB source should have a slight edge. Maybe. Again, no DPLL engaged since the master clock controls things and no leakage currents, therefore no RF worries.

For anyone else reading this, this is just fun discussion. The music sounds phenomenal on all inputs

 

[Jawed]

Nice! You upgraded to DAVE. I got the feeling you were going to stick to Hugo/Hugo 2.

I agree, getting a little too ferrite happy.

They work. The USPCB you're using perfectly transmits all the RF noise that your spaghetti generates and feeds it into DAVE, spoiling the sound. A USB cable with 20 ferrites will be better as it'll filter the noise made by your spaghetti.

Okay, I get the BLU to DAVE, but other than this, my question would be what is wrong in my system design that I would need such a fix as ferrites, which obviously isn't the fix all.

Try an optical cable (and disconnect USB) into DAVE, if you have a source with optical output.

 

[Jawed]

So combining the above, battery powered USB source should have a slight edge. Maybe. Again, no DPLL engaged since the master clock controls things and no leakage currents, therefore no RF worries.

No. I think you've forgotten this post:

https://www.head-fi.org/threads/chord-electronics-dave.766517/page-648#post-13969654

Yes I vaguely recalled that - it does indeed sound slightly brighter and I put that down then to the benefit of USB in that it gets the clock from the FPGA. But I have changed my mind on this now, as Hugo 2 USB sounds the same as optical when using headphones (it's like Dave when driving a power amp). So actually that indicates that the timing issue is not actually a benefit, as the DPLL eliminates the source jitter.

DPLL is a red herring. There is no possible edge for USB versus optical.

In summary: Rob once thought USB had an edge (RF noise gave the impression of clearer detail), which lead to the "DPLL is inferior" theory. But he realised that RF noise had coloured his preference, invalidating the DPLL theory.

Now playing: Julien Baker - Distant Solar Systems

 

[doraymon]

If you filter the noise with the equivalent of thousands of ohms using lots of ferrites (4-5000 is a good target if you can find a specification for the ferrites), then the noise doesn't cause a problem inside DAVE, it's simply too weak.

The problem is that ferries provide impedance which varies significantly with frequency so from what I understand getting 4-5k Ohms impedance on the frequency range say 100MHz - 1GHz would require tons of ferrite cores, the cable would bend under the weight of the clamps, the USB or power connector of the DAVE would be subject to extremely high bending stress and will eventually collapse in a catastrophic failure!

Now Jokes aside, the problem is that one never really know what is the frequency of the disturbing signal entering DAVE so it's like wearing an armour to block a sword and be hit by the needle between the junction of two plates.
You can't just be prepared for everything.
You need either an EMI lab or a lot, lot of experimentation!

 

[flyte3333]

No. I think you've forgotten this post:

https://www.head-fi.org/threads/chord-electronics-dave.766517/page-648#post-13969654


DPLL is a red herring. There is no possible edge for USB versus optical.

In summary: Rob once thought USB had an edge (RF noise gave the impression of clearer detail), which lead to the "DPLL is inferior" theory. But he realised that RF noise had coloured his preference, invalidating the DPLL theory.

Now playing: Julien Baker - Distant Solar Systems

Noted but as Rob says in the link I provided, regarding these RF related issues with USB sources: "This gets eliminated when the lap-top is not connected to the mains"

Break the leakage current loop and the RF issues get eliminated and there's no need for ferrites on the USB cable.

As I've said before, completely blocking leakage currents from a permanently mains connected USB source is not easy to do. In which case ferrites can definitely help. But ferrites aren't required for a battery powered USB source (disconnected from mains power).

 

[Jawed]

The problem is that ferries provide impedance which varies significantly with frequency so from what I understand getting 4-5k Ohms impedance on the frequency range say 100MHz - 1GHz would require tons of ferrite cores, the cable would bend under the weight of the clamps, the USB or power connector of the DAVE would be subject to extremely high bending stress and will eventually collapse in a catastrophic failure!
[...]
You need either an EMI lab or a lot, lot of experimentation!

The desk, here, where DAVE sits is taking all the weight of the ferrites.

And as I keep saying, I used the optical cable to check that I have enough ferrites 20 wasn't enough on the USB cable, for what it's worth, but 20 is a good starting point for experiments to find out if ferrites are worth using, if you don't have an optical source to use with DAVE.

Now playing: Billie Marten - Lionhearted

 

[doraymon]

On a different note, I just received a message that my DAVE should be with me the coming weekend...
No need to explain you guys how I feel right?