[Siz.]

Some recommend hooking the speaker level input of the sub directly at the speaker terminals as it takes into account any effect the speaker cables have, sounds like overkill to me but the theory at least makes sense ....

I can see why connectin to the speaker terminals might be convenient from a purely practical viewpoit. But I share your skepticism about audio benefits. To the extent that the speaker cables have an effect, surely this is a negative one, from picking up and transmitting random RF interference? If so, then I fail to see any possible benefit from connecting the sub to the speakers in order that it can share the same distortions!

 

[jarnopp]

Hi @jarnopp! Can you please elaborate on this? What was the issue with your converter and how did you fix it? I have created an adapter, tested it, making sure there are no shorts and it's wired for balanced drive using only pins 2 and 3. My TT2 is also at the service center at the moment because of an issue with the XLR output generating noise/clicks/pops in the right channel. I don't know what could have caused that to happen. As far as I can tell, the adapter I've made is fine.

Pins 1 and 2 were shorted together, so all 3 pins were connected. I gather this is valid for some equipment, but not standard and not appropriate for TT2.

 

[jarnopp]

This sounds like a more conventional setup, where the sub connection is made before the power amp stage. Rel are somewhat unconventional in that they like to take the signal after it has gone through your amp (so speaker level rather than line level). Nice to hear that it sounds great!

I think that I might have spooked Rel by referring to wiring up to the TT2's RCA outputs - it is perhaps understandable if they strongly associate RCA with line level ouputs, with speaker posts expected for the high level output.


Yup.

Can you say objectively lol, if your setup of tt2 and omega, is/was the right direction of travel, or if someone was to follow you, what would you suggest. I’m not sure if there’s a forum for omega speaker people...

The Omega’s are very transparent and I’m using them near field at my desk, where I also have my headphones, so switch back and forth between the two, using mScaler and TT2. It’s right for me now, especially this past year.

if I were to move or have a different need, like to fill a large room, I would likely be tempted to use multi-way speakers that are active or passive with an external amp, and be less concerned with transparency. Probably would use Qutest and no mScaler in that case, to keep costs reasonable. But each of us has to make that decision based on what we are looking for. Right now I am after maximum transparency and accuracy. (At a price point still, so don’t go asking me about getting a Dave! )

 

[AndrewOld]

Oh dear - I hope you are feeling better.


Let us start at the outset - there are always some technical reasons why things sound the way they do. That is assuming that sound quality has been characterised accurately and properly of course - I tend to ignore comments when someone says it sounds better without fully describing the differences.

On your first point with Dave and different CD players sounding different via optical it's quite easy to explain that - the CD players are mains powered, so will inject RF noise into the mains - and your system has quite a few possibilities for intersecting mains/ground loops, with CD players, Dave and the ATC speakers and/or power amps, so it's not surprising you can hear differences with different CD players. I do not think it is Dave that's causing the issue, but the CD players RF noise being picked up by the amps. I have never heard of any other audio designer talking about the importance of RF noise and how that can change the sound quality. Putting RF filters is absolutely crucial to SQ - and just a simple inductor and capacitor won't be sufficient. Eliminating the SQ problems from RF from the mains is extraordinarily complex.

Onto my lap-top - that was using USB as optical sounded identical when in battery mode (my old MSI lap-top), with nothing connected to the lap-top except for the portable hard disk.

As to the iPad - I have zero experience of using Apple myself. You say that the iPad sounded worse than your streamer - but how is the data getting to the iPad? Here we again have the possibility that increased RF noise on the mains is upsetting SQ. It would be interesting to play a file stored on the iPad, with Wi-Fi turned off and no ground or mains connection, going into Dave.

There are always technical reasons for why things sound different (when SQ tests are performed carefully, rigorously and properly characterised). In the case of bit perfect sources these are jitter levels (not an issue with my DACs) or RF noise (random and signal correlated) and audio bandwidth signal correlated noise. The RF and signal correlated noise needs to be at extraordinary low levels to not to be audible (so far, my evidence is that we are talking about below -350dB). I have never encountered a situation that has not been explained by these effects - that said it is difficult to imagine any other aberrations that could modify SQ that are source related!

Isn’t the simplest explanation of @Triode User s experience that he is using a modified DAVE into which he has put a different power supply, linear I think, and it is this power supply which has insufficient RF filtering?

 

[HeeBroG]

At the risk of going off-topic, I would be interested to know why there is a difference between the XLRs (used in single ended mode) and the RCAs, on the Dave or the TT2.

Not sure if it applies to TT2 but see post #6224 in DAVE thread.

 

[mammal]

#6224

For convienence link is here

 

[Rob Watts]

At the risk of going off-topic, I would be interested to know why there is a difference between the XLRs (used in single ended mode) and the RCAs, on the Dave or the TT2.

TT2 and Dave are different. On TT2 the XLR OP's are wired directly to high output current OP stages, XLR pin 2 is the same as RCA phono, but Dave is via 33 ohm resistors. And with Dave the negative output is not high current, so it will not drive headphones. The intent with Dave was to drive amplifiers only with the XLRs.

 

[ra990]

TT2 and Dave are different. On TT2 the XLR OP's are wired directly to high output current OP stages, XLR pin 2 is the same as RCA phono, but Dave is via 33 ohm resistors. And with Dave the negative output is not high current, so it will not drive headphones. The intent with Dave was to drive amplifiers only with the XLRs.

Mr. Watts, I have built a 4-pin XLR adapter by wiring pins 2 and 3 from the TT2's XLR outs to the appropriate pins on the female XLR4. It works fine, but there is no ground connection to the TT2 and I'm wondering if I need it. Should my adapter wire the female XLR4 adapter's ground connection to both the female XLR3 connector's ground or is that not necessary for balanced? Could you clarify this for me?

 

[Rob Watts]

If you are using headphones, then that is OK. If it's connecting to a power amp then you need pin 1 ground connected.

 

[hmartin]

Hi @Rob Watts!
Have you seen the new Linn Organik DACs? The design looks to have alot in common with your designs, but hard to tell as they are sparse on the details.

 

[ecwl]

Hi @Rob Watts!
Have you seen the new Linn Organik DACs? The design looks to have alot in common with your designs, but hard to tell as they are sparse on the details.

Hmm... I don’t know. Even though they are sparse on details, just looking at their video, it looks like they’re using an FPGA to upsample to some sort of DSD (not clear DSD128/256/512) and then putting it through a 32-element SDM DAC with shift register? Also not clear how high the tap length and what type of filter they’re using to do the oversampling.

That’s very different to me than Chord DACs where you have a high tap length WTA filter that upsamples to 16fs (705.6/756kHz) first and then upsamples to 5-bit 104MHz before feeding it through the noise shaper and the 4 to 20-element Pulse Array DAC.

However, you may be somewhat right. Mola Mola DAC is also 32-element SDM DAC with shift register but it upsamples to 5-bit 3-MHz and then converts to 32 different pulse width so that it should have a similar jitter immunity effect and lower or no noise floor modulation like Chord DACs. So it is possible Linn also implemented this DAC architecture. But even then, I believe Mola Mola can switch from say a 5-bit word of 5 to 10 suddenly whereas the noise shaper of Chord would always switch from 5 to 6 to 7 to 8 to 9 to 10 so that would still significantly reduce noise floor modulation.

That said, let’s imagine Mola Mola and Linn did implement a DAC architecture similar to Chord. There is no free lunch. If Mola Mola & Linn uses 32-elements and Chord only uses 10 to 20-elements and the FPGA’s have limited computing power (given the same power noise characteristics), that means that Mola Mola & Linn has to do less upsampling (which Mola Mola does), do less noise shaping and use shorter tap lengths just to accommodate more elements. At least that would be my assumption.

 

[AndrewOld]

Hmm... I don’t know. Even though they are sparse on details, just looking at their video, it looks like they’re using an FPGA to upsample to some sort of DSD (not clear DSD128/256/512) and then putting it through a 32-element SDM DAC with shift register? Also not clear how high the tap length and what type of filter they’re using to do the oversampling.

That’s very different to me than Chord DACs where you have a high tap length WTA filter that upsamples to 16fs (705.6/756kHz) first and then upsamples to 5-bit 104MHz before feeding it through the noise shaper and the 4 to 20-element Pulse Array DAC.

However, you may be somewhat right. Mola Mola DAC is also 32-element SDM DAC with shift register but it upsamples to 5-bit 3-MHz and then converts to 32 different pulse width so that it should have a similar jitter immunity effect and lower or no noise floor modulation like Chord DACs. So it is possible Linn also implemented this DAC architecture. But even then, I believe Mola Mola can switch from say a 5-bit word of 5 to 10 suddenly whereas the noise shaper of Chord would always switch from 5 to 6 to 7 to 8 to 9 to 10 so that would still significantly reduce noise floor modulation.

That said, let’s imagine Mola Mola and Linn did implement a DAC architecture similar to Chord. There is no free lunch. If Mola Mola & Linn uses 32-elements and Chord only uses 10 to 20-elements and the FPGA’s have limited computing power (given the same power noise characteristics), that means that Mola Mola & Linn has to do less upsampling (which Mola Mola does), do less noise shaping and use shorter tap lengths just to accommodate more elements. At least that would be my assumption.

The Linn Organik is a very interesting product, also very expensive, but it does have streaming (including Roon) and speaker/room correction capability built in. And optical ethernet, HDMI, and plenty more.

 

[hmartin]

The Linn Organik is a very interesting product, also very expensive, but it does have streaming (including Roon) and speaker/room correction capability built in. And optical ethernet, HDMI, and plenty more.

Yes, have been looking at linn due to these reasons. Seems perfect for a tv-room system. Assuming they release a less costly version that still seems good. I have a hard time thinking it can compete with msc+dave even if is much more expensive.

 

[AndrewOld]

I have a hard time thinking it can compete with msc+dave even if is much more expensive.

Absolutely no reason why it shouldn't compete.

 

[jacc]

I am just wondering if this product can effectively break the ground loop of DAVE and reducing the need for multiple ferrite?

https://ifi-audio.com/products/gnd-defender/