HOLO Audio MAY DAC

[Extreme_Boky]

That is not a good practice.

 

Connecting the USB cable connectors' metal casings together can cause ground loops IF the USB card does not have ground isolation. Highly undesired scenario. 

 

Also, no meter how the cable is constructed, the source-end PCB ground layer and the USB card PCB ground layer will connect those points together. 

 

 

[Rune]

5 hours ago, Extreme_Boky said:

My whole point of this discussion is that for that minute benefit of 1.536MHz PCM / 1024DSD with a non-Intel USB controller IC, we are paying too much of a price. This should particularly be of interest to those of us who do use Intel USB controllers and simply do not care about other controllers and possible incompatibility at the very highest of the data rates. 

I think these are valid point before considering upgrading. I have considered the following

1. Does the new card have better SQ when "only" doing DSD256/DSD512?
2. Does 1.536MHz PCM / 1024DSD sound/measure better than DSD256/DSD512. 

If the answer is no to both then there is no point of going through the hassle of upgrading even if it is free.

 1.536MHz PCM a pointless obsession of higher is better, maybe a bit like games running at over 144fps?

 

[Roasty]

I've not yet tried 1.536 with the new usb module + my AMD based pc. 

 

So far, 44.1, 384, 768 and dsd128, 256 all sound good to me. Nothing immediately different is apparent and nothing glaringly wrong is audible. But it's all based on "memory" of the old module, and I guess realistically, the only way I would confidently be able to say if there was a difference (if there was any in the first place) is if i had two dacs side by side with old/new usb module on the same rig. 

[barrows]

8 hours ago, Extreme_Boky said:

My whole point of this discussion is that for that minute benefit of 1.536MHz PCM / 1024DSD with a non-Intel USB controller IC, we are paying too much of a price. This should particularly be of interest to those of us who do use Intel USB controllers and simply do not care about other controllers and possible incompatibility at the very highest of the data rates. 

To be clear, these are not established "facts"...  these are speculative assumptions on your part.  Without accurate measurements of the actual noise levels on the ground plane in question, you cannot "know" if this USB interface is "better", or "worse" than the previous one.  Given the isolation present, it also may be that no significant amount of processor noise is going to get to the isolated side of the DAC.  Again, without actually measuring such, the rest is speculation.

No matter what, a USB receiver board is always going to have significant high frequency noise levels.  Excellent PCB design, excellent decoupling design, and isolation are all going to contribute to how much, or how little, this noise can affect actual DAC performance-looking at pictures and speculating is only that.  Holo/Kitsune claim improved sonic performance with the new interface-it is possible that they actually know what they are talking about.  It would be unusual for any manufacturer to release an improved USB interface which actual performs worse than its predecessor.

[MrDerrick]

I have ordered mine and am anxiously awaiting for it to arrive!

[Franzschubert]

4 hours ago, MrDerrick said:

I have ordered mine and am anxiously awaiting for it to arrive!

Congratulations .....  how long will it take?

[MrDerrick]

Franzschubert,

 

Hopefully this Friday or Saturday, West coast to East coast here in the States.

I am not holding my breath.

[Franzschubert]

That's fast.  I was told 4-5 weeks just to build the unit.  

[musicjunkie917]

14 minutes ago, Franzschubert said:

That's fast.  I was told 4-5 weeks just to build the unit.  

 

I think he is talking about the USB card......not the DAC.

[Franzschubert]

oops thanks

[MrDerrick]

Yes, the USB card.

[camott]

Anyone want to take a stab at comparing the Holo USB card to the JL Sounds card that some other brands like Lampizator use? http://jlsounds.com/i2soverusb.html. It's a $70 card that claims to have galvanic isolation and 2 power supply leads for USB and reclockers ... (including option for no upstream USB power by cutting a jumper). Given the pervasive opinion that the Holo USB implementation is as good as it gets, I am curious as to how/why it would be better than the JL Sounds card.

[Extreme_Boky]

Did you check the May's USB card to see how many of those supply leads, as you call them, Holo decided to implement?

[camott]

33 minutes ago, Extreme_Boky said:

Did you check the May's USB card to see how many of those supply leads, as you call them, Holo decided to implement?

 

@Extreme_Boky Sorry no I have not looked in depth at the Holo May's USB card myself. I have no idea how it handles power supply to it's card, though as I had asked last week, it appears that it *does* need USB 5v power. 

 

I guess I was just curious if other, more expert, folks like yourself had any thoughts on potential differences between the two. I wasn't suggesting one was better than the other, just curious. 

[AudioDoctor]

this is hilarious.

[elcorso]

8 minutes ago, AudioDoctor said:

this is hilarious.

Completely !!! 🤪

Roch

[AudioDoctor]

By the way, I am referring specifically to camotts comments here in this thread which did not come out of the blue, but were prompted by this discussion where he seemed pretty confident of his assertions which he asked about here.

 

 

[Extreme_Boky]

Tough crowd...

[camott]

@AudioDoctor My comment in that thread was not initially about different USB cards, it was specifically about whether or not people have heard SQ differences with different NAA devices upstream of the Holo May DAC's USB. You seem to think that the NAA endpoint can easily be heard even with a well engineered, isolated USB interface. Others don't.

 

I am just trying to engage and learn. I don't see why you need to be so condescending. But whatever. Peace out.

[GoldenOne]

On 11/29/2021 at 7:04 AM, Extreme_Boky said:

The twisted pair is for differential data transmission, so we have two wires here: USB D+ and USB D- data.

 

The third wire is +5 V DC used to power up the USB card. Its return is, in fact, the shield, which is the fourth wire. Sometimes there is a separate fourth wire which may appear as a separate wire to the shield - but it is not... the fourth wire (sometimes called the drain wire) is in direct contact with the shield.

 

So the shield is used as the RF interference noise suppression for data differential pair and as a common return for the USB +5V DC supply.

 

You also have to gain a good understanding of noise decoupling. The noise present on the USB card ground fill can not magically disappear, even if the Vcc and the Vee IC pins are perfectly decoupled. That PCB ground fill noise needs to be connected (AC or DC coupled) to something that will actually drain the PCB layer (ground fill) nose somewhere. It CAN not be the close-by May ground... because that is the whole point of the exercise - to prevent ground coupling between the source ground (PC/laptop/streamer) and May's motherboard ground layer (ground fill).

 

So, after all this discussion, I think it is more than obvious that the switching noise (the frequency and the switching amplitudes - the lower the Vcc and Vee - the better) should be kept to an absolute minimum on this small & isolated ground fill, as found on the USB card. 

May has sophisticated PLL mechanisms (4 crystal oscillators, two of which are indeed the lowest phase noise oscillators currently available - this combination of 2 standard and 2 high-quality oscillators is probably by far the best PLL implementation currently used in the audio industry) to reduce jitter, but the noise at the input is certainly not desirable. 

 

My whole point of this discussion is that for that minute benefit of 1.536MHz PCM / 1024DSD with a non-Intel USB controller IC, we are paying too much of a price. This should particularly be of interest to those of us who do use Intel USB controllers and simply do not care about other controllers and possible incompatibility at the very highest of the data rates. 

 

Also, I am just throwing a few facts here, so that we can hopefully obtain an official response from Holo distributors. They stated that the magic has happened and the new USB card sounds much better than the old one. I need to see the photos first; I reserve the right to question things/statements. More than anything, I want to see what the new card looks like, before I decide to upgrade.

 

Of course, I am not forcing anything on anyone. So, if you want to believe what you heard and know a priori that the new card sounds much better - knock yourself out and grab & install the new card.

 

 

 

I don't think holo ever said the new usb card sounds better. 

 

In fact Tim has explicitly said that the card revision is to address functional aspects in particular with non-in tel systems. And that unless you use hqplayer and have an Amd pc, the new card would make no difference to you

[Extreme_Boky]

 

Yes, that was my conclusion as well after looking at the one I have, and the partial photo of the new USB card.

But... I'd still love to see a whole photo of a new USB card, top and bottom. So, if any of you fine fellas can do that... 😁

 

 

[Extreme_Boky]

32 minutes ago, camott said:

@AudioDoctor My comment in that thread was not initially about different USB cards, it was specifically about whether or not people have heard SQ differences with different NAA devices upstream of the Holo May DAC's USB. You seem to think that the NAA endpoint can easily be heard even with a well engineered, isolated USB interface. Others don't.

 

I am just trying to engage and learn. I don't see why you need to be so condescending. But whatever. Peace out.

 

The May USB card has, what Holo seems to love and has implemented widely inside the whole DAC, a very low noise audio OP used as a low noise reference, for the two regulating series' transistors connected in parallel. This is a pre-regulation, straight after the +5V dc that enters the USB card via USB cable.

 

What follows is also quite remarkable... the USB transceiver IC has 3 separate linear regulators, the XMOS chip has 4, the Xtal oscillator has its own as well, the I2S isolator ICs also get their own regulator (for the Vdd1); Vdd2 comes from the May's mainboard.

 

So, it is leaps ahead of anything I have seen and used.

 

It's a little beauty, as is the rest of the DAC.

[AudioDoctor]

nevermind... not worth it.

[AudioDoctor]

4 hours ago, camott said:

@AudioDoctor My comment in that thread was not initially about different USB cards, it was specifically about whether or not people have heard SQ differences with different NAA devices upstream of the Holo May DAC's USB. You seem to think that the NAA endpoint can easily be heard even with a well engineered, isolated USB interface. Others don't.

 

I am just trying to engage and learn. I don't see why you need to be so condescending. But whatever. Peace out.

 

Maybe go read it again.

[musicjunkie917]

4 hours ago, GoldenOne said:

I don't think holo ever said the new usb card sounds better. 

 

Tim from Kitsune said this:

 

"Better sound quality and far better support for dsd1024 and pcm1.536mhz."

 

The first three words of that quote would suggest the new card sounds better. Of course, that is not anyone from Holo Audio in China...but, Tim does represent them.