MDAC First Listen (Part 00101001)

[Ciunas Audio]

No not at all, but the PLL derived clock (also consider its been transmitted via optical cable and the phase noise entails) is only used to "injection lock" the Dotox's discrete clock circuit.

So while the Detox's Clock circuit (even when injection locked) is very low phase noise - the USB HUB ic's internal Clock tree will be anything but. USB Hubs are not designed with ultra low Phase noise as a prime consideration.

With this in mind I'm designing the spread spectrum option mode.

Ah, this USB PLL clocking is in the proposed new design - I thought it was currently in your MDac.

The spread spectrum is an interesting idea

 

[MPerkins]

Can anybody tell me if the bulging caps members are talking about on the MDAC's are limited to a certain run or if all are affected? Wondering if there is a serial number range for the affected units?

Cheers
Mike

 

[fusion5]

Can anybody tell me if the bulging caps members are talking about on the MDAC's are limited to a certain run or if all are affected? Wondering if there is a serial number range for the affected units?

Cheers
Mike

Have a look here: http://www.pinkfishmedia.net/forum/showthread.php?p=2673281#post2673281

 

[JohnW]

Can anybody tell me if the bulging caps members are talking about on the MDAC's are limited to a certain run or if all are affected? Wondering if there is a serial number range for the affected units?

Cheers
Mike

Mike,

Fusion has linked you to my reply to the same question you asked earlier - I guess you missed it with the recent heavy thread traffic.

"As I no longer work with IAG / Audiolab I don't have access to the data - but I suspect some batches are worst then others as I have a CDQ (that uses the same caps and operates much hotter then the MDAC) and its been powered for almost 5 years without bulging caps...

But this is only circumstantial evidence."

Hard to say much more as I've only seen a few "early" units this past year or so and only 3 units sent to me with bad Caps.

 

[Rosewind]

I use an externally powered Paul Pang usb card in my music pc, so I would think that my usb out has no need for further detox work. For that reason I have not ordered a unit. Will I regret this?

Any thoughts on this?

Kind regards,
Peter

 

[iansr]

John
Sorry to be a pest but can you help me with these questions:

John
Active crossover related question; What do you anticipate will be the voltage gain of the VFET amps, 26db?

Also did you miss my question at post no 385?

 

[fusion5]

I use an externally powered Paul Pang usb card in my music pc, so I would think that my usb out has no need for further detox work. For that reason I have not ordered a unit. Will I regret this?

Peter,

The PPA USB cards use two different voltages. 5V (clock module, external USB device power) supplied via the internal Molex type connector and 3.3V (USB chip, ...) supplied via the PCIe slot. Unless you are using a modified PCIe riser card to supply both voltages from external sources, your card is only partly powered externally.

The general observation seems to be that the more optimized your audio pc is, the less improvements are gain from devices like the Regen.

My audio pc is powered by a HDPlex linear power supply which also supplies power for the SOtM USB card. And I will definitely try out John's Detox solution.

Michael

 

[jirij]

(to lighten the mood)

Yes, the Detox USB input is a standard USB type B connector.

Yes!

(hifi version)

 

[JohnW]

Any thoughts on this?

Kind regards,
Peter

Peter,

I only know the "Paul Pang" card from pictures - but from an engineering standpoint I don't see how this will have low RF or even decent Phase noise.

Without metal shielding and filtering the computer is a really Hot RF environment - its doesn't matter how good a USB card is in isolation (and I see nothing from the pictures to say that there is any sound engineering behind this card - but bare in mind I've not seen or measured one in real life).

If your happy with the sound then there's nothing to regret

At first glance, do I believe the card is soundly engineered? Sadly no

Are they any published performance results of the card?

Wait until the Detox design is completed then you can compare for yourself

 

[JohnW]

John
Sorry to be a pest but can you help me with these questions:

Ian,

Very sorry I missed your earlier post.

The VFET amplifier will have High / Low Gain mode automatically selected via the FDAC for highest system dynamic range.

High Gain is 26dB Balanced

Low Gain 6dB Blanced

 

[rtrt]

Hi John a little late so no time to check details.

One of the things about minidsp that concerned me during some reading way back, was that their levels (output I think, though it may be input too) appeared to be quite low.

From memory balanced was 2V and unbalanced was 1V which is half what i'd expected.

Are there any concerns for you there, and will those kinds of levels be a problem for the FDAC?

 

[JohnW]

rtrt,

MiniDSP expertises is in the DSP software - the Full scale output level is determined by the Analogue hardware - the FDAC hardware is very different to the mniDSP designs our FS output level is 3V SE & 6V balanced to allow for digital Pre-amplifier "Gain" headroom.

 

[hluga]

JohnW,

Just for curiosity what part of FDAC, VFET are you working on now ?

Does the prototype analog PCB meet your expectations ?

What are the next milestones ?

Best regards,
Hervé

 

[BE718]

Hi John

Remember a few weeks back when you were in China I brought up a possible problem with my MDAC? You couldnt see the plots on your phone. I have taken some further measurements for you to look at the skirts on the signals.

I did look inside and didnt see any faulty caps as you you suggested there may be.

The set-up is xlr out into a TI PCM4222evm board outputting 24bit 96kHz. Applied sig at -60 dB. I saw this originally on a Bruel & Kjaer Pulse system (first measurements I posted), so I think the measurement system can be ruled out as an issue. I have also generated the signal in two different ways, one with Adobe Audition and the other with "Test Tone Generator", at 24 bit 96kHz. Played back on Roon and JRiver software. Problem remained with whichever combination of signal and playback software.

I have also measured my Tag processor with the same signals. The noise floor of the TAG is around 6dB higher, but there are no skirts.

Opinions?

ADC noise floor and 1kHz @-60 dB. MDAC Sharp filter

 

[iansr]

Ian,

Very sorry I missed your earlier post.

The VFET amplifier will have High / Low Gain mode automatically selected via the FDAC for highest system dynamic range.

High Gain is 26dB Balanced

Low Gain 6dB Blanced

In an active set up I'll be using an additional 2 amplifiers and I need the gain to be the same (26db) for all of them, so will it be possible to have 2 selectable settings on the VFET, auto or fixed at 26db?

 

[iansr]

My other question was this:

I've ordered one slave unit. However I might want to have a 3 way active crossover configuration, so I'd need a second slave unit. But one of them would only be for bass frequencies and I think that would be over the top - I'd prefer a cheaper solution like this:
http://www.minidsp.com/products/minidsp-in-a-box/minidsp-balanced-2x4

If I got one of these instead of a second slave unit would I be able to daisy chain it in the same way and could I control its output volume in the same way from the master unit? In short, could I just use it as a drop-in substitute for a slave unit?

 

[JohnW]

JohnW,

Just for curiosity what part of FDAC, VFET are you working on now ?

Does the prototype analog PCB meet your expectations ?

What are the next milestones ?

Best regards,
Hervé

Hluga,

I'm spending most my time working to complete the FDAC Analogue PCB - updating the design and reformatting it for the FWC (I'm now able to remove various limitations imposed by the original small PCB area).

Last night I also spent a few hours working on the Detox Clock circuit.

I'm also working in the VFET amplifier Chassis design with the new vendor.

The next major milestone is to issue the FDAC Analogue PCB for T0 Production - also the Detox PCB.

During this time I'll also be arranging the prototype VFET and FDAC Chassis - and I also need to think about the Detox Chassis once I have a better idea about the PCB size / format.

 

[JohnW]

In an active set up I'll be using an additional 2 amplifiers and I need the gain to be the same (26db) for all of them, so will it be possible to have 2 selectable settings on the VFET, auto or fixed at 26db?

Yes, we can arrange Auto, or selectable 6dB / 26dB.

 

[JohnW]

My other question was this:-

"I've ordered one slave unit. However I might want to have a 3 way active crossover configuration, so I'd need a second slave unit. But one of them would only be for bass frequencies and I think that would be over the top - I'd prefer a cheaper solution like this:
http://www.minidsp.com/products/mini...p-balanced-2x4

If I got one of these instead of a second slave unit would I be able to daisy chain it in the same way and could I control its output volume in the same way from the master unit? In short, could I just use it as a drop-in substitute for a slave unit?"

not only do you face the issue level matching (Volume control) between the units but you also need to consider time alignment (Phase) between DAC's one of the purposes of the Master/Slave interface is to insure all channels are synchronised in the time domain.

If you wanted to try, then you would not need a DAC / Unit with DSP as Master FDAC would perform the DSP function and could output via SPDIF - but I'm uneasy for the reasons mentioned....

 

[JohnW]

Hi John

Remember a few weeks back when you were in China I brought up a possible problem with my MDAC? You couldnt see the plots on your phone. I have taken some further measurements for you to look at the skirts on the signals.

I did look inside and didnt see any faulty caps as you you suggested there may be.

The set-up is xlr out into a TI PCM4222evm board outputting 24bit 96kHz. Applied sig at -60 dB. I saw this originally on a Bruel & Kjaer Pulse system (first measurements I posted), so I think the measurement system can be ruled out as an issue. I have also generated the signal in two different ways, one with Adobe Audition and the other with "Test Tone Generator", at 24 bit 96kHz. Played back on Roon and JRiver software. Problem remained with whichever combination of signal and playback software.

I have also measured my Tag processor with the same signals. The noise floor of the TAG is around 6dB higher, but there are no skirts.

Opinions?

ADC noise floor and 1kHz @-60 dB. MDAC Sharp filter

BE718,

I cannot really comment on your unit without performing my own test - also the skirt you are concerned about is very low in level - its not something I've ever looked into or noticed - so I'm not sure if its an issue or not.

As I offered before the best solution is for you to send me your unit so that I can perform my own tests.