iFi's Pro iDSD (official) - NEW Firmware - MQA and more.

[AMR/iFi audio]

Easter ‘Bug Bounty’ Hunt

Hunt 'em, wherever they are...

 

 

Introduction

Our just announced firmware v5.30 is MQA capable and in one fell swoop, elevates legacy iFi products* with MQA capability for even more sonic enjoyment.

 

The MQA integration was a little difficult. It involved:

• Totally replacing the customised core code
• Re-apply tuning/core loading and
• Further fine tuning the firmware for even more precise allocated resource use in order to allow us to support MQA and384kHz.

 

To bring all this to fruition required the combined efforts of the MQA and iFi software developers or a total of +1,000 programming hours to deliver firmware v5.30 (excluding testing on all iFi legacy units).

 

Notwithstanding, there may still be one or two software bugs we have not quashed – hence we would like to involveyou, the customer – in the Easter Bug Hunt.

 

More details are to be found in this thread:

[austinpop]

@AMR/iFi audio

 

Could you clarify a couple of things regarding the 10MHz reference clock input?

1. Will it accept both square and sine wave clocks?
2. What internal clock(s) does the reference clock discipline?

Thanks.

[Govnah]

Clarification, I have read that the Pro iDSD will have Apple AirPlay, what about Android?

 

Thanks 

[AMR/iFi audio]

On 23.04.2018 at 6:33 PM, austinpop said:

@AMR/iFi audio

 

Could you clarify a couple of things regarding the 10MHz reference clock input?

1. Will it accept both square and sine wave clocks?
2. What internal clock(s) does the reference clock discipline?

Thanks.

 

Any clock, sine or square with at least 0.5V PP will work.

 

[AMR/iFi audio]

On 23.04.2018 at 7:28 PM, Govnah said:

Clarification, I have read that the Pro iDSD will have Apple AirPlay, what about Android?

 

Thanks 

 

The iDSD Pro will support DLNA playback, Airplay. The playback and setup app is available for both iOS and Android, setup is also possible via web Interface. 

[austinpop]

1 hour ago, AMR/iFi audio said:

 

Any clock, sine or square with at least 0.5V PP will work.

 

 

Excellent, thanks for your answer. That means acclaimed clocks like the Mutec Ref 10, the SOtM sCLK-OCX10, and the Cybershaft should all work fine.

 

Could you address the second part of my question? Which internal clocks will the reference clock discipline? In other words, will an ultra-low-phase noise reference clock on the input improve the sample rate clocks, the clock used for the USB input, or both?

[austinpop]

On 4/24/2018 at 3:33 PM, AMR/iFi audio said:

 

Any clock, sine or square with at least 0.5V PP will work.

 

 

Could you also elaborate what impedance the BNC input expects - 50 ohm or 75 ohm?

 

Also, I found this view of the back panel on Music Direct's website:

I am actually very encouraged that it supports an external PSU. The label says 15V/4A. A high-quality, beefy 6A or 10A LPS will likely do wonders for this unit.

[AMR/iFi audio]

On 25.04.2018 at 12:13 AM, austinpop said:

Excellent, thanks for your answer. That means acclaimed clocks like the Mutec Ref 10, the SOtM sCLK-OCX10, and the Cybershaft should all work fine.

 

As a minimum, to improve on the internal 10MHz clock oscillator, a clock should be verified to offer the level of performance (especially phase noise) of the Stanford Research Systems PERF-10. 

• http://www.thinksrs.com/products/PERF10.htm

Clocks with lower performance will likely degrade performance.

 

On 25.04.2018 at 12:13 AM, austinpop said:

Could you address the second part of my question? Which internal clocks will the reference clock discipline? In other words, will an ultra-low-phase noise reference clock on the input improve the sample rate clocks, the clock used for the USB input, or both?

 

The iDSD Pro uses the GMT clock system originally developed for the AMR DP-777, using an adjustable clock system with 0.04ppm precision and ultra low phase noise. This means ALL audio clocks (USB, S/PDIF etc.) are derived from this 10MHz clock and phase noise in the 10MHz clock will transfer directly to the audio clocks.

 

On 26.04.2018 at 1:31 AM, austinpop said:

Could you also elaborate what impedance the BNC input expects - 50 ohm or 75 ohm?

 

As per all the relevant standards (AES-3id) and/or industry practices (e.g. see SRS PERF-10) the BNC connections are 75 Ohm.

[austinpop]

4 hours ago, AMR/iFi audio said:

 

As a minimum, to improve on the internal 10MHz clock oscillator, a clock should be verified to offer the level of performance (especially phase noise) of the Sanford Research Systems PERF-10. 

• http://www.thinksrs.com/products/PERF10.htm

Clocks with lower performance will likely degrade performance.

 

The days when rubidium clocks were considered the optimal for audio are long gone. Hopefully your engineers know this, and will tell you that rubidium (and cesium) clocks, while having outstanding long-term (resistance to drift) stability characteristics, do not have particularly good short-term stability (jitter or phase noise), which is the key to digital audio. This is reflected in the published phase noise characteristics. If you look at the published phase noise specs of the Perf 10 you referenced above, it is:

Phase noise (SSB)

<-130 dBc/Hz (10 Hz) <-140 dBc/Hz (100 Hz) <-150 dBc/Hz (1 kHz) <-155 dBc/Hz (10 kHz)

 

In contrast, the published phase noise of an OCXO-based  Mutec Ref 10 is:

Phase noise	dBc/Hz
1 Hz:	≤ -116
10 Hz:	≤ -145
100 Hz:	≤ -160
1000 Hz:	≤ -166
Noise floor:	≤ -170

 

I think the numbers speak for themselves. 

 

4 hours ago, AMR/iFi audio said:

 

The iDSD Pro uses the GMT clock system originally developed for the AMR DP-777, using an adjustable clock system with 0.04ppm precision and ultra low phase noise. This means ALL audio clocks (USB, S/PDIF etc.) are derived from this 10MHz clock and phase noise in the 10MHz clock will transfer directly to the audio clocks.

 

That is great news. Since I own an aforementioned Ref 10, it sounds like the low phase noise will benefit both audio and system clocks in the Pro iDSD. This makes me want to audition it. When are you releasing it again?!

 

4 hours ago, AMR/iFi audio said:

 

As per all the relevant standards (AES-3id) and/or industry practices (e.g. see SRS PERF-10) the BNC connections are 75 Ohm.

 

75 ohm is a standard choice, so that makes sense.

[lmitche]

9 hours ago, AMR/iFi audio said:

Sanford Research Systems PERF-10

It's Stanford Research, like Stanford University not Sanford Research, thanks.

[Superdad]

13 hours ago, lmitche said:

It's Stanford Research, like Stanford University not Sanford Research, thanks.

 

Yeah I don’t recall Redd Foxx selling Rubidium clocks in the Sanford & Son junk yard.  

[anroj]

On ‎4‎/‎26‎/‎2018 at 1:38 PM, austinpop said:

 

The days when rubidium clocks were considered the optimal for audio are long gone. Hopefully your engineers know this, and will tell you that rubidium (and cesium) clocks, while having outstanding long-term (resistance to drift) stability characteristics, do not have particularly good short-term stability (jitter or phase noise), which is the key to digital audio. This is reflected in the published phase noise characteristics. If you look at the published phase noise specs of the Perf 10 you referenced above, it is:

Phase noise (SSB)

<-130 dBc/Hz (10 Hz) <-140 dBc/Hz (100 Hz) <-150 dBc/Hz (1 kHz) <-155 dBc/Hz (10 kHz)

 

In contrast, the published phase noise of an OCXO-based  Mutec Ref 10 is:

Phase noise	dBc/Hz
1 Hz:	≤ -116
10 Hz:	≤ -145
100 Hz:	≤ -160
1000 Hz:	≤ -166
Noise floor:	≤ -170

 

I think the numbers speak for themselves. 

 

 

That is great news. Since I own an aforementioned Ref 10, it sounds like the low phase noise will benefit both audio and system clocks in the Pro iDSD. This makes me want to audition it. When are you releasing it again?!

 

 

75 ohm is a standard choice, so that makes sense.

 

[anroj]

On ‎4‎/‎26‎/‎2018 at 1:38 PM, austinpop said:

That is great news. Since I own an aforementioned Ref 10, it sounds like the low phase noise will benefit both audio and system clocks in the Pro iDSD. This makes me want to audition it. When are you releasing it again?!

Pardon my ignorance but I am going to purchase the iDSD Pro and was wondering about 10MHz clocks. Can I connect the Ref 10 directly into the iDSD Pro without needing something like the Mutec MC-3+ USB?

[austinpop]

 

46 minutes ago, anroj said:

Pardon my ignorance but I am going to purchase the iDSD Pro and was wondering about 10MHz clocks. Can I connect the Ref 10 directly into the iDSD Pro without needing something like the Mutec MC-3+ USB?

 

That is my understanding.

[d_elm]

1 hour ago, anroj said:

Pardon my ignorance but I am going to purchase the iDSD Pro and was wondering about 10MHz clocks. Can I connect the Ref 10 directly into the iDSD Pro without needing something like the Mutec MC-3+ USB?

What power supply will you use with the PRO iDSD ?  15V, 4A label on the back of the unit.

[austinpop]

10 minutes ago, d_elm said:

What power supply will you use with the PRO iDSD ?  15V, 4A label on the back of the unit.

 

Something like the Uptone JS-2 would have been perfect, but not sure if it can be customized to 15V output.

[d_elm]

41 minutes ago, austinpop said:

 

Something like the Uptone JS-2 would have been perfect, but not sure if it can be customized to 15V output.

I found Music Direct list the iFi PRO iDSD and the specs indicate 9V/6.7A to 18V/3.35A for input voltage so a JS-2 at 12V/5A is a good try.

[austinpop]

21 minutes ago, d_elm said:

I found Music Direct list the iFi PRO iDSD and the specs indicate 9V/6.7A to 18V/3.35A for input voltage so a JS-2 at 12V/5A is a good try.

 If that is the case, then awesome!

 

DACs live and die by the quality of their PSU. There is no substitute for a beefy PSU with a deep reservoir of instantaneous current. I consider the PSU to be THE most important component of a DAC.

[jamesg11]

2 hours ago, austinpop said:

 If that is the case, then awesome!

 

DACs live and die by the quality of their PSU. There is no substitute for a beefy PSU with a deep reservoir of instantaneous current. I consider the PSU to be THE most important component of a DAC.

Noted you’ve been firming that view up in recent times!

 

What are the specs that indicate the ’deepness’ of the “reservoir of instantaneous current”? Capacitance in microfarads? /& ...?

[austinpop]

51 minutes ago, jamesg11 said:

Noted you’ve been firming that view up in recent times!

 

What are the specs that indicate the ’deepness’ of the “reservoir of instantaneous current”? Capacitance in microfarads? /& ...?

 

At a basic level, it refers to the maximum sustained and peak current that the power supply can deliver at the rated voltage. In large part, this correlates with the size (VA rating) of the transformer. This is assuming all other measures of quality are also delivered:

• quality of the transformer
• quality of the regulation stage(s)
• extremely low output impedance
• etc.

[lmitche]

48 minutes ago, austinpop said:

 

At a basic level, it refers to the maximum sustained and peak current that the power supply can deliver at the rated voltage. In large part, this correlates with the size (VA rating) of the transformer. This is assuming all other measures of quality are also delivered:

• quality of the transformer
• quality of the regulation stage(s)
• extremely low output impedance
• etc.

This could be a perfect use for the 5 amp Stammheim lt3045 which maxs out at 15 volts. Finding a good transformer is straight forward. Takes some diy skills however.

 

No doubt the choke filtered js-2 will sound great as well.

[anroj]

10 hours ago, austinpop said:

 

 

That is my understanding.

Thank you. That definitely helps.

[Superdad]

10 hours ago, d_elm said:

I found Music Direct list the iFi PRO iDSD and the specs indicate 9V/6.7A to 18V/3.35A for input voltage so a JS-2 at 12V/5A is a good try.

 

The UpTone JS-2 can output 7.2 SMPS at 12V continuous all day long—with instantaneous peak capability well above that. 

 

I’m excited to read that the Pro iDSD has an input range that includes 12V (the JS-2 truly shines with a real audio component such as a DAC—just ask the over 70 Mytek Brooklyn DAC owners with JS-2 )

However, the fact that the current requirement of the Pro iDSD goes down as the voltage goes up does indicate the use of DC-DC switching regulators. Not necessarily a bad thing as the iFi folks are very good at engineering with those.

[look&listen]

37 minutes ago, Superdad said:

JS-2 can output 7.2 SMPS at 12V continuous

[Superdad]

4 hours ago, look&listen said:

 

5 hours ago, Superdad said:

The UpTone JS-2 can output 7.2 SMPS at 12V continuous all day long—with instantaneous peak capability well above that.

 

What the heck!?  Darn autocorrect on my phone.  Since when is “SMPS” more probable than “amps”?   (Really no need to restate that our big choke-filtered, dual-output unit is an all-linear power supply.)