Da-capo issue (John W?)

[JohnW]

Hi Tim,

A good question - even I've had to think about the answer.

If your main playback source is CD or USB @ 44.1 KHz or 88.2 KHz - then the CD8200 should be the best. However, the MDAC (and 8200CDQ) have a slightly higher performance DAC over the CD8200. So it’s hard to say which benefit is greatest.

The one thing I will say however is that the 8200CD is extremely good - far higher performance than the Dacapo – I’m sure you will be very happy – retails around GBP500 (the price might end up GBP50 or so extra as the BOM (Bill of Materials) is slightly over target).

The 8200CD also lacks the headphone output.

I took great care in the design of the CD servo section, so that it achieves very high levels of disk playability – significant effort want into insuring the servo section would not modulate the critical DAC section via the PSU and Ground plane etc.

John

[EUVL]

John,

Would you care to make a few comments about the use of ECL logic ?
You have mentioned in one of your previous posts that you were applying ECL logic in one of your designs.

Thanks,
Patrick

[indypants]

Originally Posted by JohnW

If your main playback source is CD or USB @ 44.1 KHz or 88.2 KHz - then the CD8200 should be the best. However, the MDAC (and 8200CDQ) have a slightly higher performance DAC over the CD8200. So it’s hard to say which benefit is greatest.John

Thanks for that John. A couple more questions I'm afraid...

1) Can you elaborate on the digital in/out connectivity on the 8200CD? Apart from CD, my sources are a Squeezbox (spdif/toslink) and an SACD/DVDA player (HDMI).
2) Does the 8200CD also have a configurable clock-out, as per the spec I saw for the MDAC?
3) Forgive my lack of technical knowledge, but does a usb source also require a separate clock connection, or is this handled by the USB?

Once again, thanks for any reply!
- Tim

[JohnW]

Hi Patrick,

Sorry - I did receive your PM around Christmas, but time was more hectic then most....

On occasions I have used ECL (PECL), however have found that CMOS families such as AC (watch the ringing), VHC and UHC (both from Fairchild) have superior Close-in Phase noise for Clock Distribution (at 5V).

As usual however, there’s a catch - ECL has better "Isolation" properties - now you don’t normally hear about "Isolation" in terms of Digital Logic. If you think about a Relay or Switch - they have certain characteristics, such as "Off" isolation. When we start to use Logic IC's to perform Analogue tasks, such as DAC elements (using D-Type FF) to relatch Data to Low Phase Noise clock for Integration into Audio, the "Latch's" Isolation properties start to become very important. I have presumed this is due to the balanced internal / external nature of ECL and the Class A operation of its switching, resulting in ECL's greater "Latching" (Read Isolation) of the Data.

So while ECL has poorer "Static" virtues for Clock distribution (higher Random Phase Noise), it has better Analogue Dynamic characteristics (Better Isolation between input and output Data in D-Types Latch’s) i.e. the more perfect “Latch / Isolator” behaviour.

ECL’s “More perfect” Dynamic performance (Note these are Analogue virtues of a Digital Logic Device), are clearly audioable – however, ECL’s Poorer Close-In Phase Noise (Which of course is lies directly within our audio BW) results in measurably poorer Audio Noise performance (with 16K FFT NF Ref. to 2VRMS at -145dB to -150dB, where CMOS achieves -165dB to -175dB with care).

Results for both logic types, can ONLY be achieved with “Perfect” PSU, decoupling, Ground Planes etc. Note, any logic will only have 6dB PSU rejection with a 50% clock when integrated (and that 6dB is only because of the 50% Clock Duty) – so in reality, Logic has NO PSU rejection once integrated.

Tim,

The CD8200 has:-

2x Coax SPDIF Inputs supported to 194 KHz 24 Bits

2x Optical SPDIF input guaranteed to support up to 96KHz only – in fact “should” work at 192KHz – but we have experience so many issues with 3rd parties source devices abilities to output 192KHz over optical – we cannot guarantee 192KHz operation over TOSLINK – even though our TOSLINK input receivers (and DAC input circuits) support the full 192KHz Data rate.

1x USB up to 96 KHz, 24 Bits – Async Mode (DAC is Clock Master)

Optical / Coax output of CD SPDIF Data only.

The CD8200 does not feature clock-out, as the CD servo Section and USB are already synchronised internally – DAC is the Master Clock (As it should be).

The USB operates in Async Mode – this means the DAC is the Clock Master – and the rate Data is received from the PC is handled via flow control (and small buffer). So the DAC is not (Directly) sensitive to USB Clock Jitter etc – the clock is effectively sent from DAC to PC….. (In simplest terms).

John

[indypants]

Hi John,

The 8200CD sounds very close to what I need, and the idea of replacing two boxes with one is definitely appealing, but I was hoping for the ability to easily switch between CD and another clock-locked source. The intention at some point is to replace my squeezebox3 streamer with something more capable that can take a clock signal. At the moment, I don't know whether that replacement device will be USB (no clock requirement) or SPDIF/TOSLINK (clock requirement) - but I suspect the latter. I've yet to investigate what's available to be honest.

This leads me to some more dumb questions (sorry) ...

1) How does the 8200CD swap between CD and external source - i.e. is it done at the push of a button?
2) Am i right that the MDAC has a single clock-out interface rather than multiple switchable ones?

I suspect I'll need to investigate a USB streaming solution or (worst case) invest in a dedicated MDAC for the streamer if it turns out I do need the clock signal at a later date... ?

Perhaps I'm too hung-up on the whole clock-lock thing. It does make a heck of a difference with the DaCapo/Cardinal, but maybe things have moved on since then? As you've no doubt gathered, I'm no engineer - so I've only got my ears to go by!

As always, I'd really welcome your comments.

All the best,
Tim.

[JohnW]

Hi Tim,

Your questions are not dumb at all.

1. On 8200CD there’s a single "Source Select" button on the front panel, you cycle though the inputs. The remote control has Digital Source select Up / Down Buttons

2. MDAC has only one Clock output - however you can buy an optical splitter - then you have 2 Clock outputs then the only issue with the Clock output is that you have to select the output frequency manually - for CD its fixed 44.1KHz based, simple - however for something like the squeezebox, it could be 32KHz, 44.1KHz or 48KHz based (and multiples)... so you would have to set manually - While it would work, if you commonly play music files with different sample rates, its not that convenient...

Clock Lock is really designed for Single speed playback – however we have left the option open that allows the MDAC to communicate over a Remote control loop (RC Loop), and if we where to design a media transport of some kind in the future, the required sample rate could be automatically communicated over this RC loop.

In theory, header bits imbedded within SPDIF should indicate the playback sampling rate – however we have found that most Media Players do not set these “Bits” correctly, but there’s nothing to stop us having an option in the MDAC Menu system to allow the “Clock Lock output” to follow the sampling rate as indicated by the received SPDIF header. So if the Squeezebox or what have you, is correctly implemented you would have Auto Sample Rate selection of the ClockLock Output.

We added the ClockLock because it’s the ONLY correct method to operate a DAC; however the MDAC / 8200CD (Q) Jitter Attenuation system is so effective, that we have not heard a significant deference between Clock Lock – and standard SPDIF recovered Mode (Atleast when the DAC Jitter attenuator is set to narrower BW modes). We are looking forward to other peoples findings. There are unique Cascaded circuit’s blocks within the MDAC / CD8200 to reduce the "Analogue domain effects" of source input Jitter on the DAC substrate - effectively isolating the DAC from effects of Non-Synchronise input Data within the time domain.

Now that said, my Martin Logan CLSII have been out of action for the past year (These are my REFERANCE speakers) apart from Bass (what Bass), they resolve every detail. Maybe with these in action we would notice a greater effect between Clock Modes. The Jitter attenuation is just that – Attenuation, which means Jitter is Reduced, but not removed 100% - Clock Locking results in Clock Jitter performance that is limited only by the DAC’s reference clock input.

You cannot compare the Clock performance between Dacapo and MDAC / CD8200, I was 19 when I designed Dacapo – times have moved on, so I should not be so embarrassed…

John

[indypants]

John, I've just read and digested yor very interesting reply... thank you.

I foolishly hadn't considered the requirement to manually change the clock frequency for different media files, so thanks for bringing that to my attention. I'm pretty convinced now that despite not needing the pre-amp section (at the moment, at least), I should wait for the 8200CD(Q).

Hopefully my last question for the moment then:
Will the CD8200Q have the same clock-out as the MDAC, and be equally compatible with any media transport that you might bring us in th future

Thanks again,
Tim.

[biovizier]

While your own dac be the original conception from much earlier in this thread of a more esxpensive "ultimate" dac?

Regards,

Nick

[loris.p]

Damn
I thought I bought a working DaCapo but after one month the usual white noise
I'll try to change the caps.
Loris

Ps John(W) i sent a pm.
did you read it?

[JohnW]

Hi Loris,

I'm very sorry to hear this - but it does not surprise me. On the occasions I've attepted repair of the PCB's - they have never lasted.

Pls. don’t waste your money replacing the capacitors - I now believe that due to low voltage operation (when the PCB tracks have gone high resistance due to the leaking capacitor electrolyte eating the PCB tracks). This I believe damages the Logic IC’s within the potted DAC section, which fail after time operating under these adverse conditions.

I believe the failure sequence goes something like this:-

1. Capacitors start to leak eating the PCB tracks.

2. PCB Track becomes High resistance – DAC operates with reduced voltage – however the CLOCK section within the DAC which is supplied by a different supply rail beginnings to “Power” the DAC section, via the Clock Line and ESD protection diodes on inputs the of CMOS Logic and Supply rails (note this is intergrated at 50% duty Cycle - so half VCC)

3. After some time operating under this “Fault condition” the Clock output drivers (which are not “Series Terminated” – my bad) begin to fail…. Finally you end up with the resultant “White Noise” as the clock degrades.

This is one of the reasons I've given up on trying to repair damaged Dacapo PCB's... Your much better going for an MDAC or 8200CDQ

John