DPA little bit dac modding (7350 dac, ym3623b receiver)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi,


Firstly, most of the components are surface mount so I cannot do much with my soldering skills.

My thoughts were to replace the capacitors with some higher performance ones.

There is a 11.2896mhz crystal on a ym3623b receiver which confuses me somewhat after looking at the 3623b datasheet. :confused:

Anyone know why this clock isn't 16.9344 Mhz ?


Regards,


Ashley
 

Attachments

  • dpa1.jpg
    dpa1.jpg
    31.1 KB · Views: 1,539
Hi Ashley,

The Xtal on the YM3623 is only used to clock the receiver when there’s no input signal - on a good design; this clock is "Muted" during "Lock" conditions to prevent beating (interference) between the recovered clock and the "start-up" clock. The YM3623 does not do this by design; normally an external "Kill" circuit (a diode between the Xtal clock and the "Lock" line) is implemented.

The best upgrade you can perform is to clock lock the unit - using the "Deltran" output - which provides a 256fS clock to the transport.

The LM358 are not in the direct audio path - but lower the impendence of the internal VREF sources within the SAA7350 – so don’t worry about there quality – elsewhere on the board you will find 5524 / 32 which form the LPF / output buffer – these could be upgraded.

The YM3414 is not that great a digital filter - if your skilled - you could replace with a PMD100 / 200 HDCD filter which will elevate your Little Bit to new levels.

John
 
Cheers John,

Makes sense about the NE5534!

The unit unfortunately does not have a clock out on the back connectors :( to lock with the player. But i don't see why the clock couldn't be fed back, and this could explain why the 11.2896 crystal is in there! (feed to Phillips based player)

How would i create a circuit to feed back a 11.289mhz signal?

Is the 256fS the 8X oversampled signal?


Regards


Ashley.
 
Hi Ashley,

I’m sorry I thought that the Little Bit had a “Deltran” (Clock) output. Adding such a feature is not really a simple DIY job. The original Deltec / DPA clock-feed to the transport operated at 256Fs (44.1 KHz x 256 = 11.2896MHz). The reason the Little Bit uses the 11.2896 Xtal is not because it was important for correct operation of the DAC – but Rob Watts (the unit’s designer) uses this frequency Crystal for the Deltran function on other designs – so had this frequency to hand.

256Fs (11.2896MHz) is the Master clock frequency used in Philips I2S based CD players / digital audio chipsets.

Operating the early Toslink TX and RX devices at 11.2896MHz was beyond there rated Spec. resulting in increased clock Jitter, however as CD player servo chipsets are tolerant to Jitter in significant amounts – this is of little consequence.

On designs I developed for Pink Triangle and Cambridge Audio, I used 128Fs (5.6448MHz) for the Clock-Lock, this allowed operation of the Toslink TX & RX devices within there specified limits, and also easy clock division / multiplication of both 256 (divide by 2) and 384 (divide by 3) Master clocks commonly used in both European and Asian CD players.

I believe that Arcam used 96Fs for there Clock system – however I cannot confirm this.

If you want to develop your own clock-lock system, then I recommend you use a 128Fs clock. The Little Bit requires a 384fs (16.9344MHz) internal Master clock, which you can then divide by 3 for the Clock-Lock output.

You can use a simple L/C overtone circuit to multiply the clock from 128Fs to 256Fs or 384Fs, depending on your transport requirements.

If you do Clock-Lock, then use optical I/Os (Toslink) for the SPDIF & Clock, thereby affording EMI isolation between the DAC and CD transport – however the Little Bit is an EMC nightmare in its own right (can you watch TV with your unit powered?) – Deltec / DPA designs where REALLY poor from an EMC perspective.

FYI 8x oversampling operates @ 44.1KHz x8 = 352.8KHz

John
 
Hi John

Cheers for the detailed notes.

Yeah, I did notice that when I had the little bit on i got lines on my screen. I put it down to an over active imagination :rolleyes:

Does the EMI spread out everywhere, into the environment, down the spdif cable and into the transport, etc?



Regards,


Ashley.
 
Hi Ashley,

Yes, the RF energy is radiated "everywhere", units directly connected to the DAC via audio, SPDIF data and even though the mains cabling are directly effected - its by no coincident Deltec / DPA digital products gained significant audio improvement when used with Deltec RF mains filters - these attenuated the RF getting into the units, and also transmitted to the mains supply from the units...

Fin, get Skype working and contact me - it’s much easier to discuss your options.

Cheers,

John
 
Ashley,

I had personal experience of RF & Deltec products - there distributor in HK had a load of duff units - and after Deltec had gone under, he approached me to repair them - mostly duff input IC's damaged by earth leakage currents.

While repair these units, the neighbor on the 3rd floor below, used to complain about bad TV reception while I tested the units!

John
 
Hi John,

In reference to the Deltran system. Was this type of design actually held under a patent by DPA or any other company?

Following on from your advice on 128fs transport, I have attached a summary diagram .


Are there any potential problems with the clock lock feedback system?

Does it achieve some kind of equilibrium, and is this the 'clock lock' ?



Kind regards,


Ashley.
 

Attachments

  • dpa.gif
    dpa.gif
    3.4 KB · Views: 829
DPA Pdm series 2

Hi,

Just purchased a DPA PDM series 2 dac.

Sounds very nice.

Inside it's more complex than the little bit but still uses the saa7350 dac.From what I can tell the changes are:-

-an extra transformer, and regulators for analogue stage
-op amp stage replaced with DPA's own surface mount opamp board
-optical input removed
-lock LED indicator.
-sticker on the front "7350", and "Enhanced bitstream chipset"

I'm not sure if the saa7350 is run at 384fs in this version.

Can anyone translate this information about DPA dac's ? (not the whole article but just a summary)

Online translators see to get muddled!

http://www.audio-gd.com/audio/two/new_page_5.htm
 
DPA units & R.F

Hi John (& all),

I am really interested in what you have to say on the point of RF emissions on the dpa gear !!!

I'd love to know.....is there any way of shielding the RF in the related stages in the PDM2 ?

I was speaking to a mate the other week, and said, that although Rob Watts seemed to come up with particular design solutions...he never seeemed to take notice of other design factors (i.e. attributed sound quality in components, and now,it seems, RF problems).

My mate & I are both victims of the PDM2 variant ...;)

look 4ward to your comments (can I email you direct - pls ??)


-A-
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.