I have an older Sony cd 203es. The build quality appears to be decent however the sound is rather shrill, hollow and compressed. I have modded players using the TDA1541 and TDA1540 with tube and/or FET audio output stages. I would like to do the same with this Sony player.
Does anyone have any experience with the Sony CX20152 D/A converter chip and diy output stages?
Does anyone have any experience with the Sony CX20152 D/A converter chip and diy output stages?
I posted a few days ago on the same issue. Nobody seems very interested. Have a look, theres a couple of pictures.
These and very similar DACS were used in many Sonys around then. The CD101 used a 44.1 kHz version I believe.
If you've looked at sites of the usual suspects, you'll have found the chip is condemned by minimalists who like to hang a valve or transformer directly from the DAC.
This is because it is current PWM output, analogue but not audio. The real issue is, can you design a better sample and hold stage than the one shown in the datasheet? One that doesn't require so much filtering afterwards. Or is there another route to audio?
That's my not very expert view at the moment. I guess the objective would be to switch more cleanly and quickly than Sony could. We need to know enough about the switches and the demands they make of their associated opamps, so we can take advantage of more recent components and expertise. If the S+H section can't be improved or avoided, I doubt other changes would be worthwhile, because at present I think the switching makes all the downstream filtering necessary.
The other issue particular to this kind of DAC would be the precision of the clock that times the current sources. Other stuff is normal, like more linear amps and calmer power supplies etc.
I've just been messing around with my simulator to see if I can create the waveform I show in the earlier post, being the output of the DAC with a 1kHz full scale sine. I ended up sketching out the whole DAC, which seems quite easy, but the problem I'm having is in getting an ADC to generate the data to test it. I had an idea about using valves...
I'm surprised the Sony sounds shrill. My Kenwood's got proper grunt and a rather dull top end, I fancy. Maybe I hear what I expect, or Sony were less zealous in dealing with the HF noise.
The Kenwood's got an ace Toshiba mech, BTW, but the boards are skimpy compared to the Sony, I would think.
My Sony's (M75, M95) used TDA1541, paradoxically.
Ian
These and very similar DACS were used in many Sonys around then. The CD101 used a 44.1 kHz version I believe.
If you've looked at sites of the usual suspects, you'll have found the chip is condemned by minimalists who like to hang a valve or transformer directly from the DAC.
This is because it is current PWM output, analogue but not audio. The real issue is, can you design a better sample and hold stage than the one shown in the datasheet? One that doesn't require so much filtering afterwards. Or is there another route to audio?
That's my not very expert view at the moment. I guess the objective would be to switch more cleanly and quickly than Sony could. We need to know enough about the switches and the demands they make of their associated opamps, so we can take advantage of more recent components and expertise. If the S+H section can't be improved or avoided, I doubt other changes would be worthwhile, because at present I think the switching makes all the downstream filtering necessary.
The other issue particular to this kind of DAC would be the precision of the clock that times the current sources. Other stuff is normal, like more linear amps and calmer power supplies etc.
I've just been messing around with my simulator to see if I can create the waveform I show in the earlier post, being the output of the DAC with a 1kHz full scale sine. I ended up sketching out the whole DAC, which seems quite easy, but the problem I'm having is in getting an ADC to generate the data to test it. I had an idea about using valves...
I'm surprised the Sony sounds shrill. My Kenwood's got proper grunt and a rather dull top end, I fancy. Maybe I hear what I expect, or Sony were less zealous in dealing with the HF noise.
The Kenwood's got an ace Toshiba mech, BTW, but the boards are skimpy compared to the Sony, I would think.
My Sony's (M75, M95) used TDA1541, paradoxically.
Ian
Don't discount replacing the opamps. Going for something like the OPA2604 or OPA604 (dual or single) can make a very worthwhile improvement.
I always add the proviso (and seeing you have a scope
),
http://www.diyaudio.com/forums/anal...u-have-checked-see-its-stable-havent-you.html
These were early players and relied on brutal high roll-off filtering to get rid of the HF artifacts. That is always an issue.
The early players used a time shared DAC which also led to phase shift,
http://www.diyaudio.com/forums/digi...gen-sony-cdps-new-direct-mod.html#post2931324
I always add the proviso (and seeing you have a scope
),http://www.diyaudio.com/forums/anal...u-have-checked-see-its-stable-havent-you.html
These were early players and relied on brutal high roll-off filtering to get rid of the HF artifacts. That is always an issue.
The early players used a time shared DAC which also led to phase shift,
http://www.diyaudio.com/forums/digi...gen-sony-cdps-new-direct-mod.html#post2931324
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.