Which chip sounds better with D1

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

First I have to say thank you to Nelson Pass for contributing so much to the community and to Promitheus for the thread on D1 stage - so much useful information. :worship:

My story:
I have been working on a DAC for quite some time now. I started with a cs8416/pcm1730 with opamp I/V stage. I didn't get that to work, but then I discovered D1 and went on with another pcm1730 based design and got some signal through. The sound was very clear but with cracking sound like a bad turntable. I was told that the pcm1730 needed a virtual ground (at 2.54V) input, to work correctly so I moved on to pcm1798, which is the one I'm currently listening to.
I have attached a picture of my current system - the digital board is separate, so I can change it with not too much work. The system is prepared for balanced, but I only use half of the output.

The problem is that I don't think pcm1798 gets the full potential out of the D1 stage because the pcm1730 sounded much better (between the cracks). Can it be the difference in current output? pcm1730 has +/-2.48mA output and pcm1798 only 4mA peak to peak.
Can the nice sound be related to preamplification done in D1 stage instead of by my NAD319? Should I just try to increase the gain in my D1 stages? My current working version has voltage output levels matching my stock CD player completely.

I would of course use pcm63 if I could, but it is so hard to get now and very expensive. Which chips have you guys tried that sound really good with D1? I know that a lot of people have used pcm1704 and appears to be happy about it, but I was looking for some less expensive modern parts.

PS: I hope it is alright that I post this in Pass Labs forum even though it might belong in Digital, but I'm not looking for the ultimate dac, just something that sounds good with the D1 I/V stage and who knows D1 better than people in here. ;)
 

Attachments

  • p1040322m.jpg
    p1040322m.jpg
    90.6 KB · Views: 692
Posting your schematic would be helpful. Also search this forum and the Digital forum for the D1 I/V. There have been threads that discuss how to adjust the D1 for different current and voltage settings for other DAC chips. I seem to remember a thread regarding the D1 and the PCM1738 in recent months.

-David
 
Thank you for directing my attention to the pcm1738 thread - I might want to have another go at the pcm1730 (to my knowledge the same as 38 but with hardware control).
Sure I'll post some schematics. First the digital part.
 

Attachments

  • dachwmode.pdf
    17.5 KB · Views: 163
DC offset

I'm starting to read the datasheet for AD1955, but it looks like it needs virtual ground at some VREF - just like pcm1730. What is the best way to convert the D1 input to accept this? Would it degrade the performance a lot, if a differential buffer-stage was put in front?
 
Re: DC offset

cviller said:
I'm starting to read the datasheet for AD1955, but it looks like it needs virtual ground at some VREF - just like pcm1730. What is the best way to convert the D1 input to accept this? Would it degrade the performance a lot, if a differential buffer-stage was put in front?


You can make the D1 do non zero volt virtual grounds, just adjust the pot until it meets vref. You might have to tweak the resistor values (R24 and R31) little (lowering them) to get 2.45 volts, depending on Vgs of your FETs. This assumes vref doesn't drift over time, or dynamically.

Rob
 
Re: Re: DC offset

Nelson Pass said:
Offhand, there doesn't look to be anything wrong with the
schematics, but that doesn't mean that something's not
working properly. Unfortunately, you'll be in for some detective
work to find out.

Thank you for looking at my schematics. I'm glad to hear that you cannot spot any obvious mistakes, but the sound is just not really better than my DENON 1015 cd player with pcm61 (as far as I know) and opa2227 opamps. I'm confident that your D1 design, even with my home etched pcb, can sound much nicer - I just have to get the right chip and circuit in front.

relder said:

You can make the D1 do non zero volt virtual grounds, just adjust the pot until it meets vref. You might have to tweak the resistor values (R24 and R31) little (lowering them) to get 2.45 volts, depending on Vgs of your FETs. This assumes vref doesn't drift over time, or dynamically.
Rob

Great. That was exactly what I did with pcm1730, where I got the sonics I'm so despearately trying to get again, but without scratches in the sound which I can defintitely do without. I lowered the values from 100k to 47k and got the ~2.5V, but I mailed back and forth with a danish guy that knows a lot about pcm1730's and he said that it was impossible to use such a simple I/V stage for this chip. He was puzzled that I even got sound through.

I think that the pcm1730 chip I used could have been fried or perhaps just evil, but I'm currently trying to make up my mind :headbash: if I should move on to ad1955 or have another go at pcm1730 I think I have one or two left.
 
Re: Re: Re: DC offset

cviller said:

I think that the pcm1730 chip I used could have been fried or perhaps just evil, but I'm currently trying to make up my mind :headbash: if I should move on to ad1955 or have another go at pcm1730 I think I have one or two left.


If you have a scope and have eliminated the power supply as a noise source for your scratches I'd look at the vref out of the DAC with a scope. The datasheet doesn't mention that this voltage varies a lot, but perhaps it does, even varying with the signal. The sample circuit they show is immune to either slow or fast changes in vref out which makes me wonder. If you see variation on a scope then you might have your answer to your noise source and would have to create some sort of circuit to make the gate voltage follow the vref changes.

Just pure speculation on my part, I don't know how those vref voltage outs are supposed to behave.

Rob
 
Re: Re: Re: Re: DC offset

Nelson Pass said:


One of the special parts of the D1 was the clock PLL, and one
of the reasons we discontinued the D1 was the unavailability
of the parts for it.

I can feel another project emerging on my todo list! ;)
I'm not sure however if reclocking is worth it, if I use pcm1798.

relder said:

If you have a scope and have eliminated the power supply as a noise source for your scratches I'd look at the vref out of the DAC with a scope. The datasheet doesn't mention that this voltage varies a lot, but perhaps it does, even varying with the signal. The sample circuit they show is immune to either slow or fast changes in vref out which makes me wonder. If you see variation on a scope then you might have your answer to your noise source and would have to create some sort of circuit to make the gate voltage follow the vref changes.

Interesting ideas - sounds like something to investigate. I don't have access to a reliable scope, but maybe santa will bring me a picoscope... that would be nice, but I think it is out of his price range. :(

The upper 6 bits are treated different from the lower 18bits in pcm1730, so I it makes sense that the scratches were triggered by high digital numbers. Some cd's were much worse than others. I don't have the board any more because I needed some parts for my pcm1798 attempt, but I will redo it in the near future.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.