PCM1794 vs. PCM1798 in the real world?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
As far as I can determine, the real difference between these two converters comes down to output current and digital filter performance. Its so easy to get caught up in specsmanship and forget that circuits frequently dont deliver in practice. I have a PCM 1798 to work a design up with, but I keep wondering if the additional stop band performance and available output current of the 1794 would provide a cleaner simpler design. I think the 1798 might require a two-pole reconstruction filter to keep the quantization noise of the DS modulator reasonable, where the 1794 may not need any post filtering to reach the same performance. Has anyone implemented these converters with a passive or discrete I/V into a gain stage sans output filter and taken some measurements? Does the filter in the 1794 really provide the extra performance worth paying the difference in cost, or is it just another datasheet fairy tale?
 
Has anyone implemented these converters with a passive or discrete I/V into a gain stage sans output filter and taken some measurements?

Don't know about measurements. But www.kandkaudio.com and www.raleighaudio.com have together developed and marketed a 1794 based dac with the choice of a passive or active output stage. The transformers used in both the active or passive stage...could be seen as filters? Anyway..it sure seems like a simple passive output stage is an easy thing to do with the 1794.


Cheers,
Bas
 
Bas: I have seen your posts about the Rakk dac and it looks like a really good product. I may take inspiration from their passive x-former coupled output stage, but I think I want a lower output impedance since my preamp has a low input imput Z. I might have just bout the Rakk board because it is a pretty good deal, but I dont need a receiver chip and the supporting regulators. I am going to replace the DAC and output stage of a Sony X707ES and use a Tent clock. I think I can put my baord together for less thsan the Rakk DAC. I am interested in how you like the rakk dac though, has it met your expectations?
 
did anyone finaly compared

Hi,

I am interested in making a DAC with either PCM1794 or PCM1798 (I have a friend who is really interested to make one and is forcing me to do some designing). I did 4xTDA1543 and I really like the sound but am ready to check the "new" 24bit technology since TDA chips are extinct :( (I have one PCB left and am willing to trade it for something :) )

Did anyone finally compared PCM1794 and PCM1798? In the Farnell's catalogue the difference is around 3 times in price, and I was wondering would the combination with 4 PCM1798 be better than one PCM1794...

And now there are those DSD1794 chips that come into play as well. I think I have seen looking trough the datasheets that one can make a board that will accept all three of them so the comparison is possible if not expensive for a DIY hobby.

Has someone seen a DIY PCB with either one of chips and CS8416 as SPDIF receiver. I have some in stock and would like to use them? I am willing to make a PCB and offer one to somebody who has done the layout that I like (double layer with metallization and solder mask) :)

Pred

Thanks in advance for eventual replies :)
 
Hello pred.

I am also doing a new dac project with:
CS8416->AD1896->PCM1798->Zapfilter 2
Got all three devices as free samples!!

Only problem is how to connect circuits to get 192Khz sampling out off AD1896, in slavemode and with a 24.57..Mhz external clock.

Maybe we could help each other?

Kim
 
kimschips said:
Hello pred.

I am also doing a new dac project with:
CS8416->AD1896->PCM1798->Zapfilter 2
Got all three devices as free samples!!

Only problem is how to connect circuits to get 192Khz sampling out off AD1896, in slavemode and with a 24.57..Mhz external clock.

Maybe we could help each other?

Kim


Hi Kim,

I wanted to start a thread soon on DIY audio about the topic of PCM179x dac. The problem is my tube preamp that gives me headaches and that I want to finish first so I can start clean with this new project. On the subject of DACs I am not that good... I better with analog. But I have printed out the datasheets and it seems it wan't be that hard to do it.

If you are interested to lead the project I'll be more than happy to participate actively. At work (although I use Protel there, which is nice) I have a little spare time so my projects tend to bee long... But with help end joind forces I think it is possible to do it.

The only thing for me that differs from my previous projects is that the components are expencive so I would like that it works from the start. These small SMDs are imposible, at least for me, to desolder. I decided, also from the experience, that I would go with PCM1794 in dual mono. Yes it is expencive, but when you calculate the cost of the box, mecahnical parts and switches you see that there is no point in cuting the cost on the important part.

So if you want to do it send me an email so we can talk and not bother other people at least at the begining

Have a nice day,
Pred
 
Hello

I think it all sounds interesting about the dual PCM1794.

I just don´t know if it get to complicated then?

I´am now trying to get PCM1794 as free samples.
OK please mail me about your thought and ideas.
I can´t mail you at this bord.

Others are of cource more than welcome to join this project:)
Kim
 
Just a newbie question ... Can I find somewhere the output impedance of the pcm1794 ? ( the selection of I/V opamps depends from it and I can't find it ... )
If I understand correctly, your question is meaningless.
The chip is a current output, and needs a I/V converter.
Since the current is the same into any reasonable resistance, the output impedance is zero, or infinite, depending on your point of view.

HTH

Doug
 
Thanks for the reply Dougl, I wasn't knowing output of pcm1794 was constant current ouput.

I've read that bipo, jfet opamp was working differently with source impédances but if it's a constant current output ... I was thinking about the new opa827 for I/V with the pcm1794.
 
I built an experimental DAC which utilized the PCM1794A with passive resistor I/V that then directly fed my power amp. My extensive experimentation with the board caused the PCM1794A to fail, so I replaced it with a PCM1798 out of curiosity. I did not notice any obvious subjective difference, but, of course, A/B comparison wasn't possible. Objectively, I had measured 0.022% THD with the PCM1794A via 75 ohm resistor passive I/V, but have not got around the performing a distorion test with the PCM1798. According to their data sheets, the differences between the two devices are digital filter stop-band supression level, output current which is approximately twice the magnitude for the 1794A, and a 3dB better SNR for 1794A. Otherwise, they are pin-compatible.

My suspicion is that the 1794A internally contains two converters per channel to deliver that 3dB improvement in SNR, and so, consequently features twice the output current of the 1798. Having a greater output current magnitude can be a benefit or a problem, depending on the I/V circuit employed. With passive resistor I/V, higher output current magnitude enables lower I/V stage output impedance, which can be a benefit if the I/V then directly drives a transformer, for example. However, greater output current magnitude may place significantly greater demands on a active virtual-ground type of transconductance I/V circuit, which has to handle the output current magnitude. The lower the signal magnitude the more linear will be most active circuits.

In short, I've been equally happy with the real world sound of the PCM1798 as I was the PCM1794A. If device cost difference is a significant issue, I think you will be satisfied with the 1798. If you plan tomdo such a swap, don't forget to scale the I/V circuit gain, which is usually set by a single resistor per channel, no matter whether the circuit is passive or active.
 
Last edited:
Account Closed
Joined 2010
.

My suspicion is that the 1794A internally contains two converters per channel to deliver that 3dB improvement in SNR, and so, consequently features twice the output current of the 1798.


However, greater output current magnitude may place significantly greater demands on a active virtual-ground type of transconductance I/V circuit, which has to handle the output current magnitude.


I have a sound card based on 2 pcm1798 for four channel output using some cheap njm op-amps.I thought of making it 1 chip per channel of which i'm not sure how to deal with in the digital coupling section, but if i get solved that maybe i just replace the 4 njm i/v with 2 ada4898-2 to circumvent the problems you mentioned.
I had kinda bad experience with ada4898-1 supplying it with +9v/-12v and using a 3kohm feedback resistor.I suppose it is made to work with lower feedback resistor although it's not a current feedback op-amp or maybe it was just too sensitive to the difference in the power supply...It simply melted in a minute...
Would you have some advice for me on both the digital and analog side with this dac?


I wanted to make this sound card either a phono recording unit either a measurement card...the ADC(pcm 4220) is configured unfortunately only for 96 khz max sampling rate by the xmos or in the firmware so i can't change that.I'm not an expert on the digital side.I could work though on the digital part with some minimal guidance.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.