PCM1794 output stage

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

meta

Member
Joined 2002
Hey...

On my continued dac quest (and having realised the pin spacing on a PCM4104 :bigeyes:) the PCM1794 presented itself. As i'm after balanced outs would this circuit be okay? It uses the same i/v stage (using NE5534s) as on the application datasheet and a OPA1632 for a fully-differential output (that part from 1404 datasheet)...

PDF

Do I need to do anything else - i'm really wingin my way thru all this...:rolleyes:

Cheers.
 
M

meta

Member
Joined 2002
Why do you have 100k resistors on the outputs???
Click to expand...

I put them there following the circuit in the PCM4104 datasheet (also using opa1632).

I recommend you read through this thread http://www.diyaudio.com/forums/show...&threadid=34324 before deciding on the 5534's.
Click to expand...

Interesting - could Jockos' discrete i/v stage be used in this instance then? this is where things get confusing for me :xeye:.

or are you just wanting balanced out from one chip and not two, used in mono?
Click to expand...

Not quite getting that...I want 2 balanced outs (left n right chans) following the ti sheet for balanced mode means mono'ing the dac so I lose channel...

So, goin on schem as is - will it work?

Regards

Thanks
 
5th element

5th element

Member
Joined 2002
I mean are you wanting to use only one dac chip and not two. So you want to use a stereo output from a single chip but get balanced.

Actually using these chips is a piece of cake, there is no reason why not to use two PCM1794's in mono mode. Obviously you need more opamps and chips etc so it will cost more, so if that is an issue then it makes sense to use one PCM1794. But appart from cost. I built my first DAC (still quite new in the digi threads) to see if I could get it to work.

If I make another one I think Ill use AD's upsampler and two PCM1794s on the output. Obviously this will give me balanced which I wont actually use but it cant hurt to have a chip on each channel.
 
O

ojg

Member
Joined 2003
meta said:

I put them there following the circuit in the PCM4104 datasheet (also using opa1632).
Click to expand...

D'oh! Of course! I didn't think of this but there should indeed be resistors there, but it would work much better with 100R instead of 100k ;)

I think you should remove the 22pF cap. It is only there in the datasheet because of the voltage output of the PCM1804, not functional when you use I/V converters.

Other than that it should work great! Build it and keep us updated!

And yes you will get balanced stereo out of one PCM1794 with this circuit.
 
5th element

5th element

Member
Joined 2002
meta said:
Yes, it looks very nice :)
Click to expand...
Cheers :D

ojg said:


D'oh! Of course! I didn't think of this but there should indeed be resistors there, but it would work much better with 100R instead of 100k ;)

I think you should remove the 22pF cap. It is only there in the datasheet because of the voltage output of the PCM1804, not functional when you use I/V converters.

Other than that it should work great! Build it and keep us updated!

And yes you will get balanced stereo out of one PCM1794 with this circuit.
Click to expand...

Yes 100R and not 100k, the data sheets for both the PCM1794 and 98 specify the 22pf cap, I have used the cap in my 98 DAC and it sounds really good so it certainly doesnt mess up the sound in anyway.
 
O

ojg

Member
Joined 2003
5th element said:
Yes 100R and not 100k, the data sheets for both the PCM1794 and 98 specify the 22pf cap, I have used the cap in my 98 DAC and it sounds really good so it certainly doesnt mess up the sound in anyway.
Click to expand...

Now I see that there are more than on 22pf cap here. REmove the one between the outputs of the 5534 and keep the ones between pins 5 and 8.
 
M

meta

Member
Joined 2002
Heyhey...

Nice, I've updated the PDF link following some of the advice here :). So, I think i'll go with this circuit initially and if all is dandy try a 1dac per chan brd to see if theres any major dif...

A discrete i/v stage followed by a DOA (mebbe some iron too) would be lovely tho :D - it's just the expense/knowledge I lack there to implement correctly I think...

Thanks > again.
 
B

borges

Member
Joined 2003
Congrats on your design!

I'm using the PCM1794 with my own oversampler (Xilinx demo board) to compare it to the PCM1704. I've got a 2nd order filter on a pair of 1704's and on the R outputs on a pair of 1794's. Then I've got a different filter on the L outputs on the same 1794's. According to TI/BB, the two types of DAC accept the same input data format.

Did anobody out there ever try using a passive IV converter on the 1794? I just need to run a few more simulations to determine my component values before I solder mine in. I hooked up four bias current sources based on 431's and big surface mount NPNs to remove much of the 6.2mA DC output current.

Besides, TI/BB could easily have been more instructive on the polarity of the different outputs when using the device in mono mode. I should have read the referred thread before designing my board, because now at least half my outputs are inverted.

Greetings,
Børge
 
M

meta

Member
Joined 2002
Hey Bas...

That looks nice...

So the i/v stage (just a few transistors?) in that pdf is powered by it's own tx and psu circuit - really needed? Im after 8 chans in a single box.

It looks nice n simple tho (am i missing summin) u just go from the dac (1794 here) outs to where it says ye? No opamps at all in the signal...


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