Question About SPDIF line output Impedance

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

I am creating a CD transport SPDIF output buffer circuit, using a DS 9638 (RS422) high speed line driver chip. I am going to use only the non inverted
output of one of the 2 line drivers in the chip. The higher output voltage above the SPDIF standard is not an issue here.

I have been told by National Semiconductor (the maker of my chip) that the
driver's output impedance is 2-3 ohms. To get the correct output impedance
of 75 ohms (I am using a 75 ohm BNC jack) should I add a 72-73 ohm series resistor between the driver and the jack to get approximately 75 ohms?

If not, what would be suggested?

Thanks in advance!
 
From a brief inspection of the internal schematic it doesn't look like you've been given accurate information. The high side output impedance is going to be higher than the low side, it won't be constant. Perhaps the low-side output impedance could be 2-3 ohms, but the high side will be at least 8ohms seeing as there's an 8ohm series R. If I were you I'd go for a 68ohm resistor in series with the output - 72 or 73 ohm isn't a commonly found value.

If you're interested in creating the lowest jitter S/PDIF signal then I'd go for a CMOS driver rather than a bipolar one as the output impedance is likely to be more consistent between the two output conditions.
 
Last edited:
Having given this application a little more reflective thought, it seems a linear amplifier has a better chance of presenting a constant output impedance irrespective of output level. Hence I'd suggest something like ADI's ADA4891-1 - its available in SOT23-5 which should be small enough.
 
abraxalito:

Thanks for the chip info!

I have 2 problems. One is that I need an 8 pin DIP package as
I cannot work with surface mount components. The second
is how to determine what the output impedance of such a
device would be ( the data sheets never seem to disclose
this information ).

Is there a CMOS substitute available that comes in an 8 pin DIP?

Thanks!
 
abraxalito:

Thanks for the tip about the AD8041. I looked at the info sheet, and it looks very good.

I have seen quite a few SPDIF output waveforms from transports and players (on the web) that look malformed to various degrees. Quite a surprise that there is so much in the way of badly engineered digital outputs.

Thanks again!
 
You're welcome!

I was quite shocked when I first looked at the S/PDIF output of one of my 1980's vintage CD players on the 'scope. Its waveform was nowhere close to straight-edged and flat-topped - I figure now this was because of being sent through an under-sized 'transformer' which looked physically like an inductor.
 
Member
Joined 2017
Paid Member
abraxalito:
I have seen quite a few SPDIF output waveforms from transports and players (on the web) that look malformed to various degrees. Quite a surprise that there is so much in the way of badly engineered digital outputs.

Thanks again!

I have sample pics of my optical SPDIF at 96kHz with TOTX1952 and TORX1952(TOSLINK). Pulse response in optical SPDIF depends on the quality of an optical coupler. TOSLINK is more excellent than others but a little bit expensive. Blue is at the input of TOTX1952(transmitter), yellow is the output of TORX1952(receiver). Cheap one used in made in China product is poor performance; you can't see a square wave.

But optical coupler inevitably makes some jitter(almost 3nsec). It has no impact on the output of a DAC because I tried several methods to measure the jitter of analog output and every time failed to find the measurable jitter. I guess jitter in an optical transfer is absorbed by PLL inside a receiver chip. A coaxial interface has much less jitter than an optical one. But common mode current noise caused by the earth ground which is almost impossible to eliminate is more important than jitter which can be completely eliminated by PLL. That's why I love to use an optical one.
 

Attachments

  • tek00222.jpg
    tek00222.jpg
    120.9 KB · Views: 53
  • tek00224.jpg
    tek00224.jpg
    93 KB · Views: 53
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.