|
Digital Drive Upsamplers, DACs, jitter, shakes and analogue withdrawals, this is it. |
|
|
|
Digital Drive Upsamplers, DACs, jitter, shakes and analogue withdrawals, this is it. |
|
In Reply to: output stage: differential vs transformer posted by Joe Murphy Jr on September 26, 2010 at 09:34:12:
Hi,
Transformers correctly implemented for SPDIF outputs have no drawbacks and offer full isolation of the source from the DAC. This helps a lot if the source is also connected to an AV system or is a 'puter.
The problem is that it seems that few designers either care to or know how to correctly specify, characterise and implement transformers. For example, this is a tube driven, transformer coupled output, driving 2m of 75R Cable with Canare RCA Plugs and sockets on either end and 75Ohm termination on the far end, measured at the far end.
Or in other words, this is the signal the receiver would see after output stage, cable and two sets of Canare "75 ohm" RCA connections.
The common way of doing it (buffer gates followed by a too small coupling cap and a resistive divider taken from a datasheet schematic, without any consideration of the actual circuit in which this sits, impedances or levels then driving the transformer) is about the worst possible option I can think up, yet practically everyone is doing it like that.
The way this often looks WITHOUT the extra complications of cable and connectors can be seen on Lukasz "Lamptzator" Fikus's website .
BTW, consider the scale in all of this. The scope trace above covers around 1/3rd of one cycle in the pictures of the Lampizator site, at the level of scaling on the lampizator site it would look essentially perfect.
Lukasz councels against using transformers, but of course, he too does not know how to apply them right.
A differential output stage instead has no isolation and many other drawbacks, but it can be implemented reasonably well even by complete idiots.
It has zero advantage over any SE output stage, unless the receiver is differential and even in that case the benefits are very limited.
If the exact receiver used is well known (e.g. it is a Cirrus Logic CS8414) we may make a DC coupled driver circuit that combines correct impedances for twisted pair (AES/EBU style) with suitable connectors (almost only RJ-45) and directly drives the receiver at the receivers best levels etc.
But this requires reliably knowing your receiver chip and is not an option outside extreme DIY systems or systems with proprietary connections and if you are going to break with the SPDIF and AES/EBU standards and compatibility with said anyway, you might as well go I2S, using nice commercial differential line drivers and receivers plus correctly applied transformers on both ends to achieve isolation.
Ciao T
Thor
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to intolerable pain. Really, what more could we need?
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- RE: output stage: differential vs transformer - Thorsten 09/29/1008:51:27 09/29/10 (0)
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
