Question regarding digital pulse transformers.

0 Members and 1 Guest are viewing this topic. Read 1620 times.

Builder Brad

Question regarding digital pulse transformers.
« on: 2 Jun 2007, 11:57 am »
After listening to the Scientific Conversions SC947-02 digital pule stansformers in stalled in my Paridisea DAC and Squeezebox 2, I am comming to the conclusion that this mod sounds best when fitted to the DAC and not the squeezebox - I know I am not the only one who has come to this conclusion. There is a much harsher sound that is quite tiring to listen to when transformers are employed at both end of the cable, initially I thought it would settle down, but if anything its got worse.

I am wondering if I should be using twisted interconnects rather that the 75 ohm c-axial style of cable when transformer are in both units - the signal and earth will be totally isolated from each component and the transformers would be able to cancel out any common mode hash that is picked up by the twisted wires.

I know this would be easy to test, but I wanted to see if anyone had tried this, or had any observations ect.

Brad

art

  • Industry Participant
  • Posts: 845
    • Analog Research-Technology
Re: Question regarding digital pulse transformers.
« Reply #1 on: 4 Jun 2007, 04:51 am »
It is supposed to be a 75 ohm cable. And that also means 75 ohm BNCs. Not RCAs. Yes, it will work with them, but it won't sound as good.

BTW........Scientific Conversions is the last transformer I would use. The cheap Schott knock-off that Mouser sells (Pulse Engineering?) is better. Better still are the Newava ones that Digi-key sells.

Avoid any with a shield between the windings.

Pat

JoshK

Re: Question regarding digital pulse transformers.
« Reply #2 on: 4 Jun 2007, 01:19 pm »
Why would one avoid transformers with electrostatic shields?  Unless the it is wound bifilar'ly, not having a shield with increase the capacitive coupling.

art

  • Industry Participant
  • Posts: 845
    • Analog Research-Technology
Re: Question regarding digital pulse transformers.
« Reply #3 on: 4 Jun 2007, 03:56 pm »
Right............which is exactly what you want. Scientific Conversions, and Crystal, are both wrong. Tight coupling means fast, clean edges.

Besides that obvious problem.......if you put one on a TDR, that shield looks like a big hunk of capacitance to ground. Which is, in fact, what it really is.

Translation: a nice, big fat reflection. And one that you can not build a network to compensate for.

The only other obvious dumb thing(s) not to try is a 2:1 version. Or one with multiple taps on the secondary (for multiple impedances). Ugh!

Pat