[sdiy] Isolating an external DAC from computer Soundcard

[harrybissell]

There is a new kind of coupler called IsoLoop ... it uses a magnetic technique.
It might be fast enough...

Isolating grounds usually does not work. you need them to have signal integrity
(ie complete the circuit) unless you have other means like an optocoupler.

Perhaps you can join your circuit to the soundcard (maybe with an isolated
supply whose ground is tied to the soundcard) and isolate the signals at
the other end of the system..?

H^) harry

Linium wrote:

> Hello,
>
> After months and months spent around my highspeed computerized DAC CV
> generator with MUX, recent tests are showing that it is mostly working (i
> just need to sync well the INHIBIT signal for the 4051, cf my other post).
>
> Today whereas i was playing bass guitar along with my sequencer, my computer
> crashed weirdly and i suspect that my new invention caused this.
>
> To sum up quicly, i take spdif TTL signal and 24 mhz clock from a computer
> SoundBlaster LIVE soundcard "AUD_EXT" connector and connect it to a external
> circuit containing a DIR1703 (spdif->i2s), a TDA1543 (i2s DAC), some AOP, and
> various CMOS chip (counter/Mux/).
>
> This external circuit has its own power supply and i suspect that there is a
> need for proper isolation between the signals taken from the soundcard
> connector and the external circuit.
>
> So i guess i would need to isolate the 3 lines linking the 2 sides:
> -GND
> -SPDIF (TTL)
> -24 MHZ CLOCK (TTL)
>
> I thought about an opto-coupler but so far i don't find a high speed one able
> to support 24 mhz.
>
> Do you think that my trouble could mainly come from linking the 2 grounds
> (computer card and external supply) together ? And may be a simple
> counter-mesure would exist to somehow isolate grounds (caps) ?
>
> Or should i insist on finding an optocoupler, if so, any ref for a high speed
> one would be great :)
>
> Thx, for the reading :)
>
> Linium