Learning now! Entirely passive seems like a tough proposition since the output will have a variable impedance and a high one at that. For a home/hobby project where you can control the next stage maybe its feasible - but in commercial proactive that sounds, to me, like suicide.
Transformers are a natural for the task, but are typically big and expensive ( and most that are not , are not very good) . I have not experimented with any custom ones. I plan to use discrete circuitry, combining FETs and BJTs where each as an advantage - and keeping it pretty simple. (what was it that Einstein said? "as simple as possible, but no simpler" :-)
The output characteristics of bitstream DACs vary quite a bit too- some only give a voltage output - meaning an onboard opamp i presume. The real world gets messy. But again, still learning. All my DACish work so far has been on the other end of the chain - input, isolation, clocking etc.
Transformers are a natural for the task, but are typically big and expensive ( and most that are not , are not very good) . I have not experimented with any custom ones. I plan to use discrete circuitry, combining FETs and BJTs where each as an advantage - and keeping it pretty simple. (what was it that Einstein said? "as simple as possible, but no simpler" :-)
The output characteristics of bitstream DACs vary quite a bit too- some only give a voltage output - meaning an onboard opamp i presume. The real world gets messy. But again, still learning. All my DACish work so far has been on the other end of the chain - input, isolation, clocking etc.