Hello All,

We have a design for 5th order 1.5bit DAC for audio applications. Apparently the common-mode noise is very high. The degraded common mode noise came to light during testing when a differntial filter with single ended output was used. Because of single-ended conversion having a reference to ground, the circuit was very sensitive to filter component mismatches. Even a 0.1% mismatch was degrading the performance severely.

The designer has worked around using transformers and proved that for purely differential loads the common mode noise is not an issue at all.

We are currently debating the acceptability of this argument for real life loads, e.g headphones. Is the load truly differential ?.  We are concerned if any non-symmetrical parasitics to ground between two sides of a headphone or any speaker may not cause the commone-mode noise to show up.

Does any one have any suggestions or insight into the possible assymmetries of headphones or any speakers that may make the load non-differential ?

ic_engr

Quote:
How are you driving Differential Load with a Single-Ended circuit ? You using two identical circuits for driving diff. loads ?
Why dont you keep your circuit differential if your loads are diff. too ?

Why do you wanna trust your load to be exactly & precisely differential.
Practically, a headphone[or anything for that sake] cannot be precisely differential. What if the length of the wires differ by 1mm ?? or solder dots have different resistance ?

It's always better to keep your circuitry robust.

What is the major contributor to CM noise ?
it it because of degraded -ve PSR ??

Quote:

By 5th order , you mean 5th order Δ-Σ Modulation ?

--
Mayank