This is not exactly the right place for this topic. It falls between categories. It concerns full range speakers, but is more in the expertise domain of amplifiers. I'm planning to use it with a chipamp project, so I'll ask here.
Several speaker projects use passive low-Q output-side notch filters to cut small reponse peaks and humps out of the cabinet/driver response. Generally, these filters are on the huge/beefy side like a passive crossover would be because they have to be able to dissipate several watts.
I'm building a powered monitor project-- full range driver/enclosure with dedicated internal monoblock chipamp. Because the chipamp is built in and the enclosure will never be used with some other amp, it makes less sense to put the response filter between the amp and driver. A similarly designed passive response shaper on the input would be preferable in that it would not waste output power and could be tiny in comparison (dissipating microwatts instead of watts).
The math and basic filter design are straightforward. I'm not asking for help there. What I do want to ask about is a few practical concerns that can't be captured by simulation software:
1) Is there a reason the input filter shouldn't use off-the-shelf shielded inductors, such as the Delevan 1641 series? These are designed for RF-- but that just makes them 'overqualified' for line-level microamp audio frequency use, correct?
2) How good, really, is the shielding on shielded inductors? They'll be several inches away from a toroidal power transformer.
3) Is there a reason to prefer parallel over series notch filter topology? Parallel topology tends to be used for the response shaping filters that sit between the amp and driver, while I've only seen series topology used for driver impedence correction. My instinct tells me they're entirely equivalent for response shaping and the series topology is easier to compute/account for non-ideal components (eg, the DC resistence of the inductor). We're only talking about cuts of 1-2dB on an input impedence of 100k or so.
[edit: grammar]
Several speaker projects use passive low-Q output-side notch filters to cut small reponse peaks and humps out of the cabinet/driver response. Generally, these filters are on the huge/beefy side like a passive crossover would be because they have to be able to dissipate several watts.
I'm building a powered monitor project-- full range driver/enclosure with dedicated internal monoblock chipamp. Because the chipamp is built in and the enclosure will never be used with some other amp, it makes less sense to put the response filter between the amp and driver. A similarly designed passive response shaper on the input would be preferable in that it would not waste output power and could be tiny in comparison (dissipating microwatts instead of watts).
The math and basic filter design are straightforward. I'm not asking for help there. What I do want to ask about is a few practical concerns that can't be captured by simulation software:
1) Is there a reason the input filter shouldn't use off-the-shelf shielded inductors, such as the Delevan 1641 series? These are designed for RF-- but that just makes them 'overqualified' for line-level microamp audio frequency use, correct?
2) How good, really, is the shielding on shielded inductors? They'll be several inches away from a toroidal power transformer.
3) Is there a reason to prefer parallel over series notch filter topology? Parallel topology tends to be used for the response shaping filters that sit between the amp and driver, while I've only seen series topology used for driver impedence correction. My instinct tells me they're entirely equivalent for response shaping and the series topology is easier to compute/account for non-ideal components (eg, the DC resistence of the inductor). We're only talking about cuts of 1-2dB on an input impedence of 100k or so.
[edit: grammar]