Everyone says that air-core inductors produce less distortion than solid-core types. It's conventional wisdom, repeated in online echo chambers forever.
Is there a primary source on this? Has anyone measured inductor distortion and published their method and data? What magnitude of distortion is produced in a passive crossover inductor -- 10%, .1%, 0.001% -- of any given type, and under what conditions?
Solid core inductors allegedly saturate above a certain current level and become very nonlinear. How likely is that to happen in a typical loudspeaker setup? Again has anyone quantified this?
I am tempted to measure some inductors myself.
It sounds like solid-core (iron, ferrous, laminate, whatever) types have their advantages:
- lower cost for a given value
- lower DC resistance for a given value
- less EMF outside the inductor
All my speakers have at least one solid-core choke. I might consider replacing them if there's data to suggest that some alternative is better.
Thanks!
Is there a primary source on this? Has anyone measured inductor distortion and published their method and data? What magnitude of distortion is produced in a passive crossover inductor -- 10%, .1%, 0.001% -- of any given type, and under what conditions?
Solid core inductors allegedly saturate above a certain current level and become very nonlinear. How likely is that to happen in a typical loudspeaker setup? Again has anyone quantified this?
I am tempted to measure some inductors myself.
It sounds like solid-core (iron, ferrous, laminate, whatever) types have their advantages:
- lower cost for a given value
- lower DC resistance for a given value
- less EMF outside the inductor
All my speakers have at least one solid-core choke. I might consider replacing them if there's data to suggest that some alternative is better.
Thanks!
In the absence of distortion measurements on solid cored inductors, you could look in the first instance at distortion in transformers.
Don't forget that solid cored inductors are likely to show lower AC resistance (in addition to your list of 3 advantages) compared to air-cored because they'll in general be wound with shorter lengths of thinner wire.
Solid core will only give lower EMF outside the inductor if their magnetic circuit is a closed one - many aren't. For example, many lowest cost crossover chokes are wound on a ferrite rod.
I rather favour gapped ferrite pot cores for inductors - they have a virtually closed magnetic circuit (so long as the gap's small) with low radiation and they're pretty linear and harder to saturate than ungapped ones.
Don't forget that solid cored inductors are likely to show lower AC resistance (in addition to your list of 3 advantages) compared to air-cored because they'll in general be wound with shorter lengths of thinner wire.
Solid core will only give lower EMF outside the inductor if their magnetic circuit is a closed one - many aren't. For example, many lowest cost crossover chokes are wound on a ferrite rod.
I rather favour gapped ferrite pot cores for inductors - they have a virtually closed magnetic circuit (so long as the gap's small) with low radiation and they're pretty linear and harder to saturate than ungapped ones.
electromagnetics in free space is very linear, air is nearly as good
good metallic conductors are very linear too
eddy current/proximity effect in linear metallic conductors is complex, gives "funny" frequency^1/2 loss but still obeys the conditions for linear circuits - no new frequencies, distortion products/harmonics are created
the added loss can be dealt with by using bunch or Litz wiring
any ferromagnetic material however is a different story - all are variously nonlinear with saturation and hysteresis
good metallic conductors are very linear too
eddy current/proximity effect in linear metallic conductors is complex, gives "funny" frequency^1/2 loss but still obeys the conditions for linear circuits - no new frequencies, distortion products/harmonics are created
the added loss can be dealt with by using bunch or Litz wiring
any ferromagnetic material however is a different story - all are variously nonlinear with saturation and hysteresis
Last edited:
jpc,
I understand your concern and willingness to make an
investigation of your own. Visaton says something about
saturation currents around 6 A for coils they sell.
There are speaker manufacturers using cored coils in their
flagship models and there are also the ones using air cored
with very thick wires.
Subjectively and strictly IMO I don't think there is much to
be worried about distortion, saturation...
It could happen speakers first be destroyed before
hearing significant distortion or whatever effect may appear.
Still subjectively, think there is more reason to go with
air cored coils because people like to think they are on the
safe side and are willing to pay for that extra feeling.
I understand your concern and willingness to make an
investigation of your own. Visaton says something about
saturation currents around 6 A for coils they sell.
There are speaker manufacturers using cored coils in their
flagship models and there are also the ones using air cored
with very thick wires.
Subjectively and strictly IMO I don't think there is much to
be worried about distortion, saturation...
It could happen speakers first be destroyed before
hearing significant distortion or whatever effect may appear.
Still subjectively, think there is more reason to go with
air cored coils because people like to think they are on the
safe side and are willing to pay for that extra feeling.
Last edited:
People have measured it - one writer in Speaker Builder found the distortion was more than air core, but not necessarily a huge number. I don't have the article handy, but I deem to recall the distortion for an iron core (no laminations or special steel) was less than 1% (but I don't remember signal level, etc...)
I read once (a loooong time ago) that saturating cores turn sine waves into triangle waves.
BTW, I wouldn't hesitate to use a decent quality cored inductor for any value over 1mH. The only exception would be if I needed the insertion loss from a high DCR.
I read once (a loooong time ago) that saturating cores turn sine waves into triangle waves.
BTW, I wouldn't hesitate to use a decent quality cored inductor for any value over 1mH. The only exception would be if I needed the insertion loss from a high DCR.
Last edited:
jpc,
You're a "very naughty boy"*. You should read alan-1-b@diyAudio here.
*It is debated over and over...
cap DA and ferromagnetic core hysteresis aren't similar at all - cap DA has a linear model that works very well to explain measurements - so you again get "funny" frequency^1/2 terms and is still quite linear - no new distortion product frequencies are created with pure DA - see Pease "soakage" articles What's all this Trapped Charge and Dielectric Compression stuff anyhow? (Pease refs 1st para)
ferromagnetic hysteresis/nonlinear permeability does create distortion products in inductors using them
ferromagnetic hysteresis/nonlinear permeability does create distortion products in inductors using them
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.