Good question.
***This is a re-post from the madisound discussion board:
I came across a few related issues and would like to convey these to you gents for a sanity check. While I have some electronics smarts, it's just enough to make me dangerous. Eventually I might be effective, but is a process that will consume much time. Certainly my wife maintains that I am a "work in progress".
I came across a few snippets in Audio Express about amplifier feedback circuits. It seems these circuits exist to lower the output impedance of an amplifier. They might have another purpose too, but certainly the amplifier feedback circuits lower output impedance. While most amp manufacturers maintain this bad, I think (not sure) that most manufactures use feedback to maintain a good dampening factor for their amplifiers.
This situation exists because we speaker builders were historically incapable, and presently unwilling to build loudspeakers with a flat impedance. It is more expensive to build conjugate filters, but not terribly more difficult with the measurment tools available today. Historically, (maybe in the 1950s) folks really didn't have the gumption or tools necessary to build a good conjugate filter. Today, the issue is simply a matter of cost. Given the current culture of this matter, it's no longer "necessary" to build loudspeakers with a flat impedance. This is because all (almost all) amplifiers have plenty of feedback and very adequate dampening.
I also certainly understand that John K's experiment with heat remains very valid. Impedance and dcr will change with increased voice coil heat.
So, I think it's possible to build a loudspeaker with a very flat impedance. Good shorting rings and conjugate filters make this possible. However, this flat impedance is relatively insignificant with most amplifiers, but...maybe...(?) this would allow an amplifier with less feedback (better) and a higher output impedance to perform well?? Can I unlock better sound by properly matching a low feedback amplifier to a speaker with a flat impedance?
Historically SETs have a bad-rap, but maybe the culprit is bad speakers with steep impedance swings?
Am I on-track with these ideas?
*** This is a summary of what followed.
1. Considerable wailing and gnashing of teeth. The was significant disagreement regarding the effect of amplifier output impedance on the loudspeaker driver's Qes. This carried extensively, but eventualy this issue was gracefully put to rest. Amplifier output impedance does not effect driver Qes.
2. Some dedgree of consensus. It appears my hunches were correct regarding this issue. It appears that lower amplifier feedback and higher amplifier output impedance would perform favorably when matched with a loudspeaker with a proper impedance compensation network.
So, the next question that logically follows is... what to do about this??? There are a few amplifiers with variable feedback. Certainly it's within my ability to build conjugate network for a loudspeaker. The biggest real problem is the double impedance hump created by a the ported W18 driver in the 1801. Further exacerbating this problem is these double humps will move up/down when the port length changes. So, if I build a conjugate network for the 1801, it would only function with 1 port length. Anybody desiring to change their port length would seriously mess-up the impedance of their 1801. So, for the 1801, an impedance compensation (i.e. conjugate) network won't happen. It isn't very smart.
Further, even if I built a loudspeaker (sealed woofer) with a flat impedance the number of people able to take advantage of this would be extremely small. I speculate the number of folks with variable amplifier feeback settings is smaller than the number of folks using SET amplifiers. As such, VERY few folks would be able to take full advantage of the low feedback/flat impedance advantage.
Nonetheless, am I interested? Sure! Do I have time... holy buckets - no. I am soooo far behind with completed 1801 customers. I also have a couple 3 way speakers that really need to be ordained.
When I built the 1801s, it was primarily because I wanted this sound. It appears other folks wanted this sound too - God willing. This is great. If I build a low feedback amplifier and a flat impedance loudspeaker, it'll be because I want this sound. If other folks want this, great - God willing.
I suppose this is driven my my insatiable desire to learn more. My favorite childhood books were always "Curious George". I thought curious George was a pretty special "person". He did industrius things rooted in curiosity that always came out good. As I grow older, maybe Curious George is a good role-model.
A low feedback amplifier from Dave Ellis won't happen anytime soon. In the distant future....???
Oh, I did hear the Merlin speakers a few years ago, but didn't know anthing about the "BAM".
