Bob Cordell said:
True, and it will take a long time unless a killer paradigm overwhelms any remaining resistance.
True also, and hard core tweakers will not give away the capability to mix and match, no matter whether the results are in fact not as good as well designed powered boxes in an obvious way.
My hunch is mainstream schemes will go toward self powered speakers not only for subwoofers as it happens today, but for all boxes. A further wellcome change will be wireless digital distribution, a significant step to clear the wire mess (and more important, a boost to the wife acceptance factor). May be this could be the killer paradigm I mentioned before.
Rodolfo
ingrast said:
I agree. I've actually pursued this paradigm with my Athena active loudspeakers described on my site at www.cordellaudio.com. Each cabinet has four 125 wpc MOSFET power amplifiers built into it, along with the active crossovers and EQSS bass equalization.
The wireless audio distribution concept is also a great one; the key there is maintaining pristine audio quality by avoiding any lossy compression.
Bob
ingrast said:
The amps in an active set-up need not be built into the cabinet, so "mix-and-match" is not an issue:
http://www.linkwitzlab.com/orion_us_specs.htm
Bob Cordell said:
Bob,
My apologies for replying so late. Somehow I missed this post.
I see your point on the input referred distortion, thanks for that explanation, makes a lot of sense and helps to understand these issues.
Your last point, referring to that work of baxandall (it was Baxandall), I am quite familiar with that (in fact there is somewhere on this forum a figure of the phenomenon I posted long time ago).
The higher order harmonics do however increase with increasing feedback in this situation, beyond 10dB and further. As feedback increases, the order of the harmonics that are added also increases but with progressively lower levels. It depends on the particular configuration at which feedback point we reach the situation that we can safely assume that all harmonics are inaudible. That point certainly can lie beyond 10dB feedback.
Jan Didden
janneman said:
I think someone needs to do an experiment with a given circuit and a spectrum analyzer in which the amount of feedback around the given circuit (maybe just a simple common emitter stage) is progressively increased, with the output level held constant, to see how the various harmonics behave in amplitude vs amount of feedback. This could probably also be simulated. I'm pretty sure that at about 20 dB NFB and beyond, the amplitudes of ALL of the harmonics only go down monotonically.
We have to remember that even with a single BJT common emitter stage, harmonics exist all the way out due to the exponential Vbe characteristic - it is only a matter of what their amplitude is.
Cheers,
Bob
Bob Cordell,
---I think someone needs to do an experiment with a given circuit and a spectrum analyzer in which the amount of feedback around the given circuit (maybe just a simple common emitter stage) is progressively increased, with the output level held constant, to see how the various harmonics behave in amplitude vs amount of feedback. This could probably also be simulated. I'm pretty sure that at about 20 dB NFB and beyond, the amplitudes of ALL of the harmonics only go down monotonically.
---
Are'nt you precisely describing the Peter Baxandall's Wireless World Papers circa 1978 ? Do you need I scan them ?
---I think someone needs to do an experiment with a given circuit and a spectrum analyzer in which the amount of feedback around the given circuit (maybe just a simple common emitter stage) is progressively increased, with the output level held constant, to see how the various harmonics behave in amplitude vs amount of feedback. This could probably also be simulated. I'm pretty sure that at about 20 dB NFB and beyond, the amplitudes of ALL of the harmonics only go down monotonically.
---
Are'nt you precisely describing the Peter Baxandall's Wireless World Papers circa 1978 ? Do you need I scan them ?
AKSA said:
I agree, Hugh. It is like any tool; those who do not use it properly can have bad things happen; those who view it as a cure-all for a fundamentally poor design will get bad results, and NFB will unfortunately get painted with a broad brush. Just because someone cuts his finger off with a table saw doesn't mean that we should not use that tool; it just means that HE should not use that tool.
Tell us a bit more about what you had in mind in referring to the problem with the derivation of the error signal.
Cheers,
Bob
AKSA said:
From what I have seen a lot of Hugh's designs are from the 1970's including the old fashioned miller cap - which delays NFB. I think one of the reasons he can get away with it, is cause of the PSRR or signal to noise ratio of modern sources such as CD players/pre-amps.
I think some of them designs wouldn't be acceptable in the 70's.
Lastly, I've never heard anyone say how good they are running decent loudspeakers. What I mean by that, is technically excellent designs on the acoustic level - they almost certainly drop to seriously low impedance.
Regards
Kevin
Fanuc said:
Not so. While some very good loudspeakers do drop to low impedances, there is no fundamental reason why this need be so, and some very good speakers have very civilized impedance curves - or at least ones that do not drop below about 2.5 ohms.
Nevertheless, I agree that amplifiers should be designed with good margin against nasty loads.
Bob
Hi,
an 8ohm speaker that drops to a minimum near 2.5ohms is probably not an 8ohm but a 4 to 8ohm with a low impedance bass driver or twin bass drivers to get the bass/mid sensitivity to match the treble sensitivity.
If it were a true 8ohm speaker then a fall to 2.5ohm would indicate a very severe reactive load and would certainly needs a robust amplifer.
In either case, a genuine 4ohm capable amplifer should probably be used to drive a speaker with this impedance characteristic.
an 8ohm speaker that drops to a minimum near 2.5ohms is probably not an 8ohm but a 4 to 8ohm with a low impedance bass driver or twin bass drivers to get the bass/mid sensitivity to match the treble sensitivity.
If it were a true 8ohm speaker then a fall to 2.5ohm would indicate a very severe reactive load and would certainly needs a robust amplifer.
In either case, a genuine 4ohm capable amplifer should probably be used to drive a speaker with this impedance characteristic.