Speaker/Amp Impedence Interface.......

What happens, sonically speaking, when you have a speakers with substantial droop, say from 6 ohms to 2 ohms at 100 hz and an amp which has a substantial rise in out put impedence at the same frequency?

What happens when you have the same degrees of rise and droop but they are at different frequencies?

Anyone have any thoughts on this issue?
Depending on what the impedance differences are at the point where the speaker impedance is sagging and the amplifier impedance is rising, and whether or not it is a positive or negative differential, the most likely results would be an increase in power draw from the amp with a lack of control and poorer transient response.

As the impedances match, power output "should" theoretically increase to maximum and transient response should be perfect. That is, if the amp was properly built and designed. Since such is not the case in the real world, the problem with this is that the output of the amp can now be "modulated" by the reflected EMF of the loudspeaker to a much greater degree. In effect, the speaker is no longer acting as a power sink for the amp so much as there is an even exchange of power taking place between the two devices. The end result of this is a lack of control. This "lack of control" can also cause non-linearities in the amplifier to increase, resulting in oddball distortions and poorer transient characteristics.

While this could take place at any given frequency at any given time, the fact that low frequencies require the greatest amount of driver excursion also means that low frequencies generate the greatest amount of reflected EMF from the speaker. This is why "sloppy bass" is such a common factor in system sound, especially with speakers that lack self-damping and tend to "ring" as it is. As has been previously noted in other threads, a sealed and stuffed speaker design is typically miles ahead in terms of transient response and self-damping as compared to other common vented designs.

I would suggest looking at the white papers that Bob Carver has published pertaining to the original Sunfire subs. He talks about reflected EMF and what it takes to overcome such things AND keep the amplifier in control at all times. This is NOT to say that i have great faith in the Sunfire sub as a product and end result of his understanding of the subject, only that there is a great amount of thought that went into it. There are many products that aren't what they should be given the amount of knowledge and technology that the designer had available to them.

Other than that, the original subject as posted by Newbee is a very complex matter and could be a great source for learning about the interface between amplifiers, loudspeakers and loudspeaker cabling as they interphase with each other. Sean
In addition to Sean's good response above, if the impedance of the speakers dip at the same frequencies as the output impedance of the amp rises, then you are losing damping factor(speaker impedance divided by amp output impedance) which can result in what Sean describes above.

If you have the impedance dip from the speaker at different frequencies than the amp output impedance rises, then it is a whole different matter and not really problematic at all in most cases, if not all.

The amp output impedance should never rise to a level higher than the speaker impedance at the same frequency point(with a normal amplifier), or instability, oscillation, or blowing up of the amp could occur.
A impedance equalizer also call a zobel network may help with this situation. Usually used in crossover help maintain speaker impedance as frequency increase. There are also amp that have this network within the output stage which helps stabilize rail voltage. Without this, instability can happen just like Sean and TWL said.

Thanks for the responses. Now time for the dumb question! Would a rise in output impedence from an amp which resulted in a loss of damping as suggested by both Sean and Tom cause a loss of control of the bass driver resulting in a loss of apparent bass or would the lack of definition in the bass give it an apparent (as opposed to real) boost? Or would the tonal response remain flat, just less defined?

On a slightly, but I think only slightly different issue -
I have been listening to a medium powered PP tube amp. (As background I also have on hand three high powered PP tube amps).
The speakers I am testing have a nominal 4ohm impedence but I know nothing about its actual curves and how low or where it droops or spikes. It has 4ohm taps. My high powered amps drive these speakers to my great satisfaction, however when connected to the smaller amp the upper midrange up thru the high end gets quite ragged (the bass is OK). Would you consider this evidence of a probable impedence mis-match between the speakers and the amp, or are their other issues I'm not considering? BTW the sounds I'm referring to were always at medium levels - they didn't seem to be the result of overdriving the amp in any obvious way. As I read your comments above I would think that it was an impedence mis-match problem. The amp I'm referring to is a very popular one, often tested, and no one else has mentioned this as being a problem for them. Thanks...........
Newbee, a reduced damping factor between the amp and speaker can definitely influence the control of the bass driver(s).

Exactly how the loss of control manifests itself may not be exactly predictable, because sometimes the speakers(especially ones with lower mass and strong motors), may have very good inherent damping within the speaker design itself, or it may have poor inherent damping within the speaker itself. The better the inherent damping characteristics of the speaker itself, the less dependent it is on the amplifier to control it. In some cases, a lower damping factor can actually improve transient response.

In some extreme cases, a very low damping factor can cause the frequency response of the amplifer to rise and fall with the variations in the speaker impedance curve. We try to avoid this by proper matching of the impedances, with a goal of at least 5 to 1(speaker to amp impedance). This keeps the frequency response curve relatively flat within a db.

With 4 ohm speakers and a tube amp, it is possible that the impedance curve of the speaker in question is stressing the amplifier at certain frequency points, even at moderate volume levels. If the speaker's impedance dips to a point where the amp is uncomfortable(or approaching instability), then it could definitely have audible effects at these frequencies. Since, with multi-driver speakers and the crossovers involved, there could be impedance swings at anywhere along the spectrum, it can be determined by getting the impedance curve charts, and looking to see if there is a problem area.

Many tube amps do not have very low output impedances(especially if zero or little feedback is used), and reactive speakers with 4 ohm nominal loads could easily bring the amp to some problematic conditions.

Generally, if the tube amp has an output impedance of less than 1.6 ohms, and the speaker has impedances of 8 ohms, you will have the needed 5:1 ratio, as long as the speaker doesn't have any really reactive dips that could compromise this ratio too badly. If the amp has an output impedance of under 1 ohm, then it should be able to handle relatively benign 4 ohm loads. If the 4 ohm load drops to 2 ohms or 1 ohm, then the amp is going to have trouble.

Additionally, since tube amps typically produce lower power as the impedance drops, then it could go into clipping much sooner than when it is driving higher impedances.

As you can see, there could be a variety of reasons for what is happening in your case, and it is the matchup, and not necessarily the quality of the products that may be making the difference.
Tom, thanks for you comments...now for a laugh at my expense! The speakers I was using were 4 ohms nominal and I never thought to do the obvious, try the 8 ohm taps. Shouldn't really work right? Well the bass is a little softer but the mids and highs smoothed out considerably. Very servicable now! Again, thanks.....
I have to admit I know little about tube amps or their interaction with various speaker impedances. According to my understanding many tube amps prefer higher impedances, 8, 16 and sometimes more. The Speltz autoformers seem to make sense in that they can multiply speaker impedance to a level where tube amps can work optimally. I'm planning my first tube amp purchase in the near future so I've been reading up on the subject.