Matching Power Amps and Loudspeakers.
One of the things that always bothered me is how loudspeakers and power amplifiers are evaluated. My primary concern is the output impedance of the power amplifier and how the variation in loudspeaker system impedance varies with frequency, and how that variation in impedance affects the power amplifier.
It is impossible to tell from any data sheets that loudspeaker manufactures provide what the characteristics of the power amplifier was that was used to set the final adjustments in crossover networks to obtain a reasonably flat response at the listening position. All they usually give you is a ballpark figure of how many watts are required to drive them to reasonable listening levels.
If a amplifier was used during the final stages a a loudspeakers design, and that amplifier had a high fairly high output impedance (low damping factor) the output voltage of the amplifier may vary considerably with changes of loudspeaker impedance versus frequency, yet the loudspeaker provides a flat response on the SPL curve. Thus if the loudspeaker was designed to provide a flat frequency response when driven from such a amplifier it would not be able to provide a flat response from a amplifier with a lower output impedance (high damping factor). This is a fact because the drive voltages to the loudspeaker will vary much less with frequency and thus will cause fairly large peaks in SPL levels at certain frequencies. The lower and upper midrange frequencies would likely be affected most. The reverse is possible if the loudspeaker was designed using a amplifier with a very low output impedance.
It is quite expensive to design a speaker that has a relatively flat impedance curve across its entire frequency range, and in most cases is unobtainable. This is especially true of the low frequency end when ported enclosures are used. The high quality large value capacitors or inductors required eliminate the low frequency rises in impedance caused by the port and woofer resonance’s are very expensive. So most manufactures don’t even bother to attack this problem.
As I stated it can also be quite difficult to find a power amplifier loudspeaker combination that is optimum unless you buy or build a power amplifier that you like and build or modify loudspeaker to be happy with it. This is not always required but the possibility exists. To accomplish this matching can take a considerable amount of time and effort using some of the computer programs designed to do this type of work. Being a DIY does have its advantages if you have the resources to do this.
It nice to always have a power amplifier available that you can use as a reference standard (never get rid of it). Such a amplifier can be used as a fast A-B comparison for checking to see how well your loudspeakers like any new power amplifier that you obtain or build. You will need a pink noise source that can drive each amplifier to the same average output level along with a instant A-B switch to select which amplifier drives the loudspeakers. The one that sounds the smoothest provides the best combination. You may be suppressed at the difference you may notice even though all published specifications of the amplifier match very closely.
A additional factor also comes into play. As the frequency sensitive load provided by the loudspeaker loads the amplifier it changes in the amplifiers internal feedback levels versus frequency. In effect modulating the feedback loop. This causes the output impedance of the amplifier to be modulated as well due to the fact that the amplifiers actual output impedance is greatly affected by the amount of feedback applied. Even amplifiers claiming no feedback that use emitter or source follower output stages are affected.
Warning: If you do the noise test I mentioned with a tube amplifier you should provide a load for the amplifier that is not driving the loudspeakers. Most all solid state amplifiers are stable without a load and do not have to worry about generating high voltages that can arc across tube sockets or inside the output transformers. When a load is removed from a amplifier any negative feedback levels taken from the output will increase tremendously and this may cause some instability in poorly designed circuits. It is possible for a amplifier with such characteristic to self destruct its output devices or other parts..
Does anyone have any suggestions or comments on this issue?
John Fassotte
Alaskan Audio
http://www.audioamps.com
One of the things that always bothered me is how loudspeakers and power amplifiers are evaluated. My primary concern is the output impedance of the power amplifier and how the variation in loudspeaker system impedance varies with frequency, and how that variation in impedance affects the power amplifier.
It is impossible to tell from any data sheets that loudspeaker manufactures provide what the characteristics of the power amplifier was that was used to set the final adjustments in crossover networks to obtain a reasonably flat response at the listening position. All they usually give you is a ballpark figure of how many watts are required to drive them to reasonable listening levels.
If a amplifier was used during the final stages a a loudspeakers design, and that amplifier had a high fairly high output impedance (low damping factor) the output voltage of the amplifier may vary considerably with changes of loudspeaker impedance versus frequency, yet the loudspeaker provides a flat response on the SPL curve. Thus if the loudspeaker was designed to provide a flat frequency response when driven from such a amplifier it would not be able to provide a flat response from a amplifier with a lower output impedance (high damping factor). This is a fact because the drive voltages to the loudspeaker will vary much less with frequency and thus will cause fairly large peaks in SPL levels at certain frequencies. The lower and upper midrange frequencies would likely be affected most. The reverse is possible if the loudspeaker was designed using a amplifier with a very low output impedance.
It is quite expensive to design a speaker that has a relatively flat impedance curve across its entire frequency range, and in most cases is unobtainable. This is especially true of the low frequency end when ported enclosures are used. The high quality large value capacitors or inductors required eliminate the low frequency rises in impedance caused by the port and woofer resonance’s are very expensive. So most manufactures don’t even bother to attack this problem.
As I stated it can also be quite difficult to find a power amplifier loudspeaker combination that is optimum unless you buy or build a power amplifier that you like and build or modify loudspeaker to be happy with it. This is not always required but the possibility exists. To accomplish this matching can take a considerable amount of time and effort using some of the computer programs designed to do this type of work. Being a DIY does have its advantages if you have the resources to do this.
It nice to always have a power amplifier available that you can use as a reference standard (never get rid of it). Such a amplifier can be used as a fast A-B comparison for checking to see how well your loudspeakers like any new power amplifier that you obtain or build. You will need a pink noise source that can drive each amplifier to the same average output level along with a instant A-B switch to select which amplifier drives the loudspeakers. The one that sounds the smoothest provides the best combination. You may be suppressed at the difference you may notice even though all published specifications of the amplifier match very closely.
A additional factor also comes into play. As the frequency sensitive load provided by the loudspeaker loads the amplifier it changes in the amplifiers internal feedback levels versus frequency. In effect modulating the feedback loop. This causes the output impedance of the amplifier to be modulated as well due to the fact that the amplifiers actual output impedance is greatly affected by the amount of feedback applied. Even amplifiers claiming no feedback that use emitter or source follower output stages are affected.
Warning: If you do the noise test I mentioned with a tube amplifier you should provide a load for the amplifier that is not driving the loudspeakers. Most all solid state amplifiers are stable without a load and do not have to worry about generating high voltages that can arc across tube sockets or inside the output transformers. When a load is removed from a amplifier any negative feedback levels taken from the output will increase tremendously and this may cause some instability in poorly designed circuits. It is possible for a amplifier with such characteristic to self destruct its output devices or other parts..
Does anyone have any suggestions or comments on this issue?
John Fassotte
Alaskan Audio
http://www.audioamps.com
Amplifer and Loudspeaker Matching
We seem to focus on what diodes we use in our power supplies, but
often neglect the effect of different loudspeakers and amplifier
combinations.
Listen Test!
That is the K.I.S.S. solution to the problem.
Pick a Philosophy.
Does the engineer back his theory with evidence.
That would be my solution.
We seem to focus on what diodes we use in our power supplies, but
often neglect the effect of different loudspeakers and amplifier
combinations.
Listen Test!
That is the K.I.S.S. solution to the problem.
Pick a Philosophy.
Does the engineer back his theory with evidence.
That would be my solution.
Hi,
when we used to have good audio retailers that offered good service, part of their job was to listen to what they sold.
Their expertise was in combining complementary units that worked well together.
Gradually box shifters took over. They are now as bad as car sales persons. They both know as little about their product as each other.
A few good dealers still exist in the UK. But you have to travel to find them. Us DIYers will probably put them out of business as well.
Come to think of it, in the good old days when HiFi meant High Fidelity, it was ONLY practiced by the DIYer. If you wanted to obtain realism you had to build it yourself and then the industry started to recognise a niche market, came, sold and left again.
when we used to have good audio retailers that offered good service, part of their job was to listen to what they sold.
Their expertise was in combining complementary units that worked well together.
Gradually box shifters took over. They are now as bad as car sales persons. They both know as little about their product as each other.
A few good dealers still exist in the UK. But you have to travel to find them. Us DIYers will probably put them out of business as well.
Come to think of it, in the good old days when HiFi meant High Fidelity, it was ONLY practiced by the DIYer. If you wanted to obtain realism you had to build it yourself and then the industry started to recognise a niche market, came, sold and left again.
DIY to Comercial to DIY
I guess that is why diyAudio.com is so popular. Very few good
Hi-Fi stores left. . . we had five 20 years ago. . .now we have:
http://www.soundscapeav.com/
That is really it. . .the other two Hi-Fi stores are a joke. . .they carry
big name products but only have two listening rooms set up for
the B&W Spiral speakers.
Plus the "box" stores. . .
I guess that is why diyAudio.com is so popular. Very few good
Hi-Fi stores left. . . we had five 20 years ago. . .now we have:
http://www.soundscapeav.com/
That is really it. . .the other two Hi-Fi stores are a joke. . .they carry
big name products but only have two listening rooms set up for
the B&W Spiral speakers.
Plus the "box" stores. . .
There is no way that I am aware of to select the combination of amplifiers + loudspeakers by using "objective" criteria/measurements.
There are two main schools of thought:
- wire is wire, amps are amps, speakers are speakers and if the measured distortion is low, the amp is not clipped, and the frequency response of the speaker is "flat" that is all that matters.
- everything seems to effect the final outcome, amps all sound different, frequency response is not so important, and so on...
Of course there are gradations of the two positions which are merely illustrated above.
Fact is that from an engineering and objective point of view the components (including amp + wires + speakers) forms a complete system which is unique and should be viewed as a system. Therefore there is no single best anything. (of course the room counts too... but that's another story)
Since the entire system represents a compromise that can be analyzed from a variety of points of view, and measured via a variety of parameters, it can be seen that the real question is in deciding which compromises in which areas can be tolerated and then in what ratios or how much?
That's the art part.
The issue you raised about the ouput impedance vs speaker = some frequency response is a valid consideration. However most speakers in the commercial realm are designed with low output Z amplifiers being the intended way that they will be run. And, yes the use of amps with relatively high output impedance (DF ~1) does happen and has been discussed in various threads here on diyAudio, and elsewhere... the effects on frequency response are oft cited (pro and con). Again, my view is that the gear needs to be viewed as a system - and with the understanding that there are many compromises inherently present no matter what.
What really becomes a critical issue is a subjective matter. That is, what or which of these "compromises" and to what degree can they be accepted or tolerated by a given listener. This varies all over the map.
That it does vary all over the map is illustrated by the wide range of "acceptable" sound producing equipment that is sold and used by people. Start with the boombox and kitchen table radio and move through the box store cheep-o rack system up through the mid-fi world into the "HT" (home theater) stuff and across the gamut of "high-end" gear. The commonality to all of it is in the compromises that listeners will accept. There are some of the same compromises found in boom boxes, rack systems, mid-fi and HT as well as in high-end. Perhaps the degree varies, as does the ratios...
In high-end (and maybe in DIY) if we're sophisticated about it, have the necessary expertise & experience, we can pick and choose the type and degree of compromises we will accept.
In the end it comes back to what Count Basie said: "... if it sounds good, it is good..."
_-_-bear
There are two main schools of thought:
- wire is wire, amps are amps, speakers are speakers and if the measured distortion is low, the amp is not clipped, and the frequency response of the speaker is "flat" that is all that matters.
- everything seems to effect the final outcome, amps all sound different, frequency response is not so important, and so on...
Of course there are gradations of the two positions which are merely illustrated above.
Fact is that from an engineering and objective point of view the components (including amp + wires + speakers) forms a complete system which is unique and should be viewed as a system. Therefore there is no single best anything. (of course the room counts too... but that's another story)
Since the entire system represents a compromise that can be analyzed from a variety of points of view, and measured via a variety of parameters, it can be seen that the real question is in deciding which compromises in which areas can be tolerated and then in what ratios or how much?
That's the art part.
The issue you raised about the ouput impedance vs speaker = some frequency response is a valid consideration. However most speakers in the commercial realm are designed with low output Z amplifiers being the intended way that they will be run. And, yes the use of amps with relatively high output impedance (DF ~1) does happen and has been discussed in various threads here on diyAudio, and elsewhere... the effects on frequency response are oft cited (pro and con). Again, my view is that the gear needs to be viewed as a system - and with the understanding that there are many compromises inherently present no matter what.
What really becomes a critical issue is a subjective matter. That is, what or which of these "compromises" and to what degree can they be accepted or tolerated by a given listener. This varies all over the map.
That it does vary all over the map is illustrated by the wide range of "acceptable" sound producing equipment that is sold and used by people. Start with the boombox and kitchen table radio and move through the box store cheep-o rack system up through the mid-fi world into the "HT" (home theater) stuff and across the gamut of "high-end" gear. The commonality to all of it is in the compromises that listeners will accept. There are some of the same compromises found in boom boxes, rack systems, mid-fi and HT as well as in high-end. Perhaps the degree varies, as does the ratios...
In high-end (and maybe in DIY) if we're sophisticated about it, have the necessary expertise & experience, we can pick and choose the type and degree of compromises we will accept.
In the end it comes back to what Count Basie said: "... if it sounds good, it is good..."
_-_-bear- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.