Notes from auditioning speakers

[just listening]

Kurt:
I applaud your effort in picking up two of the same receiver to run such a quick switching between speakers. Most importantly, conducting the comparison in your own home, kudos! Too bad you couldn't do the same with the Rainmakers vs the Studio 20's, unless you can spring for a pair of used Rainmakers off of Audiogon. That way you'd only be out shipping as their used price stays very consistent from what I've noticed.

 

[R-Carpenter]

http://www.audioadvisor.com/prodinfo.asp?number=BYSWBO
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=300-970
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=300-974

Here's some selectors for people that run comparisons at home. Impedance matching will make compensation for speaker sensitivity easy. Every reviewer should use one in a short term comparison with SPL of different speakers matched.

 

[cschang]

http://www.audioadvisor.com/prodinfo.asp?number=BYSWBO
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=300-970
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=300-974

Here's some selectors for people that run comparisons at home. Impedance matching will make compensation for speaker sensitivity easy. Every reviewer should use one in a short term comparison with SPL of different speakers matched.

These add series resistance to the speaker can potentially change the characteristics of the speakers. It changes the characteristics of the speaker's crossover and impedance.

For comparisons, the level matching should be done before the amplifier, not after.

These devices are meant to protect the load on the amplifier when running multiple speakers simultaneausly.

 

[R-Carpenter]

I don't see how you can change a character of the speaker by adding a series resistance before the crossover. What do you think most of the volume controls are? Of cause if the speaker has a very low impedance deeps, series resistance can make a over all impedance higher and make the amplifiers job easier, changing amp/speaker interaction. But if you are testing with a good solid state amp like Bryston or Adcom, it shouldn't mater so much.
Ideally the output should be changed at the preamp but you are running in to the series resistance again, just at a different stage and by supplying less voltage to the speaker to match absolute SPL you also potentially changing the sound.

 

[cschang]

I don't see how you can change a character of the speaker by adding a series resistance before the crossover. What do you think most of the volume controls are? Of cause if the speaker has a very low impedance deeps, series resistance can make a over all impedance higher and make the amplifiers job easier, changing amp/speaker interaction. But if you are testing with a good solid state amp like Bryston or Adcom, it shouldn't mater so much.
Ideally the output should be changed at the preamp but you are running in to the series resistance again, just at a different stage and by supplying less voltage to the speaker to match absolute SPL you also potentially changing the sound.

The volume control on a receiver works before the amplifier stage. The actual gain on the amp stage is constant. Think of running seperates....think of the pre-amp and the amp. There are no volume controls on the amp..

If what you are stating is correct, why wouldn't you suggest using one of these devices change a speaker with a 4 ohm rating to an 8 ohm rating, to make it easier for the amplifier to drive? Better yet, why wouldn't a manufacturer add series resistance to change the rating? Simple, because doing so changes the crossover characteristics.

I am not an EE, so I can not get into the technical details, but talk to any speaker designer/engineer/EE, and they can tell you the whole story.

 

[R-Carpenter]

You can change a 4 ohm speaker to be a 8ohm speaker by placing a 4 ohm resistor between the speaker and the amp. You will effectively decrease the sensitivity of the loudspeaker. The reason why there are 4 ohm and 8 ohm speakers out there is based on the driver, driver compliment and intended purpose of a loudspeaker (which also has to do with efficiency). The 4 ohm or 8 ohm rating is also averaged. As I am sure you are aware, a speaker impedance is far from straight line. A series resistance in front of the crossover will not change the crossover and speaker itself, but will make less current run to the speaker. You can simply turn the volume control knob higher to counter act. It would effectively make the amp work
harder.
Here's an example. I have a jig build for Sound Easy FR, Impedance and other measurements. It has a series 10ohm resistor for calibration. Once in a blue moon I forget to flip the switch and take FR of a loudspeaker. It remains absolutely the same except lower.

As to the manufacturers. Well, if you are designing a speaker with driver A, B and C and the desired combination comes out as an averaged 8ohm speaker with 89db sensitivity, it is more a question where you've reached your goal.

To the tube amp experts: what exactly is a 4/8ohm switch on old tube amplifiers?

I don't mean to hijack the thread so I'll shut up now.

 

[cschang]

You just answered the question.

The speaker's impedance is not a straight line. The speaker is reactive load. By changing the impedance with a series resister, you are changing the reaction the amplifier sees.

I think if you do a google, you will find examples of the issue.

 

[fredk]

I have only seen one instance where someone measured the output from a speaker with and without a switch in the circuit. An overlay showed identical output except a very narrow band where there was a 1-2db difference. In other words, identical output with and without a switch.

I believe this was a post on avsforum.

 

[cschang]

I have only seen one instance where someone measured the output from a speaker with and without a switch in the circuit. An overlay showed identical output except a very narrow band where there was a 1-2db difference. In other words, identical output with and without a switch.

I believe this was a post on avsforum.

Was it a switch or an attenuation device?

 

[R-Carpenter]

I'll take a near field of the loudspeaker I am working on with and without resistor inline and post it. If FR is identical and only differs in the magnitude then there could be no audible difference. I use non inductive resistors, usually Eagle but can use Xicon, there's no difference.
C, I think you may be mistaking attenuator with L-Pad and a tweeter where L-pad does not work the same way as a series resistor.

 

[cschang]

I'll take a near field of the loudspeaker I am working on with and without resistor inline and post it. If FR is identical and only differs in the magnitude then there could be no audible difference. I use non inductive resistors, usually Eagle but can use Xicon, there's no difference.
C, I think you may be mistaking attenuator with L-Pad and a tweeter where L-pad does not work the same way as a series resistor.

Maybe so...but now that I think of it, I believe a variable L-Pad is what you want to do.

Also, the affects of series resistance will vary per speaker, and of course, the more resistance, the potential differences. Trying one speaker won't prove anything one way or the other. Also make sure you measure the speaker at different levels without the resistance.

Here is something that came up on a quick search:
http://amps.zugster.net/articles/attenuation

 

[R-Carpenter]

The FR of the loudspeaker does not vary with the SPL unless you are getting to the loudspeaker or the amplifier limits. There are minor effects of the voice coil inductance and it's change with the temperature increase. I spend 10 to 15 hours a week measuring loudspeakers and I can tell you now that changes in the absolute SPL have very little effect on the frequency response of the loudspeaker given that it operates in it's intended range. The question really is, can purely resistive matching be used to SPL matching for testing of different loudspeakers or is it going to have any effect on the FR of a loudspeaker? I don't see why measuring 1 loudspeaker is not going to prove a point. It has a crossover, a pair of drivers, so why not?

 

[KEW]

Kurt:
I applaud your effort in picking up two of the same receiver to run such a quick switching between speakers. Most importantly, conducting the comparison in your own home, kudos! Too bad you couldn't do the same with the Rainmakers vs the Studio 20's, unless you can spring for a pair of used Rainmakers off of Audiogon. That way you'd only be out shipping as their used price stays very consistent from what I've noticed.

I am definitely giving thought to the Rainmaker audition. As I said, I do not believe these speakers were broke in (I would guess they had less than 2 hours on them). I don't know if this really matters, or if it matters more on certain speakers than others, but the uncertainty is troubling.
I think I'll be checking with the Rainmaker shop and see if they will let me take them for a day (or try Audiogon).

 

[KEW]

http://www.audioadvisor.com/prodinfo.asp?number=BYSWBO
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=300-970
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=300-974

Here's some selectors for people that run comparisons at home. Impedance matching will make compensation for speaker sensitivity easy. Every reviewer should use one in a short term comparison with SPL of different speakers matched.

Since each speaker had it's own dedicated receiver, matching SPL was no problem. Maybe I'm missing something?

 

[cschang]

The FR of the loudspeaker does not vary with the SPL

That is not true, look at the measurements at Soundstage. On the more recent measurements, they show the deviation in FR at different SPL.

Look at chart number 4 here:
http://www.soundstagenetwork.com/measurements/speakers/bw_cm1/

Also, the more resistance you need to add to match the speaker, the more potential to affect that crossover. Not all crossovers are the same, not all speakers are the same, they use different drivers, etc. Testing one to prove your theory, only proves one thing, that your theory holds true or not with that particular speaker. Now measure a dozen, and we are getting somewhere.

 

[cschang]

Since each speaker had it's own dedicated receiver, matching SPL was no problem. Maybe I'm missing something?

RC is saying that using one of those devices will accomplish the same thing you did. I am saying that is not true.

What you did is a much better way of comparing.

 

[R-Carpenter]

Since each speaker had it's own dedicated receiver, matching SPL was no problem. Maybe I'm missing something?

Not everyone has dedicated dozen of receivers, I don't so I level match. It was just a suggestion.
I am saying it's a cheaper way of achieving the same goal.

C, I don't know what loudspeaker they measured and if they got to x-mas limits. The purpose of this test is show deviations with the increased stress. 70/90 db is strange comparison. Anyway, lets get to a resistor between the loudspeaker and the amplifier. Even if a stressed loudspeaker shows very slight fluctuations ( The stress may be mechanical, thermal or otherwise. Meaning? it could be anything from a weak sidewalls to under powered smoking resistors) in FR, it should show same fluctuations with the resistor inline.
C, you are failing a simple test even before it takes place. Are you even interested to look at the data?

 

[cschang]

Here is another read regarding resistance, in this case it is in the speaker wire and what it can affect:
http://www.roger-russell.com/wire.htm

It is about 1/4 down that page.

"As the wire resistance increases, it becomes significant compared to the speaker impedance. It will affect the areas of lower impedance values first and eventually will be audible. Speakers with small impedance variations versus frequency, and that don't dip below the nominal impedance, will be more tolerant of higher resistance in the speaker wire. On the other hand, speakers with large variations in impedance that dip below the nominal value will be much easier to notice. If the speaker has constant impedance versus frequency, the only change will be reduced output."

This resistance in the wire is much less than what you are adding with a resistor. Both are inline series resistance.

 

[cschang]

C, I don't know what loudspeaker they measured and if they got to x-mas limits. The purpose of this test is show deviations with the increased stress. 70/90 db is strange comparison. Anyway, lets get to a resistor between the loudspeaker and the amplifier. Even if a stressed loudspeaker shows very slight fluctuations ( The stress may be mechanical, thermal or otherwise. Meaning? it could be anything from a weak sidewalls to under powered smoking resistors) in FR, it should show same fluctuations with the resistor inline.
C, you are failing a simple test even before it takes place. Are you even interested to look at the data?

Come on RC...the name of the speaker is on the very top of the page...B&W CM1. You can discount what I am saying, but at least read the info I am posting.

The bottomline is you said FR doesn't change with SPL, and I said it does.

Your test isn't really going to prove much one way or the other. With a quick search, I have given you a couple of things to read.

 

[R-Carpenter]

Yeeeahh, I mentioned operating range and not going over x-mas. Speaker wire has inductance and capacitance if you want to be scientific about it. So, the effect is different in comparison to a resistor that has significantly higher resistance (hence the name) and minuscule capacitance. Even given this facts, the audibility of the speaker wire differences has been extensively debated and found non existent in DBT.
Again, come back to my point. Can a simple resistor be used to level match? I think it can, you think it can't.