4 ohm load vs 8 ohm load? What is prefered?


Example; Mcintosh MA6500 integrated amp will output 120 wpc
when using 8ohm speakers. But if I use a 4ohm speaker system it will output 200wpc.
What would the difference be assuming both speaker systems are equal in audio fidelity?

Thanks in advance for any insight.
markeetaux
It seems like the MA6500 will work well with either load, so it shouldn't make much difference whether you use nominal 4-ohm or 8-ohm speakers... as long as the speakers' sensitivity specs are not too different. Given the same efficiency, there will be more available power into 4-ohms, though it's likely that the better speaker (whichever one it is) will still sound better (unless you are clipping the amp at your preferred listening level).
If they are equal in effencey then there be no difference at the same level, the 4 ohm would have a little more power available and play a couple of DB louder. Neither is preferred in itself. A transistor amp will try to double its output when the impedance is halved. Most don't have the capability, the Mac should be producing 240 instead of 200 but that is not bad. So with lesser quality transistors low impedance's should be avoided, even high quality ones can have problems with VERY low ones, say under one ohm. Since tubes usually have transformers to match their impedance they do not operate in this manner. The impedance derives from the particular design of the speaker and neither is preferred in itself, they are a consequence and not a design goal except to make them usable by the amplifiers in their price class. Typically, the more drivers the lower the impedance.
The way I understand it is given the same sensitivity the amp would have to double down into 4 ohms for the 4 ohm speaker to play as loud as the 8 ohm speaker. The MA6500 almost gets there so it would be very close. I doubt if anyone would even notice.
With 4ohm you gain 15.7% of perceived loudness but loose almost 100% of damping factor (120 vs. 230). Choose speakers that sound better to you - forget numbers.

Perceived Loudness = k^(1/3.5) where k is a ratio of power.
Kijanki, I doubt that anyone can pinpoint the difference between a damping factor of 120 vs. 230.
Only poorly designed, very boomy speakers require a damping factor above 20. These are just numbers, as you pointed out.
The OP should decide based on listening tests.
Casouza - I agree. I believe that great specifications are often a contrary indicator of sound quality. I would, for instance, avoid amps with thd=0.00001% at any cost.

Higher df might be of some value since xover inductor in series with the woofer is in order of 0.08 ohm limiting df already to about 100 resulting in total df=50. DF gets much worse with frequency and might get worse for small signals. It is all, most likely, speaker design dependent.
01-12-10: Kijanki
>With 4ohm you gain 15.7% of perceived loudness but loose almost 100% of damping factor (120 vs. 230).

1. FTC amplifier ratings are an over-reaction to the 1970s amplifier spec wars which have little bearing on actual music. A nice jazz recording has 18dB between average and peak power. FTC power ratings are five minutes of sine waves with 3dB crest factors. That's preceded by an hour of "preconditioning" with a 1KHz sine wave with average power 1/3 rated power (prior to 2000) which is actually the hard part (they reduced it to 1/8 power).

When you plan to use the amplifier for music instead of science experiments you should get peaks 3dB higher into a 4Ohm load than an 8Ohm load.

2. The damping factor is only relevant when you solder the output devices to the speaker terminals and use active cross overs. Keeping the active cross-overs but adding 8' of 12 gauge cable will drop the numbers to 83 and 42.

In the 4 Ohm case you'll have at worst a .2dB peak at the bass driver resonance which is generally believed to be inaudible so it doesn't matter.

Of course, with an odd order passive Butterworth electrical filter the damping factor will be just 2.4 at the cross-over point and even order Linkwitz-Riley filters will have a damping factor of 1. Moving into the stop band the damping factor will be less than one.

If you think damping factor is important you need to stop using passive cross-overs instead of worrying about amplifiers.
>01-12-10: Rsfphil
The way I understand it is given the same sensitivity the amp would have to double down into 4 ohms for the 4 ohm speaker to play as loud as the 8 ohm speaker. The MA6500 almost gets there so it would be very close. I doubt if anyone would even notice.

If the speakers had the same _efficiency_ that would be the case although most speakers are rated using _sensitivity_.

A 90dB/2.83V/1 meter speaker will play 90dB measured a meter away with a 2.83V RMS (4V peak sine wave) input regardless of its load impedance.

That will take 1W into 8 Ohms or 2W into 4 Ohms.

A 90dB/1W/1 meter efficient speaker has 90dB/2.83V/1 meter sensitivity with an 8 Ohm nominal impedance and 87dB/2.83V/1 meter sensitivity when the impedance is 4 Ohms.

Sensitivity is the relevant number unless output transformers with different taps get involved, since music which doesn't make your ears hurt from the volume has average power so far below the peaks you're just looking to keep those from clipping.
01-13-10: Drew_eckhardt
>When you plan to use the amplifier for music instead of science experiments you should get peaks 3dB higher into a 4Ohm load than an 8Ohm load.

Assuming equal efficiency.

I think the big thing is to not worry as long as you don't drop below the amplifer's rated impedance (some can become unstable and oscillate).

With competent design and high phase angle exceeding the limits might get you thermal protection (wait for the amp to cool off) or blow rail fuses (a bit of a pain when there are no external fuse holders and there are a dozen screws holding the amp cover on).
"should get peaks 3dB higher into a 4ohm load than an 8ohm load."

- 15.7% louder is easier to imagine than 3dB.

"The damping factor is only relevant when you solder the output devices to the speaker terminals and use active cross overs"

- I don't know where you getting this from.

"Of course, with an odd order passive Butterworth electrical filter the damping factor will be just 2.4 at the cross-over point and even order Linkwitz-Riley filters will have a damping factor of 1"

- we are not talking about filter's "alpha" but about damping factor defined as ratio of speaker impedance to amplifier's output impedance and yes it is important up to some point. I'm trying to show that it is limited for woofer by resistance of inductor in series to about 100 and would be nice not to lower it even more.