Sidebars

Doing ABX tests yourself

You don’t have to have an ABX switcher or TEF analyzer to do these tests, just a friend and a lot of patience. You also need a good analog meter that you can use to measure the output of the amplifier(s), and at least a homemade double-pole double-throw switch you can use to switch the stereo speakers between two different amps. Take caution with tube amps doing this. Some of them become very unhappy when driven with no speaker connected.

	Flow chart for simple ABX test system.

To match levels, observe the level of a 1 kHz. tone with the meter leads attached to the output of Amp A (in parallel with the speakers). Now do the same thing with Amp B, trimming the input levels for Amp B to yield exactly the same levels you got with Amp A (try for .1 dB). Force be with you.

Have your friend flip a coin ten times to determine whether X is going to be A or B for each of ten trials. Then have him/her run the switch for you, playing A and B at your request and then X. Write down your answer before your friend tells you which X is. Your friend should be out of sight behind you, and he/she should be absolutely quiet and still while you are doing this. You will find you become fiercely competitive at this, and absolutely sure you can do it better, and you will start keying off noises, facial expressions, anything you can to get the right answer. This is fatiguing and difficult work. Take a break after each set of ten. Quit after 100 trials, because by that point your brain will have acquired the consistency of overripe zucchini Jell-O.

About Cables

Now, I know there are going to be questions about this. For the record, I used 16-gauge lamp cord with soldered ends for this test, plus normal patch cords for the line level signals.

At the same time, I know that there is a sizable community of folks (some of them are close friends and people that I highly respect) that believe that differences between speaker cables are both audible and significant. Therefore, before you get out your poison pen to write a long mean-spriited letter to Nick cancelling your subscription, please note the following: I acknowledge that it may be that the use of exotic cabling might change the audibility results of the tests that I ran. However, keep in mind that all amps had the same cabling which was generic stuff in broad general use, thus satisfying the Range Rule. It is reasonable to view my test method as generally valid and the use of exotic cabling as a special case.

At some point in the future I hope to do ABX testing of cables. In the meantime, if you use 16-gauge lamp cord for speaker cable, you can safely anticipate that amps will sound pretty much the same. If you use 12-gauge oxygen-free counter-wound copper with braided surface or some other hi-tech interconnect, I suspect amps will probably still sound pretty much the same, but I don’t know that for sure.

About speaker sensitivity

This is a big one. Speakers have widely differing sensitivities, usually expressed as the Sound Pressure Level, in dB, generated at 1 meter from the speaker on axis, when an audio signal of 2.83 volts RMS is applied to the speaker. Very small, inefficient bookshelf speakers have a sensitivity where 2.83 volts RMS yields 82 dB SPL or thereabouts. Normal medium hi-fi speakers tend to have a sensitivity of between 88 and 93 dB SPL per 2.83 volts RMS input. Sound reinforcement systems can have outputs of over 100 dB SPL per 2.83 volts RMS input. Why 2.83 volts RMS? Because, for an 8-ohm speaker, 2.83 volts RMS will yield 1 Watt of power, which makes for easy arithmetic.

This leads us to an engineering sort of solution. If you wish to monitor at 90 dB SPL, with 20 dB headroom for those troublesome moments of kick drum passion, then you need a system that doesn’t run out of sauce until 110 dB SPL. If your speaker has an efficiency of 88 dB SPL per 2.83 volts RMS (equivalent to 1 Watt, sort of), and you have two of ‘em (for stereo, remember) and you listen to them from three feet away, you will need 100 Watts per channel (20 dB is 100 times the power), which will give you 108 dB SPL at each speaker at the onset of clipping. You get to add three dB for having two such speakers (i.e. double the power), which you means you’ll end up with 111 dB SPL at the onset of clipping. If you get greedy, and say you’d like to be able to actually listen at 100 dB at three feet with 20 dB of headroom, then you are going to need some serious power: 1 kilowatt per channel!

FYI, my custom reference speakers have an efficiency of about 87 dB per Watt, and with the Bryston’s 250 watts, clipping begins on transients of about 114 dB SPL. I find I can get up to sustained levels of about 95 dB before clipping becomes a problem at all, and at 100 dB it begins to be troublesome on some music. This is where low-powered amps fall apart, of course. 20 dB of headroom for a 35 Watt amplifier means that its nominal level should be at .35 watts, or about 82 dB SPL for my speakers. At that level a 35 Watt amplifier will do fine, and my experience with the Crown confirms this. However, by the time I get to 90 dB SPL, the Crown begins to overload a lot, depending on the recording.

Please note that clipping (harmonic distortion) doesn’t just sound bad in these applications. It can be seriously destructive to tweeters, because clipping represents power being shifted to higher frequencies (including ultrasonics) because it no longer can be applied to a fundamental frequency. And at 4 AM, when you’re mixing heavy metal, it may be difficult to figure out whether that clipping you’re listening to comes from the tracks or the amp/speaker until you see little puffs of smoke coming from the tweeters and things begin to sound an awful lot duller.

So, to my way of thinking, the confidence of knowing that the amp is never going to be the problem is significant. Therefore, I believe in having more power than I’m ever really going to need. My reference level these days has been reset to 85 dB SPL (this was in 1994), which means I have about 30 dB of headroom. Neither the amp nor the speakers ever are stressed at these levels. So, if I hear a distortion problem, I know it’s someplace earlier in the signal chain. ‘Nuff said.

About Speaker Impedance

The problem of speaker impedance adds complication, confusion and mayhem to the above. Most amplifiers are rated for the power they deliver into an 8 ohm load, with other power expressions for 4 and 2 ohm loads, as well as for bridged mono operation. These latter expressions tell you something about the current handling capacity of the amplifier, and the closer to linear proportions (i.e. the power at 2 ohms is 4 times the power at 8 ohms) the better the amp is at providing large amounts of current, which is generally desirable if you want to work at sustained high power levels.

However, loudspeakers (a) don’t usually have exactly an 8-ohm nominal impedance, and (b) speaker impedance varies as a function of frequency, so that when you send pink noise (f’rinstance) to the speaker, the amp is trying to maintain equal acoustic output at all frequencies regardless of the power required by the speaker at each frequency. So, when you do your power calculations, you have to add significant fudge factors to account for the variabilities of impedance. It is not unreasonable to assume a nominal impedance of six ohms when the speaker is rated as an eight-ohm speaker, but keep in mind that at some (usually low) frequencies it may drop as low as two ohms, meaning that the amp must put out three times the power at that frequency to maintain the same level as it does at frequencies where the speaker is at six ohms. Note that you cannot directly measure the impedance of the speaker without specialized equipment. If you try to measure a speaker with a Volt-Ohm-Multimeter (VOM), the resistance you see will be DC resistance only, and not relevant at audio frequencies.

About Damping

Damping is the ratio of source impedance of the amplifier to the actual impedance of the loudspeaker. In fact, it is an expression of how effectively the amplifier will control the electrical “inertia” and feedback of the speaker, sort of like a shock absorber does on a car. The greater the damping, the less the speaker itself will effect the audio signal as it changes over time.

This is why we believe that the difference we heard in reverberance with the Yamaha was due to its lower specified damping factor. If our speculation is right, the most significant indicator of comparative sound quality in a modern amplifier, within the power limits of the amp, may be its damping factor. Therefore, damping may be a particularly useful spec to consider, and a higher damping factor is better.

Post a Comment