Megatest speaker cables – 32 cables listened to and measured!

Test setup and explanation measurements

The test setup for measuring speaker cables is largely the same as that of the interlink test. For example, the LCR measurements are similar in setup AND the response measurements and bottom plots are also largely similar, except for the termination impedances. These are 0.3 Ohms on the output (output impedance of the power amplifier) and 8 Ohms on the input (the average speaker has a nominal impedance of 8 Ohms).

You can read about our insertion, explanation and methodology in the article on rca interconnects.

Something different

However, there are a few things we did and had to do differently. It is not possible nor useful to send a 2ns pulse through a loudspeaker cable with the right terminating impedances. 0.3 Ohm on our pulse generator is not possible: there is nothing left of the pulse. So we had to set that up a little differently. It also shows that with such low output impedances and relatively low input impedances, not much remains of such signals anyway.

Now that pulse test is mainly about propagation time and propagation variation. However, it is unfortunate that we cannot measure spectral decay anymore. This turned out to be very interesting in the interlink test. We are going to see how we might be able to do that again differently in order to make this work.

On a speaker

We also did some – electrical! – response measurements on a real speaker. This is much more interesting than on a dummy, since all speaker cables measure perfectly on a dummy, because it is non-inductive and has a dead straight response by itself. A speaker in the real world is not like that. In short: it is much more interesting to use a real speaker.

Also, the speaker forms a circuit with the amplifier – in this case a Bryston PowerPac 300 (damping factor of 500 at 20 Hz) and that makes that there is interaction between the two. How an amplifier reacts to a speaker has to do with, among other things, the damping factor of the amplifier and the impedance variations of the speaker. Remember, however, that the speaker cable is part of this circuit. So we are measuring the entire circuit in this setup.

Now it is important to realize that these measurements will not be a flat line. The speaker is not straight (we are NOT measuring the acoustic response, but the electrical response!) AND the interaction of the amplifier and speaker is also influenced by the impedance and other properties of the cable.

The main thing to look at, is the difference between the speaker cables. You can clearly see in the overviews that each cable has a certain signature. The differences in response are often subtle (with a few exceptions), but noticeable and audible (our opinion).

Phase?

We also compared some speaker cables in terms of phase. To do this, we output a 500 KHz tone via the Rigol function generator on output 1 and 2. The output phase was exactly equalized (with a simple push of a button, what a great option that is!). The output impedance of the generator is 1 Ohm. On the input of the Tektronics scope, we put 8 Ohms.

We can see on the scope that some cables are almost identical, but others clearly show a shift (phase shift).

How this can be linked to propagation time and variation is a good question. That requires more research. Also, we would find it interesting to see how this varies by frequency. Unfortunately, time did not permit that; we have been testing cables for several months now. We have to draw a line somewhere. Perhaps we will pick that up again later.

Noise

Another element we looked at is how much noise the speaker cable picks up. After all: we don’t want noise, right? And every manufacturer gives it their own twist (pun intended).

We measure noise the same way we measure interlinks. So with a connection between the function generator and the scope, without putting a signal active. You are now basically simulating a system that’s sitting idle. We didn’t send a sine wave through it this time, because we end up seeing the most in the pattern without any signal running through the cable anyway.

What we have noticed is that the ‘silence’ is audible in the listening tests and mostly visible in the green line (the actual signal). The blue ‘blur’ doesn’t really appear to be audible…. Probably because it is too random and is so low in level that it is no longer audible to us. At least not in our test system.