A ProAc Studio 200 mod

Trying to smooth an edge...

A very good friend of mine owns a pair of ProAc Studio 200. Although I don’t know very much of their history and character, my immediate impression when I meet them is that they sound harsh - too harsh. High-hats are too bright and sibilance all too present.

Some speakers are known to sound harsh in certain environments. Rather than trying to find out the real cause of this experience we decided to give it a try in my lab. It was immediately obvious from the anechoic SPL response that the speaker has a pronounced peak centered around 12 kHz.

While a peaky treble does not necessarily cause sibilance, I considered it worth-while trying to calm the treble response down by removing the peak.

The impedance of the ProAc Studio 200 is shown in Fig. 1. Both impedance and phase are very well behaved and indicate a simple crossover design. The impedance has its minimum at 6.35 ohms.

The anechoic response is shown in Fig. 2. The microphone is placed 1m in front of the speaker and at the same height as the tweeter center. Obviously the treble peak is a good 6 dB high.

Any attempt at removing the peaky behavior will end up in modifying the crossover filter. Fig. 3 shows the original high-pass filter section (I refrain from showing the component values for copyright reasons). As can be seen, the filter is a simple one and a good example of a high-quality approach. ProAc, right?

The basic topology is a third-order high-pass with an amplifier side resistive attenuation. The inductor is damped by a resistor to lower its Q. The driver side capacitor is nearly twice as large as the amplifier side capacitor, resulting in a slightly sloping tweeter response.

Rather than trying to modify the slope of the response, I thought that adding an LCR trap tuned to the center frequency of the peak and with a Q corresponding to its height, might compensate for the peak. I consider it essential, though, not to affect amplitude or phase response in the crossover region.

Simulation showed that the trap will in fact lower the filter Q, resulting in a little lower output in the crossover region. I tried to correct this by removing the original damping resistor and simulation confirmed it might work. Thus, I implemented it and measured its response.

The modified high-pass filter section is shown in Fig. 4. Inductor L7 is an air-cored coil of 0.074 mH, C6 is 2.2 µF and R8 is 8.2 ohms. While we were at it, we also decided to substitute “audiophile” grade capacitors for the original ones. This may seem unnecessary, but I’ve previously had good experience from such substitutions – even with ProAc speakers.

Fig. 5 shows the measured system impedance response. The red curve shows the original response, and the blue curve shows the response of the modified speaker. Fig. 6 shows the same comparison for the phase response.

What happened to the SPL response? Fig. 7 shows the response of the original speaker (red) and of the modified (blue) one.

As far as I can judge, the addition of the trap causes very little damage to the original impedance, phase and SPL responses. If this conclusion is a valid one, which I really hope it is, then the original crossover performance characteristics should have been pretty well preserved.

This is the more difficult part to report. The first impression is definitely that the speaker has sort of “calmed down”. The original speaker is more pronounced and more demanding than the modified. On the contrary, the modified speaker is a little more relaxed and, at least to my opinion, more musical.

To get a second opinion, I conducted a blind A-B test, first with my wife, then with my daughter. Surprisingly, both of them remarked (and independently so!) that the original speaker was more “flat” than the modified one. The modified speaker was deemed more pleasant and no doubt the preferred one.

Although the speakers still are sibilant on a few of the records I’ve played, many more records are presented in a much more pleasant and enjoyable way. My own conclusion will be that this modification is a careful and successful one.

The real proof will not show, however, until my friend gets his speakers back home and the new capacitors have been given some time to burn in.

Figs. 8, 9 and 10 show the crossover board modification in some more detail.

I still have the speakers in my home. After a week and a half the new capacitors started to open up and the sound I was hoping for is coming out now. The speakers are smooth and articulate, a bit bright to my liking, but the fatiguing harshness is definitely gone.

Imaging impresses me, it is sometimes stunning. It beats my mind how these sharp-edged front panels can let go of the sound so easily, even outside the stereo pair. Playing music with them is a real joy. I thought I wanted to share the news.

Feeling a bit nervous now, I have invited my friend to an acceptance test listening session…

For various reasons I still have the speakers in my home. Believe it or not, I can report that the sound is even smoother now. Sibilance is gone for sure. The brightness I reported above is also gone. In fact, although the speakers have a “masculine” character, they are humble, fair and not the least trying to show off. If you can see what I mean…

I now consider this mod successful.

My friend’s acceptance test was also successful – approved. He has brought his speakers back home and I wish them all the best. Thanks for the acquaintance!

[ previous ] [ back to top ] [ next ]