Ok so I mentioned that once some parts came in I was going to attempt to improve the response of my Focal Electra CC900 center channel speaker. One of those, I wonder if my limited know how, limited equipment, and no regard for budget could improve upon a company like Focal's own design. While one could argue that my responses don't reflect anything more than the limitations of my gear, they have all been relatively consistent with other peoples measurements of the same items in the past, and always match my predicted responses pretty well. I'm pretty confident in these results.
Ok this is the before and after response curves of the two speakers. I apologize for the extra smoothing that the one got over the other. I just figured out that if I have any smoothing on, even 1/24th, that it applies 1/6th smoothing to a combined curve. I had no smoothing on the original, and when I made the combined curve, I ended up with no smoothing applied. However, I accidentally left the smoothing on at 1/12th when I took the new ones. While it means that the graph is glossing over some of the finer details of the response, my goal was to fix a 10db trough in the response, not some minor glitches. I think its pretty evident that my new crossover did that. While it still has a terrible off axis response, its still better than it was.
Here is the crossover schematic with new values, and the predicted response based on that.
Ok this is the before and after response curves of the two speakers. I apologize for the extra smoothing that the one got over the other. I just figured out that if I have any smoothing on, even 1/24th, that it applies 1/6th smoothing to a combined curve. I had no smoothing on the original, and when I made the combined curve, I ended up with no smoothing applied. However, I accidentally left the smoothing on at 1/12th when I took the new ones. While it means that the graph is glossing over some of the finer details of the response, my goal was to fix a 10db trough in the response, not some minor glitches. I think its pretty evident that my new crossover did that. While it still has a terrible off axis response, its still better than it was.
Here is the crossover schematic with new values, and the predicted response based on that.
Here is the response over a 45 degree angle. The dark blue is on axis, and the lighter green is 45 degree off axis. Every other one was as even as I could move it, around 9 degrees or so. If this looks totally wrong, it could be, I'm not really sure what I'm doing here. I made a circlular ring for the mic and marked off 4.5 degree markings, with 15, 30, and 45 in bold. The ring fits around the mic stand I'm using so that every time I move the mic to one of the ticks on the circle its moved the mic that many degrees off axis.
No, no, no. That wasn't the point at all. The point was to show what goes into a very expensive high end speaker in the first place, and what can be done with it. To show what factory frequency responses of even high end speakers look like, and again, that you can "improve" upon it.
I don't know, I thought I saw a lot of diy websites where they would writeup modifications they made to a commercial speaker, so i did the same. I also found it very interesting how little the response changed off axis given that this is an mtm design. 45 degrees off axis places the listener at the ends of my couch and then some, and it suggests that there would just be a somewhat smooth roll off of the highs, not the expected uneven lobing.
I'm also using this project as an excuse to learn my measurement suites better so that when I have something more interesting to post, I will be better prepared. I've put money down on the 12" Summa kit from Dr. Geddes, and I am talking with him about purchasing some 15" BB "subs" in his bandpass boxes. I intend to take a full set of measurements for all of this, both full in room (At listening position), and quasi-anechoic. I also hope to give useful distortion measurements, but I'm thinking my particular software is limited with whats built in, and I'm not sure I'm up to doing Geddes suggestion. I suppose we shall see.
-Matt
I don't know, I thought I saw a lot of diy websites where they would writeup modifications they made to a commercial speaker, so i did the same. I also found it very interesting how little the response changed off axis given that this is an mtm design. 45 degrees off axis places the listener at the ends of my couch and then some, and it suggests that there would just be a somewhat smooth roll off of the highs, not the expected uneven lobing.
I'm also using this project as an excuse to learn my measurement suites better so that when I have something more interesting to post, I will be better prepared. I've put money down on the 12" Summa kit from Dr. Geddes, and I am talking with him about purchasing some 15" BB "subs" in his bandpass boxes. I intend to take a full set of measurements for all of this, both full in room (At listening position), and quasi-anechoic. I also hope to give useful distortion measurements, but I'm thinking my particular software is limited with whats built in, and I'm not sure I'm up to doing Geddes suggestion. I suppose we shall see.
-Matt
well I've been trying to listen to everything I could on it. Music, movies, etc. As for Dialog, I mentioned that it had a sort of cupped hand sound to it, making things seem a bit muffled, chesty, etc. I also thought voices were too drawn to the speaker, not disappearing. Ok, well the good is that, it now sounds very neutral, and voices no longer sound like the actors have their hands over their mouths or have a cold. The bad is, it hasn't done much of anything for the tendency of voices to stick to the center. My biggest issue with surround sound and center channels is that center images don't ever seem to hover in the middle like they do with a good set of stereo speakers. Instead, they seem tethered to the center, and that still seems true here. I keep wanting to say its better in that area too, but I wouldn't be willing to wager money on it. Sometimes it seems to float better, and some times it sounds like the actors have mouths on their knees.
As for music, things seem more neutral now, and so instruments take on a more natural timber. That is of course a good thing. However, I still think that it creates a sort of "U" shaped soundstage up front, where instruments dip down to the level of the center (The center tweeter is 6"s below the mains).
The change in timbre as you get more and more off axis sounds consistent with the measurements, which is to say, highs sound like they roll off, but overall, smooth and coherent. Overall I would say it was worth the 30 dollars in parts. Actually, I ended up unwinding the inductors in the original for two of them, and so I really only changed one resistor, 4 capacitors, and modified 2 inductors, and replaced 2 inductors.
I decided to take cabinet resonance measurements as well. I haven't posted them because there seems to be no standard for this, and the type that I can display is something that can only make sense relative to another measurement taken in the same way. In other words, it would be somewhat worthless to post. Anyway, adding some more dampening did very little, but did reduce the overall cabinet energy by about 1-2 db's or so. There were no cabinet resonant modes to speak of, so it seems like a well constructed cabinet.
As for music, things seem more neutral now, and so instruments take on a more natural timber. That is of course a good thing. However, I still think that it creates a sort of "U" shaped soundstage up front, where instruments dip down to the level of the center (The center tweeter is 6"s below the mains).
The change in timbre as you get more and more off axis sounds consistent with the measurements, which is to say, highs sound like they roll off, but overall, smooth and coherent. Overall I would say it was worth the 30 dollars in parts. Actually, I ended up unwinding the inductors in the original for two of them, and so I really only changed one resistor, 4 capacitors, and modified 2 inductors, and replaced 2 inductors.
I decided to take cabinet resonance measurements as well. I haven't posted them because there seems to be no standard for this, and the type that I can display is something that can only make sense relative to another measurement taken in the same way. In other words, it would be somewhat worthless to post. Anyway, adding some more dampening did very little, but did reduce the overall cabinet energy by about 1-2 db's or so. There were no cabinet resonant modes to speak of, so it seems like a well constructed cabinet.
cabinet measurements
would you mind explaining how you obtained your cabinet resonance measurements?
i am interested in doing something similar myself.
thanks
would you mind explaining how you obtained your cabinet resonance measurements?
i am interested in doing something similar myself.
thanks
I use an accelerometer stuck to the side of the cabinet with thin double stick tape. Its part of my ATB PC Pro suite. However, it looks very similar to a mic element hooked to a long 1.8mm stereo cable. I think you could make your own or even buy the ATB one (20 dollars) and use it with any number of software. With my package all I can really do and record is look at the frequency response from 30hz to 3khz or so. I would like the ability to add time too and look at the spectral decay (To see stored energy), but haven't figured out how to do that yet. Anyway, you just attach the accelerometer to the cabinet side, measure just like you would for a normal frequency response measurement, and look for peaks. The peaks imply a resonance, and those are what you are trying to tame. If there is no peak to speak of, then there is really no resonance to speak of, and thus nothing to worry about. If there is a really high resonance for some reason, such as above 1khz, you might find that just acoustic foam or wool batting will work. However below that point you will need bracing and dampening. You don't need to worry about cabinet modes below say 75hz or so because the wavelengths are too long (This is my understand anyway).
- Status
- Not open for further replies.