I have tried the B&C DE7...because all the rumors about compression drivers
I wish I hadn't spent my money that way...
Totally un-hi-fi compared to normal domes...maybe my expectations were based on what I read here; yes they do hit very loud ( for a 10 W device)
but also directivity is a great issue; I should try some other horn shapes and
also the 'directivity EQ' but I really miss the 15-20 KHz range .
I wish I hadn't spent my money that way...
Totally un-hi-fi compared to normal domes...maybe my expectations were based on what I read here; yes they do hit very loud ( for a 10 W device)
but also directivity is a great issue; I should try some other horn shapes and
also the 'directivity EQ' but I really miss the 15-20 KHz range .
What horn did you try with these?
I've head the DE5 one before, and also didn't like it very much, but I suspect this has much more to do with the attached horn, then with the driver itself. But so far I do not know of a CD horn desgined to go low (<1500Hz) that fits on a 0.5" throat...
Pico,
I wonder how much of the loss in the 15Khz and up range is caused by the polyester diaphragms they are using. I would imagine they are somewhat heavy and not all that stiff. What they are calling polyester I am just guessing is what others would call Mylar? It would be interesting if the diaphragm could be replaced with a stiffer material such as aluminum or beryllium how that would affect the top end response.
I wonder how much of the loss in the 15Khz and up range is caused by the polyester diaphragms they are using. I would imagine they are somewhat heavy and not all that stiff. What they are calling polyester I am just guessing is what others would call Mylar? It would be interesting if the diaphragm could be replaced with a stiffer material such as aluminum or beryllium how that would affect the top end response.
Well, we're talking about a particular kind of driver, and I still have to
discover the properties of compression drivers, and I believe I paid too much for a pair of cheap plastic & neo magnet; the horns were 8€ each, so I preferred to make my own
And I tried 2nd order at 3 KHz and 3rd order at 5 KHz. I recently finished an elliptical mouth horn....but still the geometry
might not be optimal...being little in size since the Fc is high; I might give it I try, if I am required to do so...maybe in these days, but in my room I much prefer other type of drivers .
discover the properties of compression drivers, and I believe I paid too much for a pair of cheap plastic & neo magnet; the horns were 8€ each, so I preferred to make my own
And I tried 2nd order at 3 KHz and 3rd order at 5 KHz. I recently finished an elliptical mouth horn....but still the geometrymight not be optimal...being little in size since the Fc is high; I might give it I try, if I am required to do so...maybe in these days, but in my room I much prefer other type of drivers .
At 1 kHz the spread is 10 dB and at 10 kHz it is 20 dB. That is not really Constant Directivity. I have seen worse, but I have seen better as well.
Last edited:
That is the drivers resonance at 2 kHz and at about 1 kHz (compression drivers have two resonances just like a ported loudspeaker does, but there is only one output, not two.) The dip is just the area between these two peaks. All compression drivers do that.
I really do think that 1" is optimum because I do like to cross as low as possible and the DE250 is just about done as low as I go in the Abbey (thanks goodness if for Hi-Fi!). I do not think that a 1/2" throat driver could get down as low as 1200 Hz let alone the 1 kHz in the Abbey.
The DE250 is simply in a "sweet-spot" of compression drivers as I have tried lots and lots of others and none is any better (many are nearly as good or comparable, but cost more - sometimes a LOT more.)
Yes, but the waveguide is only 8 inches in diameter. The polar pattern begins to widen below about 1800 Hz.
By the way, the increasing slope in frequency response at angles greater than the -6 dB coverage angle are seen in every waveguide, so comparing the levels at 0 and 90 degrees is not really useful I'd say.
Last edited:
Yes, I agree. Polar maps are more convenient. The data should be normalized to one axis, although 0 degrees may not be the best to normalize to. Perhaps I'll make a polar map with Arta and post it here.
Perhaps not but it should be adjusted somehow because of the large difference in level between 1 and 20 kHz. I could measure it again with the simulated transfer function from a passive filter I've designed...
Its true that they all show the same data, but for me it is just easier to grasp when it is laid out as a map with color signifying level. If the "brightness" of the color is analogous to level (as it is in my color maps) then the eye sees things just as they are. No need to look at the lines and try and try and figure out if they are parallel or not. It just becomes obvious at first sight exactly what is going on.
In my polar maps the resolution is very high, which is not true of any others that I have seen. I am not sure why but any polar maps done with ARTA always look a lot smoother than when I measure the same speaker. I can't say exactly where that comes from, but I do know that the techniques that I use are unique - they were developed precisely to yield very high resolution polar maps. Lower resolution ones are not as informative.
That is exactly what I would do and do myself. Show it like it will be in actual use. Not some derived version that distorts the reality. What format is the data in? If its something that I could transfer into my program I could do that.
To me the point is to see the data like it will actually be used. Not manipulated in some way that distorts the real picture.
I agree Gedlee that I would much rather see an unsmoothed response most times rather than the 1/3 octave smoothing or more coarse smoothing that is often used. You may see an overall trend but may miss something important that is hidden in a smoothed curve. I have always been of this mind even in most mathematical calculations, I would rather do any rounding in a multi-step equation at the end rather than have multiple rounding errors. Simplicity is not necessarily going to give you accuracy in my opinion.