Martin Poulin
Here’s a very good dome tweeter from Transducer lab.
It's the N26CR-T (T stands for the optional titanium voice coil upgrade)
​
http://www.transducerlab.com/index.php?pag=pro&p=p&i=45
The tweeter is very good above 2.5 kHz as seen by it's distortion plot.
​
I did a quick measurement in stock form, with no equalization whatsoever.
Please disregard the accuracy above 15 kHz, as I used an inexpensive USB microphone.
​
I glued* a wave guide salvaged from the ScanSpeak H2606 tweeter**. The waveguide loads very well above 2500 Hz.
​
*with temporary silicone so I can remove easy everything if need be.
**same horn can be taken from Perless H26TG-35 for less money.
​
You can see both set-ups measured at 50cm.
red-stock green is horn no EQ blue is with eq
red-stock green is horn no EQ blue is with eq
Results:
​
-
The SPL gain of the waveguide @ 3 kHz is about 9 dB!!!
-
The SPL gain is still about ~3 dB @ 10 kHz
-
The loading of the horn fell like a rock after 14 kHz, so the EQ must be used to compensate.
​
With a bit of rough equalization, I then tried to match the SPL of the horn loaded version vs the original stock, (always at 50cm distance) and try to maintain my ~95 dB output level.
​
Once equalized, the wave-guide didn’t show much surprise. The distortion dropped significantly from 2.5 kHz to 7 kHz. That was expected as the horn is most efficient in this band. The nasty 3rd harmonic distortion is much lower than without the horn.
​
Above 10 kHz, I had to use a progressive frequency augmentation. The distortion remained similar to stock because the tweeter’s voice coil now sees a similar drive voltage. In other words, the horn loading effect is pretty much gone.
​
Once equalized, the time domain measurements aren't bad.
Whilst it does take longer to stabilize, it remains quite acceptable.
The wave-guide benefits seems to outweigh the drawbacks.
Beryllium tweeter horn loaded:
My Flagship N28BER used here.
http://www.transducerlab.com/index.php?pag=pro&p=p&i=54
​
The horn showed similar improvement as did the ceramic dome version above.
Distortion is much less, and 3rd harmonic is almost gone in the normal pass band (above 3 kHz).
​
If you have DSP capacity to re-equalize:
-
Horn loading a dome tweeters is a total no-brainer.
-
Not sure about other horn models, but the "recycled" one I used from a H2606 tweeter, worked very well.
​
If you're still using a passive crossover, you're in a world of pain, again.
Achieving a flat responses with DIY passive compensation would prove very difficult and expensive.
The extra complexity, phase shift, and insertion loss will probably make things worse for you.
​
Obsolescent passive crossover lovers. Can somebody please explain to me why you're not in active yet? It’s like having a two horse buggy as a means of transportation. Call it "pure" if you want. I call it plain archaic, I use a car.