The TWO34 has its fundamental resonance below 900Hz, so it works well with a first-order crossover if you are setting the nominal crossover frequency to 3000Hz or so.
It WILL need a series L_C_R network across it to compensate for the rise in impedance at resonance. (Otherwise you'll find that your high-pass filter isn't behaving as you expected it to).
I use this tweeter myself with such a crossover with no problems.
(If no-one responds with the resonance compensating network values, I'll check this evening with what I used and post it tomorrow).
Chris Morriss.
It WILL need a series L_C_R network across it to compensate for the rise in impedance at resonance. (Otherwise you'll find that your high-pass filter isn't behaving as you expected it to).
I use this tweeter myself with such a crossover with no problems.
(If no-one responds with the resonance compensating network values, I'll check this evening with what I used and post it tomorrow).
Chris Morriss.
To make a first order crossover work adequately, it would also be wise to place a network across the bass unit to compensate for the rise in impedance as the frequency increases. For the HM170G8, a series R-C network of 8.2R + 12uF is about right.
The resonance compensating network across the tweeter, and the inductance compensating network across the bass unit means that the drive units now present a much more constant impedance load to the crossover.
With these networks in place, you can then calculate the value of the series 'L' for the bass unit filter, and 'C' for the tweeter filter, assuming a constant 8R load. (It won't be perfect, but it will a LOT better than if the speaker impedance had not been compensated for.)
The resonance compensating network across the tweeter, and the inductance compensating network across the bass unit means that the drive units now present a much more constant impedance load to the crossover.
With these networks in place, you can then calculate the value of the series 'L' for the bass unit filter, and 'C' for the tweeter filter, assuming a constant 8R load. (It won't be perfect, but it will a LOT better than if the speaker impedance had not been compensated for.)
Ouroboros said:For the HM170G8, a series R-C network of 8.2R + 12uF is about right.
I used 10ohm + 10 uF zobel on the woofer. And a 18 liter enclouser tuned to 50hz.
for a x-over around 2500hz use 2.0 - 2.2 mH air-core inductor.
See the freq-plot below. In my room measuring <200hz isn't possible the curve is smoothed with 1/6th octave. Used tweeter is a 19mm Scan Speak.
See the freq-plot below. In my room measuring <200hz isn't possible the curve is smoothed with 1/6th octave. Used tweeter is a 19mm Scan Speak.
An externally hosted image should be here but it was not working when we last tested it.
http://pageperso.aol.fr/petoindominique/doc_htm/hm170g8.htm
The rise of the bass unit response above 1kHz indicates it will
work well in combination with a "Baffle Step Compensation"
inductor, giving a final rolloff of 2 to 3 KHz.
(a parallel resistor can be used for fine tuning)
BSC is mandatory for a good tonal balance.
The tweeter is far more more efficient and will need padding
down with series and parallel resistors. The parallel resistor
will suppress the resonance impedance peak and the network
for impedance compensation will probably not be needed.
sreten.
The rise of the bass unit response above 1kHz indicates it will
work well in combination with a "Baffle Step Compensation"
inductor, giving a final rolloff of 2 to 3 KHz.
(a parallel resistor can be used for fine tuning)
BSC is mandatory for a good tonal balance.
The tweeter is far more more efficient and will need padding
down with series and parallel resistors. The parallel resistor
will suppress the resonance impedance peak and the network
for impedance compensation will probably not be needed.
sreten.
Attachments
Hi everybody....I still in project for HM170G8.
I need cutting at 3150hz so i spec down the inductor value...that will no problem with cutting frequency...
the main problem is centre frequency at 1-1.25khz khz ia peak at 3db.someone have schematic in passive x-over for reduce this peak?...all is very linear except this freq...
now i use 6db with zobel in woofer x-over
I need cutting at 3150hz so i spec down the inductor value...that will no problem with cutting frequency...
the main problem is centre frequency at 1-1.25khz khz ia peak at 3db.someone have schematic in passive x-over for reduce this peak?...all is very linear except this freq...
now i use 6db with zobel in woofer x-over
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Loudspeakers
- Multi-Way
- Diy Audax