Hi, I'm building my first set of speakers using a set of Faital 12FH500-4 horns and Faital HF206-6 compression drivers. I am using these components primarily because I got them ridiculously cheap and I've heard goods things about Faital. I plan to use a sealed enclosure.
I've never designed a crossover before, so I'm hoping if folks could chime in on the the attached photo, which shows my proposed design. I want to reduce the horn volume by 6db to start, so I included an L-pad. I also included a zobel on the woofer. The crossover point is 1,000hz.
Thanks for your help.
I've never designed a crossover before, so I'm hoping if folks could chime in on the the attached photo, which shows my proposed design. I want to reduce the horn volume by 6db to start, so I included an L-pad. I also included a zobel on the woofer. The crossover point is 1,000hz.
Thanks for your help.
Uhm, gonna need more info than that to get a xover 2nd opinion.
Did you model the circuit and the frequency response?
Did you model the circuit and the frequency response?
There are many ways to model, ultimately using measurments but if this is your first time you might consider getting the factory responses and impedance data from the manufacturer and turning them into data files to use in a simulator. Example of simulators include Xsim and vituixcad.
The position this puts you in is tuning the crossover network to shape the responses. Keep in mind the effect of the baffle. Once you gain experience you can expand into the other areas of crossover design such as the physical aspects.
The position this puts you in is tuning the crossover network to shape the responses. Keep in mind the effect of the baffle. Once you gain experience you can expand into the other areas of crossover design such as the physical aspects.
I've tried to use Virtuix Cad using the FR and Impedance curves from Faital, but not sure how to interpret this, which includes a modified L-Pad value from what I originally posed.
The results that you have provided indicate that the horn tweeter is running about 5dB too hot. That's not surprising, as the SPL data for the two drivers that you are using indicates a 10dB difference in their sensitivities.
You will need a more complex circuit that includes some equalization of the horn tweeter's response. If you just pad down the tweeter's response a bit more, then you will encounter premature high-frequency roll-off. That's not unusual, as after seeing a JBL design, it seems that compression drivers and their horns require quite a bit of EQ to get to a reasonably flat response in their design passband.
You will need a more complex circuit that includes some equalization of the horn tweeter's response. If you just pad down the tweeter's response a bit more, then you will encounter premature high-frequency roll-off. That's not unusual, as after seeing a JBL design, it seems that compression drivers and their horns require quite a bit of EQ to get to a reasonably flat response in their design passband.
As a first remedy you could try bypassing the series resistor of the L-pad with a small cap or a cap with (small) resistor in series.
Edit: also model the baffle diffraction influence for the woofer response!
Last edited:
I plan to use this Eminence horn from Parts-Express.
[td]Eminence H2EA 2" Cast Aluminum Exponential Horn 60x40[/td]
I've heard that the horn reduces the compression driver by a few dB. Any way to model that in Vituix?
I'll also look into model the baffle diffraction. That might take some research.
| 290-551 |
I've heard that the horn reduces the compression driver by a few dB. Any way to model that in Vituix?
I'll also look into model the baffle diffraction. That might take some research.
I assume that you used the measurement data from Faital. There will be differences between the Eminence (90x40°) and Faital (60x40°).
But as a rough first guide, the data is OK. Attached is a circuit suggestion for the high frequency range.
But as a rough first guide, the data is OK. Attached is a circuit suggestion for the high frequency range.
Yes, used the Faital data. Here is the revised sim based on your revised design above. Thoughts on this as a starting point for my build? I figure that after I have the cabinets built I can test it with this and then adjust values as needed. Does that make sense?
You mentioned that you want to build a closed box. Do you already have an idea of the net volume for the 12HF500?
In any case, you will lose approx. 6dB of SPL in the low pass due to the baffle step, which means that the values for the high pass must be adjusted accordingly. You also need to be careful with the impedance minimum at 200 Hz. It could be that your amplifier does not like the actual 3 ohms in the long term.
Otherwise, build a test cabinet and measure!
In any case, you will lose approx. 6dB of SPL in the low pass due to the baffle step, which means that the values for the high pass must be adjusted accordingly. You also need to be careful with the impedance minimum at 200 Hz. It could be that your amplifier does not like the actual 3 ohms in the long term.
Otherwise, build a test cabinet and measure!
As mentioned, the impedance minimum at 200 hz might be partly adressed by assigning the correct resistance to the large coil 1,27mh. The value in VituixCAD seems a little too low for such a large coil. (I could be wrong here - maybe you have bougth a very expensive coil with low resitance? )
Also, the Z-value needs to be adjusted when the acoustic centers are from different distances in relation to the listener. That is usually the case. In the draft crossover the drivers are both assigned 0 as value.
Also as mentioned you need to adress the baffle step issue. At high frequencies, a loudspeaker tends to project sound only in the forward direction that is, it has a 2 pi hemispherical radiation. At low frequencies, a loudspeaker undergoes a transition by which it begins to radiate sound in all directions, or a 4 pi spherical radiation. Therefore the mid/treble region needs to be further surpressed in relation to the level in the bass . From which freq the mid/treble should be surpressed with approx 6db depends on your baffle width.
Moreover, consider to make the speaker a bass reflex speaker or the bass will roll off downwards from ca 100hz and you will need a subwoofer.
Also, the Z-value needs to be adjusted when the acoustic centers are from different distances in relation to the listener. That is usually the case. In the draft crossover the drivers are both assigned 0 as value.
Also as mentioned you need to adress the baffle step issue. At high frequencies, a loudspeaker tends to project sound only in the forward direction that is, it has a 2 pi hemispherical radiation. At low frequencies, a loudspeaker undergoes a transition by which it begins to radiate sound in all directions, or a 4 pi spherical radiation. Therefore the mid/treble region needs to be further surpressed in relation to the level in the bass . From which freq the mid/treble should be surpressed with approx 6db depends on your baffle width.
Moreover, consider to make the speaker a bass reflex speaker or the bass will roll off downwards from ca 100hz and you will need a subwoofer.
Last edited: