Good morning everyone, how long?
I have reached the final steps of building my 3-way box and would now like to check the speaker attenuation part.
I thought a lot about using L-pads to attenuate my tweeter and midrange.
Tweeter 91.2 dB 4 ohms
Midrange 88 dB 4 ohms
Woofer 87 dB 4 ohms
I obviously want to match all the sensitivities with that of my woofer, resulting in an attenuation of 4.2 dB for the tweeter and 1 dB for the midrange. I did the calculations by hand and also used online calculators. The results always come out very similar for the resistor values, although these values are extremely difficult to find and I don't even know if this is correct...
Is there any option other than the L-pads? Could I use just one resistor for the tweeter? I saw this in some places in my country, but I don't know if it's the best option.
I have reached the final steps of building my 3-way box and would now like to check the speaker attenuation part.
I thought a lot about using L-pads to attenuate my tweeter and midrange.
Tweeter 91.2 dB 4 ohms
Midrange 88 dB 4 ohms
Woofer 87 dB 4 ohms
I obviously want to match all the sensitivities with that of my woofer, resulting in an attenuation of 4.2 dB for the tweeter and 1 dB for the midrange. I did the calculations by hand and also used online calculators. The results always come out very similar for the resistor values, although these values are extremely difficult to find and I don't even know if this is correct...
Is there any option other than the L-pads? Could I use just one resistor for the tweeter? I saw this in some places in my country, but I don't know if it's the best option.
You can also create custom resistor networks by combining standard resistor values in series or parallel to achieve the required resistance.
Are the dB values those stated by the manufacturer(s), or measured in your cabinet?
I ask because some manufacturers' data may not be accurate.
Further, when mounted in a cabinet, the woofer output will drop due to 'Baffle Step Compensation', so it may not be 87dB - and if it isn't, the tweeter and mid will need different levels of attenuation.
As to the differences in output, I must admit that I can't hear 1 dB difference but others may well be able to do so.
Geoff
I ask because some manufacturers' data may not be accurate.
Further, when mounted in a cabinet, the woofer output will drop due to 'Baffle Step Compensation', so it may not be 87dB - and if it isn't, the tweeter and mid will need different levels of attenuation.
As to the differences in output, I must admit that I can't hear 1 dB difference but others may well be able to do so.
Geoff
Hi Pedroga,
You are here considering the nominal speaker impedance, right ? But your speaker may have a different impedance at the transition frequencies : it would be worth to check, who knows ? A notable variation in resulting impedance (more than 0.25R to 0.5R) could impact the performance of your crossover.
I will take an example : on one of my 3-ways enclosure, I use a 8" midrange BEYMA 8M60N.
This speaker is given as a 8 ohms nominal impedance.
I use it in the 300Hz to 4500Hz range.
The impedance varies from 8R @300Hz to 14R @4500Hz.
With a 8-points average impedance calculation, I have 12.54R on the considered band of frequency.
That's 4.54 ohms more than the nominal 8R value...
So I took the 12.54R averaged value to calculate my attenuator RS and RP resistor values.
Then I calculated the equivalent resistance of the speaker plus its attenuator at the two transisiton frequencies, taking the impedance value for each transition frequency to calculate the HPF and LPF components of the midrange pass-band section of the crossover.
For the tweeter and the woofer, I don't have to average the impedance, since there's only one transition frequency for each, so I take the impedance value at the matching transition frequency for calculating the attenuator (if needed) and the HPF and LPF crossover sections.
That's what I do for manual calculations to minimize gaps...
But it's me, OK ? 😉
T
I measured all the values with the crossover itself installed on them, so they are already separated by frequency range