I'm studying winISD and I noticed that when I compare two drivers in the same project in winISD in the sound pressure tab (SPL), the sound pressure disparity is reasonably large and I would like to know how to improve this... I looked here on the forum and some websites that the use of resistors to reduce the power that would go to the driver in order to equalize the volume between them is totally reprehensible as it considerably ruins the sound quality, another solution that I was reading about was crossovers (which is the one I think about most). I was thinking about the work that I could also use to use equalizers to send more bass signal, for example, to the subwoofer or even reduce the signals from the midranger and tweeter, but thinking about it I think that wouldn't be a very good solution.
This would not apply to attenuating a tweeter.
A tweeter is designed with a high sensitivity so that it can be attenuated by a series resistor to match the lower sensitivity of a woofer.
This is standard practice to match up the SPLs, and the resistor does not degrade the sound quality of the tweeter.
What you wouldn't do is to attenuate a woofer since a series resistor, apart from wasting a huge amount of power, would affect the audible performance of the woofer.
What about the midrange? How would you match the woofer? I was thinking and I think it would be better to see a woofer with a higher sensitivity
If a crossover schematic has already been decided upon then, in order not to upset the crossover function, the tweeter and the midrange should each be attenuated using two resistors in an L pad arrangement.
An L pad arrangement of series and parallel resistors is shown in front of both the tweeter and the midrange in the following schematic:
However, what you need to know is how to use software to design an integrated crossover/attenuation circuit from scratch.
Over to the speaker design experts!
An L pad arrangement of series and parallel resistors is shown in front of both the tweeter and the midrange in the following schematic:
However, what you need to know is how to use software to design an integrated crossover/attenuation circuit from scratch.
Over to the speaker design experts!
Back in my days of a passive 5 way horn system I'd made, I used L-Pads on all the channels above the Tapped Horns (least efficient at 105dB/W).
I started out putting the drivers particulars into a simple online L-Pad calculator and then asked it for attenuation value Horn resp gave for the efficiency of the mid bass horn (110dB/W).
This got me ball park, and then I was able to tune by ear a dB up or down here and there.
I used 5 and 10W speaker resistors.
Being an efficient system the power handling of these was never a problem - might require beefier ones in radiating non horn speaker system.
For the mid horns, upper mid horn and horn tweeter, I used the manufacturers relative quoted efficiencies on standard horns.
Again this got me in the ball park, and then I tweaked to taste.
With a selection of resistor values, I found I had all the variants I required.
I can't remember which online calculator I used.
Something like this though
https://diy-fever.com/misc/l-pad/
The exact driver impedances didn't need to be spot (professionally measured) as I was adjusting by ear anyway.
I started out putting the drivers particulars into a simple online L-Pad calculator and then asked it for attenuation value Horn resp gave for the efficiency of the mid bass horn (110dB/W).
This got me ball park, and then I was able to tune by ear a dB up or down here and there.
I used 5 and 10W speaker resistors.
Being an efficient system the power handling of these was never a problem - might require beefier ones in radiating non horn speaker system.
For the mid horns, upper mid horn and horn tweeter, I used the manufacturers relative quoted efficiencies on standard horns.
Again this got me in the ball park, and then I tweaked to taste.
With a selection of resistor values, I found I had all the variants I required.
I can't remember which online calculator I used.
Something like this though
https://diy-fever.com/misc/l-pad/
The exact driver impedances didn't need to be spot (professionally measured) as I was adjusting by ear anyway.
That sounds like LLB (lots and lost of b---) as resistors are used in almost every speaker crossover, whether DIY or commercial, I've seen or read about. They're used in different ways in L-pads, Zobel filters and to contour driver responses.
The only exception which springs to mind is Pearl Acoustics "Sibelius", which uses a single, well chosen driver. That might apply to some other single driver designs too.
As requested by AllenB, a link(s) would be useful so we can see the reasoning behind that statement
Geoff
The only exception which springs to mind is Pearl Acoustics "Sibelius", which uses a single, well chosen driver. That might apply to some other single driver designs too.
As requested by AllenB, a link(s) would be useful so we can see the reasoning behind that statement
Geoff
padding a tweeter is pretty normal.
Dont care for magic " in the signal path" nonsense.
No need to pad midrange with driver choice.
Easy example is 12" woofer has way more senstivity
than a 3" around 6 dB or more
So easy to combine with baffle step losses.
WinIsd is just for theoretical transfer function of filter.
Not crossover design.
The woofer alignment shown is not a alignment
Pretty common for woofer to be 6 dB higher
than the midrange. Since your up against baffle step.
And expect to deal with up to 6 dB of losses.
Might as well start with drivers that have FRD files.
So you can get idea how speakers actually sum.
Winisd is just a quick reference to find box volume.
You hit .5 to .7 Qtc and there you go.
Nothing more to piddle around with.
Summing driver on and off axis
requires crossover designers software
and the baffle gets simulated in full space.
Which will include off axis information.
Winisd assumes Half Space.
You model in full space since the baffle provides
half space loading. And need to be aware of the baffle step
or stepped response that will happen.
VirtuixCad
Dont care for magic " in the signal path" nonsense.
No need to pad midrange with driver choice.
Easy example is 12" woofer has way more senstivity
than a 3" around 6 dB or more
So easy to combine with baffle step losses.
WinIsd is just for theoretical transfer function of filter.
Not crossover design.
The woofer alignment shown is not a alignment
Pretty common for woofer to be 6 dB higher
than the midrange. Since your up against baffle step.
And expect to deal with up to 6 dB of losses.
Might as well start with drivers that have FRD files.
So you can get idea how speakers actually sum.
Winisd is just a quick reference to find box volume.
You hit .5 to .7 Qtc and there you go.
Nothing more to piddle around with.
Summing driver on and off axis
requires crossover designers software
and the baffle gets simulated in full space.
Which will include off axis information.
Winisd assumes Half Space.
You model in full space since the baffle provides
half space loading. And need to be aware of the baffle step
or stepped response that will happen.
VirtuixCad
Last edited:
This is the site I read about
https://www.tnt-audio.com/clinica/spl_matching_e.html
Wouldn't that really take away any quality from the sound?
Nope... padding is extremely normal
and with more experience.
Proper dual resistors used as voltage divider
can help in flatten the response/ impedance curve.
Quality" is the driver itself and not crossing to low
and distorting the tweeter by doing so at high levels
Can be applied to the fantasy marketing
of using huge over priced 400 watt or 600 watt
inductors for 30 to 60 watt speakers.
That dont see more than 20 watts at home.
But they are big and expensive and are " upgrades"
when a speaker is what a speaker is.
And a 600 watt ball of copper wont fix it.
People like to imagine problems in audio.
Then imaginary problems that don't exist.
Become problems.
But then they mount a tweeter dead center on
the baffle with way to much center to center spacing.
And the " real" problems get ignored
Sure a 3 ohm resistor is so so horrible.
But they completely mis the horrible 30 ohm peak
they make at crossovers points.
Even with computer aided design and cheats.
You see garbage 30 ohm peaks at crossovers.
But a 3 ohm padding resistor is " bad"
Keep in mind the 400 dollar coil with .1 ohms
is way more important than turd crosssover design.
= Nope
and with more experience.
Proper dual resistors used as voltage divider
can help in flatten the response/ impedance curve.
Quality" is the driver itself and not crossing to low
and distorting the tweeter by doing so at high levels
Can be applied to the fantasy marketing
of using huge over priced 400 watt or 600 watt
inductors for 30 to 60 watt speakers.
That dont see more than 20 watts at home.
But they are big and expensive and are " upgrades"
when a speaker is what a speaker is.
And a 600 watt ball of copper wont fix it.
People like to imagine problems in audio.
Then imaginary problems that don't exist.
Become problems.
But then they mount a tweeter dead center on
the baffle with way to much center to center spacing.
And the " real" problems get ignored
Sure a 3 ohm resistor is so so horrible.
But they completely mis the horrible 30 ohm peak
they make at crossovers points.
Even with computer aided design and cheats.
You see garbage 30 ohm peaks at crossovers.
But a 3 ohm padding resistor is " bad"
Keep in mind the 400 dollar coil with .1 ohms
is way more important than turd crosssover design.
= Nope
The idea of using the aliexpress tweeter has already left my head... I'm going to invest a little more in them and buy something much better and with more "linear" responses
I had already read about resistors to make my projects, but they never had to do with sound.If a crossover schematic has already been decided upon then, in order not to upset the crossover function, the tweeter and the midrange should each be attenuated using two resistors in an L pad arrangement.
An L pad arrangement of series and parallel resistors is shown in front of both the tweeter and the midrange in the following schematic:
View attachment 1310633
However, what you need to know is how to use software to design an integrated crossover/attenuation circuit from scratch.
Over to the speaker design experts!
The issue is that you would have to use the L-pad on the midranger and tweeter, both to attenuate and bring the woofer to the level.
I think my statement is correct now.
Does the brand and quality of the resistors matter when assembling the Lpad and crossover?
When I made my guitar pedals, their quality mattered.
Even for manufacturing the amps
Some loudspeaker constructors would have you think that the type of resistor can make an audible difference.
At audio frequencies, where resistors are simply resistors, I recommend the use of ceramic wirewound types in loudspeaker crossovers for their power handling capabilities.
Last edited:
I would ignore the information on that site.
The statements that "resistors: added brightness...roughness...coloration...sibilance and produced at times an uninvolving sound" are simply expressions of opinion devoid of substantiating scientific data.
A case is being made in the article for not using attenuating resistors by suggesting that resistors are not perfect.
It may surprise you to learn that none of the components in a loudspeaker system can be said to be perfect!
Besides, the alternative to resistors that is presented is highly impractical and potentially deleterious to the sound if not undertaken correctly.
I'm puzzled as to how a resistor would 'add brightness'.
Some speaker builders would have you believe that a capacitor or resistor which costs five, ten or more times as much as as a perfectly serviceable good quality part will dramatically improve the sound, blah blah. They will make a significant difference: to your bank balance.
For example, I buy good quality ceramic wire-wound resistors for A$2.75, and have sometimes gone for 'Superes' at A$6 each when the cheaper ones were n/a, and I can't tell the difference. I've even used so called "junk" sand cast resistors and found them OK.
However, there is another brand which cost A$55 (!?) each and I can't see the value in that - it costs more than some decent drivers!
Geoff
Some speaker builders would have you believe that a capacitor or resistor which costs five, ten or more times as much as as a perfectly serviceable good quality part will dramatically improve the sound, blah blah. They will make a significant difference: to your bank balance.
For example, I buy good quality ceramic wire-wound resistors for A$2.75, and have sometimes gone for 'Superes' at A$6 each when the cheaper ones were n/a, and I can't tell the difference. I've even used so called "junk" sand cast resistors and found them OK.
However, there is another brand which cost A$55 (!?) each and I can't see the value in that - it costs more than some decent drivers!
Geoff
Last edited: