In a sealed box, the lower the Qt, the less the "hangover". The lower the Qt, the more the bass is rolled off. The basic choices for a box with no sub operating below, attempting to offer bass extension.
Where
there is a sub below, so bass extension is not needed
best transient response (minmal hangover) is the aim ~
however the passband of the box (in my case starting at about 80 Hz) being designed is crossing in at close to it's Fc:
In my case (drivers are Lambda 15TDM), a quite large box is needed to get Qt below 0.6, so:
How far above Fc is the reduced hangover of a low Qt is apparent?
Maybe one octave (ie worth pursuing with a bigger box for lower Qt), or only part of an ocatve around Fc - maybe mot woth worrying about?
Cheers
Where
there is a sub below, so bass extension is not needed
best transient response (minmal hangover) is the aim ~
however the passband of the box (in my case starting at about 80 Hz) being designed is crossing in at close to it's Fc:
In my case (drivers are Lambda 15TDM), a quite large box is needed to get Qt below 0.6, so:
How far above Fc is the reduced hangover of a low Qt is apparent?
Maybe one octave (ie worth pursuing with a bigger box for lower Qt), or only part of an ocatve around Fc - maybe mot woth worrying about?
Cheers
I have now read the article at www.winebase.com.au/audio/index.html on Sealed box design, which looks at a critical factor many speaker design books don't mention: calculating the effect of the amp.
Using a valve amp, with much higher output resistance: 100 litres might be needed!
Hence wanting to be 100% certain that Qt is equally relevant above Fc, before building a 100 litre midrange!
Thanks
Using a valve amp, with much higher output resistance: 100 litres might be needed!
Hence wanting to be 100% certain that Qt is equally relevant above Fc, before building a 100 litre midrange!
Thanks
Hi
Good point about the amplifier damping factor affecting driver Qt. Have you thought about adding series resitance between the driver and the amp thereby decoupling the amp and swamping out the amps variable Rout. Should help to give a smaller box for your target system Q. At the expence of lowering the sensitivity of the midrange, but might not be a problem if using a woofer below.
Good point about the amplifier damping factor affecting driver Qt. Have you thought about adding series resitance between the driver and the amp thereby decoupling the amp and swamping out the amps variable Rout. Should help to give a smaller box for your target system Q. At the expence of lowering the sensitivity of the midrange, but might not be a problem if using a woofer below.
Hi
I didn't think of series resistance. As the drivers are very efficient, that's an option. Fyi specs:
1w Spl 97.8 dB
Fs 35 Hz
Qms 5.09, Qes 0.35, Qts 0.33
Vas 312 Liters; Sd 855 cm2; Xmax 3.3 mm
Re 6.6 ohms;
Le 0.2 mH
. . lowering the sensitivity I'd have to pad down the tweeter, but being c. 100 dB, that's ok.
Using Unibox's Rs factor, to get Qt of 0.55 with no series resistance, volume calcs as 135 litres.
With series resistance of 1 ohm (amplifer, cable and series resistor), the volume calculated increases to 208 litres . . this compounds the problem
Have I misunderstood, or any other ideas?
Thanks
I didn't think of series resistance. As the drivers are very efficient, that's an option. Fyi specs:
1w Spl 97.8 dB
Fs 35 Hz
Qms 5.09, Qes 0.35, Qts 0.33
Vas 312 Liters; Sd 855 cm2; Xmax 3.3 mm
Re 6.6 ohms;
Le 0.2 mH
. . lowering the sensitivity I'd have to pad down the tweeter, but being c. 100 dB, that's ok.
Using Unibox's Rs factor, to get Qt of 0.55 with no series resistance, volume calcs as 135 litres.
With series resistance of 1 ohm (amplifer, cable and series resistor), the volume calculated increases to 208 litres . . this compounds the problem
Have I misunderstood, or any other ideas?
Thanks
Adding resistance in series with a driver in a closed box increases the Qtb... if the resistance has a positive value.
If the resistance could have a negative value, the Qtb will decrease. There is no real component such as a resistor of negative value, however you can simulate it using positive feedback around a power amplifier. It needs a few components. It's quite easy to get a closed box speaker with a low Qtb, less 0.5.
As the bass response would then be decreasing with lowering frequencies, a boost ahead of the power amplifer would provide fairly linear frequency response.
This can be easily simulated with a 12 dB/o hi-pass filter and a bass boost of 6 dB/o. I used Tina demo version. Results are quite astonishing : very extended bass region, very low group delay.
Some drivers speakers have very low Qts, see Volt for example, and do not need negative resistance at all.
The initial idea was submitted by Russell Breden in Electronics World.
If the resistance could have a negative value, the Qtb will decrease. There is no real component such as a resistor of negative value, however you can simulate it using positive feedback around a power amplifier. It needs a few components. It's quite easy to get a closed box speaker with a low Qtb, less 0.5.
As the bass response would then be decreasing with lowering frequencies, a boost ahead of the power amplifer would provide fairly linear frequency response.
This can be easily simulated with a 12 dB/o hi-pass filter and a bass boost of 6 dB/o. I used Tina demo version. Results are quite astonishing : very extended bass region, very low group delay.
Some drivers speakers have very low Qts, see Volt for example, and do not need negative resistance at all.
The initial idea was submitted by Russell Breden in Electronics World.
A series resistance effectively increases the Q of the driver and also reduces sensitivity.
Alternatively, it can be seen as a voltage divider, which lowers the signale to the speaker, but it lowers it less if the driver impedance is larger (ie at resonance).
The two viewpoints are equivalent and produces the same results if the calculations are done.
Below is the effect on two systems of a series resistance of 2 ohms.
Alternatively, it can be seen as a voltage divider, which lowers the signale to the speaker, but it lowers it less if the driver impedance is larger (ie at resonance).
The two viewpoints are equivalent and produces the same results if the calculations are done.
Below is the effect on two systems of a series resistance of 2 ohms.
An externally hosted image should be here but it was not working when we last tested it.
infinia
> 100-120 liters is not bad for a 15 woofer
yes, though it's not used as woofer, it's 80 Hz up, I'd call it a mid-bass. In fact it (the Lambda 15TDM) is to be used instead of an earlier planned mid-bass horn (deferred due to lack of time).
Compared to a mid-bass horn it is smaller, but when I use a sub as well, 120 litres is bigger than desirable/ WAF.
Svante
Thanks, your graph shows the where series resistance is useful - to flatten response around resonance.
forr
Just to be clear I don't need to extend bass; but yes, lowering group delay is the part of the aim.
I don't have the Electronics World article:
When you say "boost ahead of the power amplifer would provide fairly linear frequency response" " is it a form of EQ?
Are there any ill negative?
effects using 'positive feedback around a power amplifier'?
Cheers
> 100-120 liters is not bad for a 15 woofer
yes, though it's not used as woofer, it's 80 Hz up, I'd call it a mid-bass. In fact it (the Lambda 15TDM) is to be used instead of an earlier planned mid-bass horn (deferred due to lack of time).
Compared to a mid-bass horn it is smaller, but when I use a sub as well, 120 litres is bigger than desirable/ WAF.
Svante
Thanks, your graph shows the where series resistance is useful - to flatten response around resonance.
forr
Just to be clear I don't need to extend bass; but yes, lowering group delay is the part of the aim.
I don't have the Electronics World article:
When you say "boost ahead of the power amplifer would provide fairly linear frequency response" " is it a form of EQ?
Are there any ill negative?
effects using 'positive feedback around a power amplifier'? Cheers
Hi Rick57,
---When you say "boost ahead of the power amplifer would provide fairly linear frequency response" " is it a form of EQ?---
The simulated boost I played with consists of a simple frequency dependant dividing network using two resistors and a capacitor connected to ground. If the voltage gain available from the preamplifier+amplifier is insufficient, a little op-amp with some voltage gain would do the trick.
---Are there any ill negative? effects using 'positive feedback around a power amplifier'?---
From my current simulations, ill effects of positive feedback begin to be just noticeable when the DC resistance of the voice-coil has been reduced to 1/3 of its initial value by the negative resistance of the amplifier output. They are due to the inductive component of the voice coil and can be reduced with a capacitive compensation in the positive feedback loop.
However when playing with positive feedback at the experimental stage, circuit protections are recommanded (a polyswitch in series with the driver for example).
---When you say "boost ahead of the power amplifer would provide fairly linear frequency response" " is it a form of EQ?---
The simulated boost I played with consists of a simple frequency dependant dividing network using two resistors and a capacitor connected to ground. If the voltage gain available from the preamplifier+amplifier is insufficient, a little op-amp with some voltage gain would do the trick.
---Are there any ill negative? effects using 'positive feedback around a power amplifier'?---
From my current simulations, ill effects of positive feedback begin to be just noticeable when the DC resistance of the voice-coil has been reduced to 1/3 of its initial value by the negative resistance of the amplifier output. They are due to the inductive component of the voice coil and can be reduced with a capacitive compensation in the positive feedback loop.
However when playing with positive feedback at the experimental stage, circuit protections are recommanded (a polyswitch in series with the driver for example).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.