Impedance compensation in crossovers

EspenE · 10th January 2007, 10:20 PM

Many crossovers in published designs use compensation networks to flatten overall network impedance: typical is a LCR network placed directly across the input terminals to flatten the (often) rising impedance at crossover point.

This is supposed to make a more friendly amplifier load.

How important is this? Any opinions or experiences?

EspenE

Svante · 10th January 2007, 10:44 PM

The only time it is important is when the output impedance of the amplifier, or the cable resistance is high. If so, a varying impedance will affect the frequency response.

The load itself is easier if the impedance peak is there, since the impedance is higher.

Curmudgeon · 11th January 2007, 12:22 AM

I use low output impedance amps, so it's not an issue for me for that reason. However, Zobels and conjugate networks also help the crossover sections.

For a simplified example, a midrange speaker will have an inductor and a capacitor in series to set the upper and lower rolloffs. At low frequencies, the capacitor will increase in impedance, and the loss is the ratio of that impedance to the driver's impedance. But without a conjugate, there will be an impedance peak at the mid's low frequency resonance, and the loss from the input capacitor at that frequency will be much less than it should be. A similar issue occurs at the high end, especially for drivers with a strongly rising high frequency impedance from voice coil inductance.

Svante · 11th January 2007, 03:19 PM

Quote:

Originally posted by Curmudgeon
I use low output impedance amps, so it's not an issue for me for that reason. However, Zobels and conjugate networks also help the crossover sections.

True, but a slightly different question. If one wants to use passive filters "off the shelf" they typically behave better if the load is purely resistive. some filters will, however, present a non-resistive load to the amplifier, even if the drivers are replaced with perfect resistors.

EspenE · 11th January 2007, 06:54 PM

Thanks for the answers.

Yes, I'm aware of the use of zobel/conjugate network inside crossovers. What I'm after, is the use of compensation network "ouside" the actual crossover.

Consider the impedance curve attached: a network without impedance compensation with a broad impedance peak of 25 ohm at about 2500 Hz.

The same network with a LCR network of 10 ohm / 33 uF / 0,12 mH directly across the input gives flat impedance 5-8 ohm throughout.

On paper, the compensated network of course looks the best.

But does it matter in practice?
Only when amplifier output impedance is high?
How "high"?
Tube-amplifier-high?

Thanks,
Espen

Curmudgeon · 11th January 2007, 07:21 PM

o. With low amplifier output impedance, (high damping factor) and low cable resistance, I certainly agree that there would not be an effect.

If the amp and cable have a combined resistance of 1 ohm, a 5 ohm load will have about a 0.8 dB loss, and essentially zero into the 25 ohm load, resulting in about a 0.8 dB error. I'd consider that a second order effect, but if the speaker has a peak in that region you might not want the addition.

I hope I'm more helpful this time

sreten · 11th January 2007, 08:05 PM

Hi,

The strange thing that people use these to make the speaker " valve friendly ".

IMO anything but. Valve amplifiers love high impedance. What needs
to be done is balancing the speakers response with a significant
source resistance - this is the genuine "valve friendly" speaker.

/sreten.

10th January 2007, 10:20 PM	#1
EspenE diyAudio Member Join Date: Jun 2003 Location: Oslo	Impedance compensation in crossovers Many crossovers in published designs use compensation networks to flatten overall network impedance: typical is a LCR network placed directly across the input terminals to flatten the (often) rising impedance at crossover point. This is supposed to make a more friendly amplifier load. How important is this? Any opinions or experiences? EspenE

10th January 2007, 10:44 PM	#2
Svante diyAudio Member Join Date: Feb 2004 Location: Stockholm	The only time it is important is when the output impedance of the amplifier, or the cable resistance is high. If so, a varying impedance will affect the frequency response. The load itself is easier if the impedance peak is there, since the impedance is higher. __________________ Simulate loudspeakers: Basta! Simulate the baffle step: The Edge

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Active Crossovers & Baffle Step Compensation	Mikett	Multi-Way	21	8th June 2014 11:40 AM
Impedance compensation	DR ANALOGUE	Multi-Way	1	7th May 2006 05:51 PM
Impedance Compensation in Active Systems	ShinOBIWAN	Multi-Way	11	13th February 2006 11:39 AM
Impedance compensation	Polimorph	Multi-Way	1	6th June 2004 04:33 PM
Impedance compensation calculations based on T/S params?	Jim85IROC	Multi-Way	0	5th February 2004 09:42 PM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

11th January 2007, 12:22 AM	#3
Curmudgeon diyAudio Member Join Date: Nov 2005	I use low output impedance amps, so it's not an issue for me for that reason. However, Zobels and conjugate networks also help the crossover sections. For a simplified example, a midrange speaker will have an inductor and a capacitor in series to set the upper and lower rolloffs. At low frequencies, the capacitor will increase in impedance, and the loss is the ratio of that impedance to the driver's impedance. But without a conjugate, there will be an impedance peak at the mid's low frequency resonance, and the loss from the input capacitor at that frequency will be much less than it should be. A similar issue occurs at the high end, especially for drivers with a strongly rising high frequency impedance from voice coil inductance.

11th January 2007, 07:21 PM	#6
Curmudgeon diyAudio Member Join Date: Nov 2005	o. With low amplifier output impedance, (high damping factor) and low cable resistance, I certainly agree that there would not be an effect. If the amp and cable have a combined resistance of 1 ohm, a 5 ohm load will have about a 0.8 dB loss, and essentially zero into the 25 ohm load, resulting in about a 0.8 dB error. I'd consider that a second order effect, but if the speaker has a peak in that region you might not want the addition. I hope I'm more helpful this time

11th January 2007, 08:05 PM	#7
sreten diyAudio Member Join Date: Nov 2003 Location: Brighton UK	Hi, The strange thing that people use these to make the speaker " valve friendly ". IMO anything but. Valve amplifiers love high impedance. What needs to be done is balancing the speakers response with a significant source resistance - this is the genuine "valve friendly" speaker. /sreten.