Hi Eric
On the one hand somebody can assume that the wavelengts of audio frequencies are that immensly large that only a small fraction of a wavelength can be placed into a speaker cable, so cable impedance wouldn't matter.
If one takes a really high "audio frequency" of 100kHz approx. then it's wavelength is around 3 km.
1 kHz' wavelength is a hundred times that much.
So voltage drive of a speaker would be quite fine from cost-value ratio.
On the other hand the really overwhelming dynamic range of 120 dBs of our hearing system can cause little errors being audible, that some measurements or simple considerations wouldn't expose. There have been lots of discussions on cables in audio magazines and also EW+WW and there are as many opinions as there are "experts".
I think the damping factor should be as low as possible but not at the cost of any other parameter. One has to consider that the usual DF measurement is made with a steady-state sinusoidal frequency and a PURELY RESISTIVE load (maybe there is a need for a new standard ?!). But how will an amp react when loaded with a real life speaker ? I think that's the reason why usually really beefy amps have more control over the load, simply because they are left virtually unimpressed by what is connected to them.
Regards
Charles
P.S. regarding the advantages of active speakers I forgot to mention "short cables" !!
On the one hand somebody can assume that the wavelengts of audio frequencies are that immensly large that only a small fraction of a wavelength can be placed into a speaker cable, so cable impedance wouldn't matter.
If one takes a really high "audio frequency" of 100kHz approx. then it's wavelength is around 3 km.
1 kHz' wavelength is a hundred times that much.
So voltage drive of a speaker would be quite fine from cost-value ratio.
On the other hand the really overwhelming dynamic range of 120 dBs of our hearing system can cause little errors being audible, that some measurements or simple considerations wouldn't expose. There have been lots of discussions on cables in audio magazines and also EW+WW and there are as many opinions as there are "experts".
I think the damping factor should be as low as possible but not at the cost of any other parameter. One has to consider that the usual DF measurement is made with a steady-state sinusoidal frequency and a PURELY RESISTIVE load (maybe there is a need for a new standard ?!). But how will an amp react when loaded with a real life speaker ? I think that's the reason why usually really beefy amps have more control over the load, simply because they are left virtually unimpressed by what is connected to them.
Regards
Charles
P.S. regarding the advantages of active speakers I forgot to mention "short cables" !!
Hi Charles,
I understand that characteristic impedence ought to be irrelevant in the audio band, but modern digital audio causes RF throughput.
Yes, Self I think referred to Interface Intermodulation Distortion.
Also other articles referring to loop inductance causing distortion.
Yes, I am discounting the steady state sine wave condition.
Rather the pulse condition.
And yes stout amplifiers are typically less load dependant.
Is this a case of a big supply quencing the return pulse without significantly modulating the power rails and causing PSRR consequent artifacts ?.
High loop NFB amp implies high number of gain stages which implies FB propagation delay and opportunity for ringing condition.
Any such ringing would cause sonic masking of the original impulse (not silence after an impulse) and loss of real clarity.
I find this to be a sonic effect in high NFB amps imo, and fatigueing.
What happens to the cone during the overshoot condition.
Ie - does putting a sudden brake on the cone cause mechanical ringing in the cone, and is a gentle braking more sonically correct ?.
Nelson gives musing about air being non symmetrical, so should the braking rate of the cone be matched to the characteristics of air ?.
Eric.
I understand that characteristic impedence ought to be irrelevant in the audio band, but modern digital audio causes RF throughput.
Yes, Self I think referred to Interface Intermodulation Distortion.
Also other articles referring to loop inductance causing distortion.
Yes, I am discounting the steady state sine wave condition.
Rather the pulse condition.
And yes stout amplifiers are typically less load dependant.
Is this a case of a big supply quencing the return pulse without significantly modulating the power rails and causing PSRR consequent artifacts ?.
High loop NFB amp implies high number of gain stages which implies FB propagation delay and opportunity for ringing condition.
Any such ringing would cause sonic masking of the original impulse (not silence after an impulse) and loss of real clarity.
I find this to be a sonic effect in high NFB amps imo, and fatigueing.
What happens to the cone during the overshoot condition.
Ie - does putting a sudden brake on the cone cause mechanical ringing in the cone, and is a gentle braking more sonically correct ?.
Nelson gives musing about air being non symmetrical, so should the braking rate of the cone be matched to the characteristics of air ?.
Eric.
Hi Jens
Of course can you make your amps output resistance positive or negative or even a complex impedance. But if you don't take care "it might get interesting".
I will put together what might be useful to you and may be post it here (is anybody else interested ?) or mail it to you.
But I have to close now.
Regards
Charles
Of course can you make your amps output resistance positive or negative or even a complex impedance. But if you don't take care "it might get interesting".
I will put together what might be useful to you and may be post it here (is anybody else interested ?) or mail it to you.
But I have to close now.
Regards
Charles
activating (activizing) bass branch...
Charles,
the schematic of my bass XO is posted in the loudspeaker forum. If you forget about the bass enhancement, it is a simple 3rd order low pass.
Instead of worrying which kind of core for the inductors will give me the best peformance, it itches me to place the LCLR filter (rescaled for nice small air coils) before the amp and drive the speakers directly.
Is the interaction of the speaker with the XO purely resistive? In order not to change damping and tonality, it would then suffice to place 0R43 in series with the power amp. Doesn't do anything for the damping factor but gets rid of the distortion in the ferrite coils and electrolytic.
If not, could I place L1 and C1 before the amp and keep L2, maybe in series with 0R19?
Eric
Charles,
the schematic of my bass XO is posted in the loudspeaker forum. If you forget about the bass enhancement, it is a simple 3rd order low pass.
Instead of worrying which kind of core for the inductors will give me the best peformance, it itches me to place the LCLR filter (rescaled for nice small air coils) before the amp and drive the speakers directly.
Is the interaction of the speaker with the XO purely resistive? In order not to change damping and tonality, it would then suffice to place 0R43 in series with the power amp. Doesn't do anything for the damping factor but gets rid of the distortion in the ferrite coils and electrolytic.
If not, could I place L1 and C1 before the amp and keep L2, maybe in series with 0R19?
Eric
Wow, what a debate! Lots of opinoins, few who really knows.
To be short: Most amps have decently low output impedance (high damping factor). Some tube amps and non-feedback amps with MOSFET's may have rather high output impedance.
As Nelson says, speakers (some of them) may get influenced by different kinds of damping factor and therefore sounds different (not necessarily bad) with different amps.
To be short: Most amps have decently low output impedance (high damping factor). Some tube amps and non-feedback amps with MOSFET's may have rather high output impedance.
As Nelson says, speakers (some of them) may get influenced by different kinds of damping factor and therefore sounds different (not necessarily bad) with different amps.
Hi Eric
Uh, that bass enhancement circuit is even 2nd order ..... !
But apart from that, the load a driver presents to a crossover is never purely resistive. If you want to have a resistive load on the crossover you will have to correct the driver's impedance. If you don't want to do this then you have to account for the frequency dependant load when designing the crossover.
In other words: This lowpass will have a different amplitude response when driving a resistor than a loudspeaker driver.
Just try to measure the response at the driver's terminals.
If you want to use a passive filter in front of your amp then you would also have to design it for higher impedance !
Your resistor problem:
If it were a closed box then I would say, just try it, the bass will sound tighter and more precise if the tuning wasn't already of very low Qtc. In the latter case (quite seldom) the bass response would drop too early and sound too weak.
But I think your speaker is a TML (a favourite with T+A), isn't it ?
I don't know what will happen then, although I have a simulation program for TMLs. But I don't know how accurate this is (some people say TMLs couldn't be calculated at all).
As far as I remember TMLs don't like drivers with too low a Qts.
Maybe you should ask planet10, he is aTML expert.
Regards
Charles
P.S. In the meantime I did the maths again for the simulated output resistance. If anybody is interested I will post it.
Uh, that bass enhancement circuit is even 2nd order ..... !
But apart from that, the load a driver presents to a crossover is never purely resistive. If you want to have a resistive load on the crossover you will have to correct the driver's impedance. If you don't want to do this then you have to account for the frequency dependant load when designing the crossover.
In other words: This lowpass will have a different amplitude response when driving a resistor than a loudspeaker driver.
Just try to measure the response at the driver's terminals.
If you want to use a passive filter in front of your amp then you would also have to design it for higher impedance !
Your resistor problem:
If it were a closed box then I would say, just try it, the bass will sound tighter and more precise if the tuning wasn't already of very low Qtc. In the latter case (quite seldom) the bass response would drop too early and sound too weak.
But I think your speaker is a TML (a favourite with T+A), isn't it ?
I don't know what will happen then, although I have a simulation program for TMLs. But I don't know how accurate this is (some people say TMLs couldn't be calculated at all).
As far as I remember TMLs don't like drivers with too low a Qts.
Maybe you should ask planet10, he is aTML expert.
Regards
Charles
P.S. In the meantime I did the maths again for the simulated output resistance. If anybody is interested I will post it.
Seems to be complicated to activate the woofers. Still seems to me I could save money on quality caps and baked air coils by activating mid and high range. I could measure amplitude and phase at the drivers and tailor a passive low power filter with the same resonse that sits in front of the amp. I am not sure whether I need to include the series resistance.
Would the midrange (2x approx 3 inch cones) and tweeter (i inch dome) also be sensitive to the resistance they see? Would they profit at all from being damped by the amp?
Eric
Would the midrange (2x approx 3 inch cones) and tweeter (i inch dome) also be sensitive to the resistance they see? Would they profit at all from being damped by the amp?
Eric
Hi Eric
Though this is a little off-topic in this DF thread I will try to answer it.
The series resistances have influence on the tuning of the tweeter and midrange as well (damping and amplitude response).
In the latest "Klang und Ton" they mentioned that their flagship speaker has a less smooth (but still excellent) amplitude response in it's active version but that the overall improvment achieved by activatin would well be worth the effort.
I have never done an activation of a passive loudspeaker. All speakers I used in active configuration were active from the beginning so I don't see any reason for not going active.
Is there any chance that you can borrow the additional amps and a (maybe adjustable) crossover to make a try (at least the crossover wouldn't be that difficult and expensive to build just for this trial)?
Many "activists" are using adjustable Linkwitz Riley crossovers and achive excellent results even thogh they are not exactly tailored to the speakers used.
Regarding the simulation of output impedances you seem to be the only person interested in. Since it is a lot of work I may probably fax my hand notes to you because it would give me some work to put it in a neat electronic form.
The passive equalization circuits work as complex frequency-selective transformation like the ones that are common for matching purposes in RF-circuits.
Regards
Charles
P.S. You still owe me the answer if your speaker is a TML.
Though this is a little off-topic in this DF thread I will try to answer it.
The series resistances have influence on the tuning of the tweeter and midrange as well (damping and amplitude response).
In the latest "Klang und Ton" they mentioned that their flagship speaker has a less smooth (but still excellent) amplitude response in it's active version but that the overall improvment achieved by activatin would well be worth the effort.
I have never done an activation of a passive loudspeaker. All speakers I used in active configuration were active from the beginning so I don't see any reason for not going active.
Is there any chance that you can borrow the additional amps and a (maybe adjustable) crossover to make a try (at least the crossover wouldn't be that difficult and expensive to build just for this trial)?
Many "activists" are using adjustable Linkwitz Riley crossovers and achive excellent results even thogh they are not exactly tailored to the speakers used.
Regarding the simulation of output impedances you seem to be the only person interested in. Since it is a lot of work I may probably fax my hand notes to you because it would give me some work to put it in a neat electronic form.
The passive equalization circuits work as complex frequency-selective transformation like the ones that are common for matching purposes in RF-circuits.
Regards
Charles
P.S. You still owe me the answer if your speaker is a TML.
Dear Charles,
I put your response and my answers in the T+A thread in the speaker forum.
The only thing that is relevant here is the question whether a mid or high range speaker would profit from high damping. Well, frequency response may show more jagged features as you said. How about transient behavior etc.?
Just for clarity: low output resistance = low Q? Looks strange at first sight but made sense once I realized that a low output resistance will actually dampen the oszillation of speaker cone & air and hence lower its Q.
But a damped oscillator has flatter and broader resonances. Why would damping then cause a more jagged frequency response?
Eric
I put your response and my answers in the T+A thread in the speaker forum.
The only thing that is relevant here is the question whether a mid or high range speaker would profit from high damping. Well, frequency response may show more jagged features as you said. How about transient behavior etc.?
Just for clarity: low output resistance = low Q? Looks strange at first sight but made sense once I realized that a low output resistance will actually dampen the oszillation of speaker cone & air and hence lower its Q.
But a damped oscillator has flatter and broader resonances. Why would damping then cause a more jagged frequency response?
Eric
Hi Eric
Unless the driver's Q is already very low, things will most likely be better with better damping.
The response issue is simply because a good designer will arrange the COMBINATION of driver AND crossover network to perform as wanted.
If you now go and connect maybe a third-order butterworth or 4th order LR crossover in front of your amplifiers the original tweaked response doesn't apply anymore.
In other words: the optimal active crossover would be the one that is tailored to the drivers in use, but you can still achieve good results with tuneable universal crossovers.
Regards
Charles
P.S. Have you got enough amplifiers to make a test-run ? I might probaly have a simple crossover you could use (i.e. I would want it back afterwards) and tweak but this is depending on the current filter functions of your speaker (my crossover is 3rd order butterworth).
Unless the driver's Q is already very low, things will most likely be better with better damping.
The response issue is simply because a good designer will arrange the COMBINATION of driver AND crossover network to perform as wanted.
If you now go and connect maybe a third-order butterworth or 4th order LR crossover in front of your amplifiers the original tweaked response doesn't apply anymore.
In other words: the optimal active crossover would be the one that is tailored to the drivers in use, but you can still achieve good results with tuneable universal crossovers.
Regards
Charles
P.S. Have you got enough amplifiers to make a test-run ? I might probaly have a simple crossover you could use (i.e. I would want it back afterwards) and tweak but this is depending on the current filter functions of your speaker (my crossover is 3rd order butterworth).
Hi Charles,
sorry, wasn't around this weekend and didn't get to post anything in the loudspeaker forum.
Yes, the T160E seems to be a transmission line.
I can build an active filter and a sufficient number of power amp boards very quickly, but thanks for the offer.
I am very firm in digital and analog audio electronics but did not get too much into speaker design until I realized what kind of junk can be inside even top of the line speakers. Therefore I am currently reading up on speaker and crossover design.
As a quick fix, I am still wondering whether I should redo my passive crossover with decent components (which will cost several 100 €) or go (partially) active without too much tweaking.
From what I have gathered so far from your posts and some reading, measuring voltage vs. frequency at the speaker terminals and tailoring an active filter with the same resonse should work well for mid and high range. The benefit would be mainly in lower distortion, the benefit of high damping on mid and hi not being so evident. Correct?
For the bass, the same approach but with added series resistance should work, foregoing for the time being the positive effect of higher damping until I can work out a more intricate filter.
Eric
sorry, wasn't around this weekend and didn't get to post anything in the loudspeaker forum.
Yes, the T160E seems to be a transmission line.
I can build an active filter and a sufficient number of power amp boards very quickly, but thanks for the offer.
I am very firm in digital and analog audio electronics but did not get too much into speaker design until I realized what kind of junk can be inside even top of the line speakers. Therefore I am currently reading up on speaker and crossover design.
As a quick fix, I am still wondering whether I should redo my passive crossover with decent components (which will cost several 100 €) or go (partially) active without too much tweaking.
From what I have gathered so far from your posts and some reading, measuring voltage vs. frequency at the speaker terminals and tailoring an active filter with the same resonse should work well for mid and high range. The benefit would be mainly in lower distortion, the benefit of high damping on mid and hi not being so evident. Correct?
For the bass, the same approach but with added series resistance should work, foregoing for the time being the positive effect of higher damping until I can work out a more intricate filter.
Eric
As for the dumping coefficient, it is important to understand that its value is indicated on the terminals of the amplifier itself. At the other end of the speaker cable at the speaker terminals, the dumping coefficient may drop to normal values. If the cable is not high enough quality. It is also important to understand how the high value of the dumping factor is obtained.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.