Delay effects of crossover??

[Ken L]

If I cross over at say1200hz with a 3rd order high pass and 2nd order low-pass - what effect would the components have on delay and time alignment?? (between Mid and hf driver?)


what about 1st order around 10 K with 2.4 mfd cap and .1 inductor?

I guess I'm having trouble giving up on the configuration discussed in this earlier thread

Center to Center of Drivers 29" apart in proposed config - Problems??

Also, I'm thinking about sticking the DDS Waveguide in center of it coaxially, but this still has the hi freq driver aobut 15" in front of the 8" full range

Regards

Ken L

[theChris]

the delay near the crossover point will be of course determined by phase, and this would be determined by Q. however there will likely be mechanical and electrical mechanisms at work, changing the acoustic response. you'll notice that crossover's like the linkwitz riley crossover are designed to give a flat acoustic response if the speakers are located close and would normally be in-phase.

so from an engineering standpoint it isn't as easy as it looks. people have crossover modeling software to help them out.

[John Sheerin]

Hi Ken,
Well, the mismatched electrical slopes would be going in the right direction (more delay on the highs), but I'm not sure if it would really be enough. That would be a handy place for that digital crossover...

What I'd really like to mention is that the polar response of the Azura horn will go down the toilet above ~300hz if you stick the DDS waveguide in the throat (from FEA simulation). So depending on your room, you might not like the results - on axis response could be a bit attenuated relative to off axis, although in the quick look I took, it seemed like there might be some diffraction around the waveguide putting some sound back into the very center of the pattern. I actually really liked the Azura horns I heard a few weeks ago with just Lowther DX4's - they only went to 13khz, but it was flat response and sounded really good.

John

[Ken L]

Originally posted by John Sheerin
Hi Ken,
Well, the mismatched electrical slopes would be going in the right direction (more delay on the highs), but I'm not sure if it would really be enough. That would be a handy place for that digital crossover...

Definitely gotta go passive with this one.

I actually really liked the Azura horns I heard a few weeks ago with just Lowther DX4's - John

Yep, that's my most likely choice from here.

I'm trying to find a deal on some PM5's or PM4's. There were some PM5's on e-bay - but only 25% off for used.

Can't seem to stop thinking about the waveguides _grin_

Better find some Lowthers soon _ bigger grin_

Thanks for your post

regards

Ken L

[Thunau]

Originally posted by John Sheerin
Hi Ken,
Well, the mismatched electrical slopes would be going in the right direction (more delay on the highs), but I'm not sure if it would really be enough. That would be a handy place for that digital crossover...


John

I believe that it's the low frequencies that get delayed by a LP filter in relation to the highs. If the LP is a 3rd order, its group delay will be bigger than the group delay introduced by a 2nd order HP.
I assume we are talking acoustic slopes, because if not, than that particular system could theoretically result in a properly phase tracking 4th order acoustic system anyway- depending on what the drivers are doing acoustically with no electrical filters applied.

[Lovan]

Originally posted by Thunau


I believe that it's the low frequencies that get delayed by a LP filter in relation to the highs. If the LP is a 3rd order, its group delay will be bigger than the group delay introduced by a 2nd order HP.
I assume we are talking acoustic slopes, because if not, than that particular system could theoretically result in a properly phase tracking 4th order acoustic system anyway- depending on what the drivers are doing acoustically with no electrical filters applied.

It´s the other way around, HP 3rd order and LP 2nd order.

Isn´t the order determing Q?

[Ken L]

Originally posted by Thunau

I assume we are talking acoustic slopes

I was intending to ask about electrically

Originally posted by Lovan

It´s the other way around, HP 3rd order and LP 2nd order.
............confused:

Well, 3rd order HP and 2nd order lowpass is what I intended to ask about.

And, even though I may well have been off on a tangent and may not do the project, the question of how much delay electrical components impart to a crossover is a something I have always wondered about.

regards

Ken L

[hancock]

Originally posted by Ken L
If I cross over at say1200hz with a 3rd order high pass and 2nd order low-pass - what effect would the components have on delay and time alignment?? (between Mid and hf driver?)

Regards

Ken L

Hi Ken,

The best thing I can recommend is to download the demo version of lspcad and do some playing around. Put in the amount of physical offset. Put in some different starting crossovers and let the solver go to work. Remember you want a lower slope on the driver that is in front. For a small amount of offset you can do this with a crossover with the same order LP and HP but with different Q's. For more offset you'll have to use different orders. The max physical offset you can correct for is about 3/4 wave. Be sure to check out the vertical polar response to see the reduction in lobing issues you get from a properly designed crossover.

John

[Ken L]

Hi John

Originally posted by hancock
... Remember you want a lower slope on the driver that is in front...

This is a good point to know

Originally posted by hancock
For more offset you'll have to use different orders.

This also

Originally posted by hancock
The max physical offset you can correct for is about 3/4 wave.

You're still talking front to back offset here?

So after 3/4 of a wave you just line up with the next wave?? and you're one hz off but in phase??

Thanks for your post.

regards

Ken L

[hancock]

Originally posted by Ken L


You're still talking front to back offset here?

Hi Ken,

Well, it's really more general than just front-to-back. What is really important is the amount of delay between the arrival times of the drivers at the listener. So, for instance, if the two drivers are aligned acoustically in the vertical plane, but the listener is above axis, that is equivalent to the listener being on axis and the drivers being offset.

So after 3/4 of a wave you just line up with the next wave?? and you're one hz off but in phase??

Well, in this case, if you do manage to get the drivers "in phase" at the crossover point, you will minimize lobing issues at the crossover point, but you will still have lobing above and below this point. However, it will still be an improvement because lobing is most problematic at the crossover point (since both signals are equal in power). In general, even if you can only partially compensate for the acoustical offset with the crossover, it will be an improvement.

If you want to go a little deeper on the subject, I recommend reading some of these:

"Loudspeaker Driver Phase Response: The Neglected Factor in Crossover Design" by Marshall Leach Jr.

"Active Crossover Networks for Noncoincident Drivers" by Sigfried Linkwitz

"Time Offset and Crossover Design" by Dennis Fink

"Is Phase Linearization of Loudspeaker Crossover Networks Possible by Time Offset and Equalization?" by John Vanderkooy and Stanley Lipshitz

"Power Response of Loudspeakers with Noncoincident Drivers--the Influence of Crossover Design" by Vanderkooy and Lipshitz

"Acoustic Alignment of Loudspeaker Drivers by Nonsymmetric Crossovers of Different Orders" by B. Hillerich

All these articles are in the JAES Loudspeaker Anthologies, a highly recommended purchase if you want to take this hobby to the next level.

John