time alignment on passive crossover?

[Ken L]

The Azurahorns I'm using are 19" deep- the problem is that there is no way to place a tweeter anywhere but somewhere around the rim of the horn - which then makes it _more_ than 19" out of time alignment- Tri-amping is out -

I can tri-amp temporarily with the Behringer to determine an actual delay target if needed.

But then I would have to figure out how to passively implement it.

If I solve the time alignment problem it's a quicker, cheaper and possibly optimum solution due to removal of the upper octaves from the mid -

If I go the 8" full-range route such as Lowther Alnico or AER BD3, it's a hell of a lot more expensive, I get time alignment and point source coherence - but a big part of me wonders if that is too much to ask from one driver. Response on a chart with a test signal is a lot different than playing seven octaves of musical detail.

On the other hand if try I to implement 1500 microseconds of delay (I guess that's 1.5 milleseconds) while crossing over a to Fostex T90A at 12k with a 2nd order slope, will I have to add so many components that I start adding distortion? And how do I go about it?

If I can do this well, the mid-range has no slopes entering the range between 340 and 3 K and is basically covering only a little more than three full octaves and at the 12db down points is covering only a little less than five.

Any guidance wouold be appreciated. I did search and looked at Linkwitz site but that left me still wondering _big grin_

Or am I looking for a solution where there is no problem? Seems like the components in a crossover will add at least some delay.



Later

Ken L

[Bill Fitzpatrick]

Although I can't help solve your problem I'm glad that you are considering it. Seems that most people don't.

I remember in my early hi-fi years someone played a recording with castinets on some Khorns and pointed out the slight echo (or maybe it should be called a double tap) due to the delay of the mid-range.

Some will say that this isn't an issue but I heard what I heard.

[pjb]

Sounds like you want to look into calculating a lattice network.

SoundEasy for instance will calculate a first or 2nd order lattice to provide delay for the tweeter in your case.

I'm not going to pretend I know how to calculate one outside design software!

[roddyama]

Originally posted by Ken L
Or am I looking for a solution where there is no problem? Seems like the components in a crossover will add at least some delay.

Later

Ken L

Hi Ken,

I have an x-over at 1200Hz between the 10” mid-bass and the mid horn. I have always made attempts to offset the horn closer to the listening position to “time-align” the horn driver with the mid-bass. This makes a difference in the smoothness of the transition from the direct radiator to the horn. This is at 1200Hz. It’s a little trickier the lower the x-over frequency. The phase shift through the x-over region becomes a large part of the time delay, and it’s a moving target with frequency. I’m not sure what the effect will be with a 3kHz x-over. You might try temporarily mounting the tweeter separately and experimenting with placement. You should be able to roughly simulate the effects of adding a time delay to see if it is worth the worry.

If you do this, you’ll have to come back and let us know the results.

[hancock]

Originally posted by Ken L

Or am I looking for a solution where there is no problem? Seems like the components in a crossover will add at least some delay.

Hi Ken,

I have good news and bad news.

bad news...

You can implement a psuedo-delay in the crossover itself, but practically speaking that is only good for about 3/4 of a wavelength. If you have 19" of offset and are crossing over at 12kHz, that is 17 wavelengths. You can use a lattice filter to implement a passive delay, but to get 1.5ms of delay and get the knee of the delay above 40kHz you're going to have some massive inducters. That means significant insertion loss.

Even if you get the the tweeter time aligned on axis, you still have to worry about lobing off axis. I don't know what vertical offset you have, let's assume the tweeter is 10" above the mid. That is 9 wavelengths at crossover. If the mid and tweeter are time aligned, then the first null will be at 3 degrees above or below axis. In other words, you would have to have your head in a vice to get the advantages of time aligning a super tweeter.

the good news...

Since you can't really get time alignment with a supertweeter--especially with a horn--you don't have to bother. In practical terms, it is best to think of a supertweeter as a separate speaker adding some missing upper-octave energy rather than as a driver integrated into the system.

Sorry to be the bearer of bad news. This is why I stick to the TAD 2001--you don't need a supertweeter--even though a 4 inch compression driver plus supertweeter has definite advantages distortion-wise.

John

[slowmotion]

Hi Ken L,

I'm just thinking here, but how about putting the Fostex T90A
inside the Azurahorn, right in front of the main driver?

Don't know how big the Fostex is, and I don't know how
the tweeter will disturb the output from the mid range driver,
but you will solve most of the time alignment problems.

Just a thought

cheers

[Ken L]

Originally posted by roddyama


If you do this, you’ll have to come back and let us know the results.

Definitely will Roddy

Originally posted by hancock


Since you can't really get time alignment with a supertweeter--especially with a horn--you don't have to bother. In practical terms, it is best to think of a supertweeter as a separate speaker adding some missing upper-octave energy rather than as a driver integrated into the system.

John

This makes a lot of sense John

Originally posted by slowmotion

I'm just thinking here, but how about putting the Fostex T90A
inside the Azurahorn, right in front of the main driver?

I had thought of this more than two or three times - a la hammer dynamics - but mounting complications due to the depth of the horns and the size of the tweeter keep me from trying it.

If I can, I think I'll borrow a pair and play with them to see how it does.

This post was made mainly to determine solutions and problems adding a supertweeter and whether or not to go the supertweeter route or the higher cost Full Range driver route with alnico Lowthers or AER's, etc.

Yesterday Jim Dowdy and I went over to Jack Roberts for a listening session and to debug hum from my 12B4 Preamp.

- Jack is one of the only three people I know of in the US that are actually running their Azurahorns - Dave Slagle is the other one. Jack was using his homebrew Parafeed 300B which a number of the Bottleheads say is the best 300B they have ever heard, along with his homebrew 12B4 preamp.

he is using vintage Philco 8" woofers and had no subs - which he needs -

In the range that the horns cover - basically down to 120 or so - it was the most beautiful and coherent sound I have ever heard - These horns just seem to be _magic_ Kind of like the Behringer - whatever problems they might have just seem to quickly be overwhelmed by positives and by taking you into the music.

Jack's setup gives up a little in detail and transient response to the two best TAD horn systems I have ever heard, Classic Audio Reproductions with TAD's and a Sierra Brooks Grand 32 with tad drivers. But really doesn't give up very much at all - Both of those systems have never had thecrossovers optimized IMHO - Jack's vintage drivers have a slight overshoot at some frequencies due to age or spider/cone or something.

While they give up the very slight loss of detail, the dispersion, smoothness and _magic_ and plain beauty of the sound of the Azurahorns still made this this the most beautiful and pleasing sound I had ever heard- If the crossovers were optimized on the other two systems there would be no best or better - It would only be choosing your oerferences for which compromisess you prefer. The Azurahorns will always lose out on detail - but only slightly - but will have the edge on smoothness, dispersion and _magic_, IMHO anyhow. 8" cones will never have the speed of the smaller Berylium. To me, compression drivers have always had a very slight harshness, though.

It showed me how much of the detail comes from the amp/pre- A really eye-opening session -

it also showed me how much can truly be done with point source coherent drivers -

I'm leaning toward Lowther PM5's instead of tweeters. Of course, who knows what I will do when I actually have the cash?

Thanks for your posts guys. It helps a lot on sorting out the details of how to best optimize my setup

I will post info on how well the final solution does.

You took the time to reply, I would like to share with you that more and more I believe that the Behringer digital crossover is really good - I will not be able to say so with certainty until I get a decent pre in my system because there might be something I could hear then that isn't there now - but as the new wears off I am left with much better integrated sound and a little tweaking is bringing additional depth in the soundstage and smoothness -

Later

Ken L

[misterears]

19" implies 1.5ms and though that might look bad on a step or impulse response, it's not a big deal subjectively. The biggest problem is getting a reasonable 1/3 oct smoothed system response at various listening locations (unless you're a one-sweet-spot listener). I'm not a time alignment believer on the top end....it's polar (if you care), extension and 1/3 smoothed response...IMO.

[Defo]

I would go for Decware DFR-8 or Tangband 1772 instead of Lowther, which is both priced and hyped up to the skies... IMO

[Defo]

Just noticed the date of this thread... Sorry