trying to integrate an ET-703 tweeter with a TAD 2402. Could do with the dimensions for the 2402 cabinet...in particular the dimension from the diaphram of the 4001 driver to the front of the cabinet!
Trying to time align the ET-703 you see!!
Regards
Richard
Richard,
In order to time align, getting things on the same plane doesn't get your drivers time aligned. Their will be some phase shift through the crossover. What you need to do is align them acoustically, taking the crossover into account.
What sort of crossover did you intend for the ET-703? The ET-703 doesn't have the greatest looking response curves all by itself.
Cyclotronguy
Hi,
Well the TAD's sound very nice as-is. However i was thinking of a 3rd order crossover in the 10khz area. The 703's are pretty flat up to 45khz.
I'm a crossover novice...i don't really want to have a complicated crossover it makes things worse in the opinion of a few pro companies.
I built some custom stands for them this week in oak..the sound has improved immensely!
Originally Posted by Cyclotronguy
That is a great idea. I have used them that way and they blend in quite nicely.
Nope... ask anyone who has ever measured a pair... they have a broad peak around 15KHz and fall off to around 25KHz where they level off and continue on out, but don't sweat it... they sound great.
Also, I wouldn't worry too much about the time alignment thing... it is far more critical in the mid band. Physically pushing the tweeter back will likely cause all sorts of diffraction and reflection issues so short of using digital delay, it really isn't worth it.
Widget
Hey Mr.W
Thanks for the advice.
Yes i imagine they have done their homework regarding the midband.
Do you suggest then just sitting them near the front and testing by ear? I don't want to outlay the cash you see, only for it to go pear shaped! Although i suppose it would still be a great learning experience.
Do you use digital correction? I read that one is less likely to notice time alignment issues with a tweeter than any other component.
Regards
Richard
I'll have to respectfully disagree with Widget on the tweeter time alignment issue. I have found that placing the acoustic centers of the mid driver and tweeter the same distance from the listener is critical to obtaining the best imaging from a system.
An error between a mid driver and woofer is usually a small fraction of a wavelength at the crossover frequency, or perhaps a single wavelength in the case of midrange and midbass horns. With a midrange horn and tweeter though, mounting the tweeter at the mid horn's mouth will result in an arrival error of several wavelengths at the crossover frequency. This really matters, due to the way our ear-brain systems determine localization of sounds. If the high frequency components of a sound arrive several wavelengths before the midrange components our ability to localize the sounds is compromised.
The way I have aligned components in the past is to begin by aligning the acoustic centers of tweeter and midrange, then experiment by moving both tweeters by 1/4" increments until the sweet spot is found. This spot might be either forward or back of the starting point. If all system components are placed accurately in relation to the listening chair (a stretched string works much better than eyeballing it) and driver polarities are correct, there will be a distance for the tweeters where the imaging firms up and the tweeters blend most perfectly with the mids and disappear as sources. The effect is something like focusing a camera lens, and it is a really cool effect when the system "pops" into focus.
I do agree with Widget that surfaces near the tweeters need to be treated to prevent closely timed reflections. Sheets of soft felt or towels work well for this.
Another solution I've not seen discussed much of anywhere... and I will send anyone interested to the ALK website, where Al discusses simple crossovers followed by traps.
In this design you have as for example a nice second order crossover with a trap. As a result the slope starts out at something like 70 dB / octave and resorts to 2nd order by the time it is at -24 dB.
As a result their is very little argument between the two drivers (in our case a mid and tweeter) with-in the crossover range. You can then play around with phase (absolute and reverse) while looking at impulse response in something like MLSSA and get a pretty nice response.
This little twist in crossover design, doesn't address time alignment but it does minimize the interference issues between drivers that would normally have a large overlap... and perhaps that's part of the issue.
Cyclotronguy
Using steeper slopes, 3rd, 4th, or higher at crossover frequencies around 10KHz, I just don't hear the image collapse or sharpen when I have moved the tweeter in and out... it is easy to see the effects of proper alignment on an impulse measurement or in the ripples you get due to comb filter effects, but I am not personally convinced that they are as important as some would have you believe.
If alignment can be achieved without any other sacrifice then by all means do it, but typically there are other factors that limit our ability to achieve this. All of that said, I wouldn't simply align the VCs... you really need to shoot an impulse response at the intended listening distance and height and physically adjust the system for the cleanest pulse.
Widget
While this technique doesn't actually "fix" anything, it certainly can reduce the effects of driver interaction... I have not tried any of Al's networks, but I have used DEQX with 300dB/octave slopes which also reduce the drivers' interactions... I am not sold on this digital technique... not damning it mind you, I am just not sure it is the panacea that some claim it to be.
Widget
Absolutely agree.
I have found that my intuitive (and often psychoacoustic...) expectations are not met in real world measurements. Taking real impulse measurements and adjusting accordingly, remeasuring and listening, moot VC (voice coil) physical alignment alone.
There is so much occurring in the crossover(s) that measuring the response post crossover is clearly what matters, and then adjusting accordingly. The results thereby are much more uniform and less subjective.
bo
"Indeed, not!!"
With the frequencies you are dealing with, above 10kHz, the wavelengths are so small that you end up with interference and comb filtering even if you get the voice coils time aligned. In this regard I agree with Cyclotronguy's recommendation to try a steep slope crossover and not be so concerned about time alignment.
While physical time alignment may help, I would expect lots of nasty reflections from the supertweeter if it were mounted anywhere behind the front baffle of the speaker. However, with the crossover so high the supertweeter dispersion is probably already confined by the throat of the horn, so you may not need to deal with diffraction based on your mounting scheme.
Your kidding right? You're willing to spend $2000 on a pair of supertweeters that will add 1/2 octave to the audible spectrum if you are over 35 years old, on a multi-thousand dollar speaker system, but you're not willing to take any measurements to develop a proper crossover network? You will save yourself a lot of headaches if you purchase SoundEasy for $229, and a Behringer microphone and preamp for $100 and take some measurements. If you add a MAudio (Delta 410 $157) soundcard you will also be able to emulate any crossover design you choose, listen to how it will sound with your speaker, and tweak the component values before you actually buy the components. This will give you more crossover design power than any speaker engineers had before the 1980 or even 90s.
If you really want to get serious you could use Soundeasy crossover abilities to design a digitally delayed, transient perfect, 3-way, fully active, equilized crossover for your TAD 2402.
Other people may tell you different, but the odds of developing a good crossover for a compression driver and horn without measurement tools is slim to none. If someone tells you differently, I would challenge them to show you measurement to prove it. Other people may tell you that an RTA is good enough, but for the same money you could have Soundeasy with many times more resolution and the essential phase information you need for crossover design.
I was speaking mostly of using crossover points between midrange and tweeter of roughly 5kHz. to maybe 7kHz. Impulse measurements are certainly the way to go... its just that since I've had the capability of making impulse measurements I have not been using standalone tweeters.
As I recall the new Everest speaker brings the little beryllium supertweeter IN at 20kHz. When we were given a demo at JBL, Don and I dutifully made the trip up to the speaker to put our ears to the tweeters and we both heard zip, zero, nada from them! In this case I doubt the tweeter misalignment makes any difference.
Thanks for all this guys i really appreciate.
Today i installed a Berning ZH270 OTL...it is a match in heaven for the TAD's and i wonder whether i now need an ET-703
Got the Berning on low and normal feedback for different recordings and the synergy is incredible. Ultra transparent, accurate and musical.
The stands allow the speakers to transmit their full bass energy..so my chesterfields are now rattling apart, not my floor!
I'll still be investigating the options here for the future. Maybe i can experiment first with less expensive tweeters and the pc soundcard option...all sounds really interesting.
thanks again
Richard
Very nice system Richard.
IMHO the benefits of the tad tweeter far out weight the issues of crossovers.
Ian
Steve,
IMHO 5khz is a far more sensible place for a tweeter Xover unless you wanna kill insects and give the neighbour's dorgy a breakdown!
I mean most of us are pushing or past 50. I'd wanna see some real value for money of I was going to spend that sort of money on a tweeter.
Ian
There are currently 1 users browsing this thread. (0 members and 1 guests)
Posting Permissions
Reply With Quote
