Time Aligning Drivers, how to?

Status
Not open for further replies.
@mitchba, kind of nice that you stopped by...and that you wrote the book on Accurate Sound Reproduction using DSP.

It is a highly recommended book, by the way, no I don't have it yet. I will shortly.

Thank you. If you have any questions about the content, feel free to ask.

There will be another Audiolense article at computeraudiophile.com next week and will show how to achieve what you want in about 30 minutes. But it is commercial software.

rePhase has capability as well:

EASY FIR crossover PC based+DRC

http://www.diyaudio.com/forums/mult...on-eq-fir-filtering-tool-209.html#post5116511

Guide to Speaker/Room Correction Using Freeware and JRiver

Have fun!
 
If using PC based media player software, like JRiver for example, then all FIR filtering (e.g. digital XO, time alignment, amplitude and excess phase correction) is performed using JRiver's 64 bit convolution engine, all in the digital domain. There is no additional AD DA conversion step. Two of the links posted above show how. Enjoy the music!
 
ericsquires, cask05, GM, pano, mbrennwa, mitchba:

Yes, the Cornwall is of the type pictured...cask05. I think that is a pic of the cornwall II, which is what I have, as it has the Mid Driver that is ceramic, which the Klipsch folks mention would be a little bit smoother in the mids.

From there, I removed the horns and coated them with automotive undercoat to prevent ringing etc. I probably should have also used help rope and wrapped them horns with it also, but didn't have a source at the time. NOTE: Not Coated the Drivers!

Then I redid the Crossovers.

Finally, I epoxied heavy gauge steel on the back panel. And cut out circles for Mid Driver Magnetic clearance and small holes for the original stuffing and stapeled it back in.

Probably should have used the new stuffings but what I have works.

I've heard guy discussing time alignment here and thought that might be a simple way to do it with these speakers. Obviously not.

=====================================================

So, the next project is the MTM towers with the two side mounted woofers per cabinet. That is where I'll use a Symnet DSP unit and tri amp these speakers. I've got a pic here somewhere that I will post.

This was the original incarnation, I have the boxes less drivers, amps, crossover. Same boxes as pictured, a gift from the designer himself. So I'm resurrecting them. I have different drivers for them, in the process of verifying their functionality, and installing them as wife, life, and child permit.

I'm using the driver's the designer said that I should use for these speakers which now use more modern drivers: Accuton C30-6-024 Tweeters, Seas U18RNX/P H1571-08 Midwoofers, ETON 11-581/50 HEX Woofers.

There were some issues with the ceramic Tweeters and ceramic Midwoofers that had some "techs" blowing them out. It was mentioned that when they go, you hear the sound like an old glass Christmas ornament's twinkle, tinkle, etc and know that is the sound of an $300-$1000 driver just pit the dust.

Eventually, if needed I'll add a custom subwoofer.

The Cross over parameters are shown below. Note: This is the starting configuration for the crossover points.

Thanks for all the info I'll keep y'all posted as I get start to get these going.
 

Attachments

  • LiveSoundStage3.jpg
    LiveSoundStage3.jpg
    57.3 KB · Views: 453
  • 01. CutOff 10 250 2500 inf.jpg
    01. CutOff 10 250 2500 inf.jpg
    85.3 KB · Views: 379
Last edited:
Sync,

You cannot take a multi-way speaker and just "time align it" becuase what you need is time AND phase, and that requires matching eletro-acoustical responses in time and phase.

This is very rarely done with commercial speakers, but they usually follow certain principles. First order on the tweeter, second order on the midrange. And for that you need drivers which will let you do it. 🙂

Best to just let RePhase or Dirac do it for you.

Best,

E
 
...From there, I removed the horns and coated them with automotive undercoat to prevent ringing etc.

All you need to do is clamp the midrange horn mouth to a small baffle. PWK tried the coating with absorptive material and reported that it did nothing to damp the horn ringing. (The PWK report on the K-400 horn--the big brother of the Cornwall's K-600 horn discusses this.)

I've heard guy discussing time alignment here and thoughtthat might be a simple way to do it with these speakers. Obviously not.

Actually that's why I took the time to post those three different pictures of moving the tweeters back on top to time align them. It's that simple.

Many people have done it. You certainly will neither be the first, nor the last. The woofer is less than a half wavelength in front of the midrange horn/driver (K-600/K-55) so Klipsch phase aligns the woofer to the midrange. I can post plots if you're interested.

One of the reasons why the Cornwall does so well with a lot of people is that the loudspeaker is almost time aligned except for the tweeter/midrange interface, but it is phase aligned. The easy relocation of the tweeter takes care of 95% of the issues with time alignment.

Try it.

Chris
 
Cask05, I'm interested in doing that. Are you saying to move it along the top or to move it within the cabinet?

I'd like to see the plots.

Coating the horns seemed to help. They are much deader than they were before.

I was also under the impression that Dick Moore, RIP former member here before he passed on, worked for Klipsch and tweaked the cornwall IIs from what I recall. He had mentioned that he did work on them to improved them.

Cheers,
 
Move the tweeter outside of the cabinet, to on top at the rear of the cabinet, directly over the joint between the midrange driver and horn. Align using either your ears (get someone to move the tweeter slowly while someone else listens for the soundstage to suddenly open up and come into focus, or invert the phase of the tweeter channel and measure using an SPL meter for minimum SPL on axis.

Here is a spectrogram of the Cornwall with corrected EQ to flatten its response to +/-3dB, but also showing the time delay of the midrange horn/driver relative to the tweeter. The delay is about 0.7-0.8ms, which corresponds to about 10 inches of displacement backwards. Try moving the tweeter 10 inches backward, then move slowly back and forth while listening. It should go into and out of focus within 0.25 inch of correct time alignment.

Chris
 

Attachments

  • Cornwall Correcting EQ spectrogram.jpg
    Cornwall Correcting EQ spectrogram.jpg
    421.8 KB · Views: 368
Last edited:
By the way, that 0.8ms tweeter/midrange delay corresponds to about 4-5 full wavelengths at the nominal crossover frequency of 5500 kHz--which is a lot of time/phase offset.

The bass bin crosses at about 800 Hz, so a 1/4 wavelength alignment tolerance says that you could be misaligned by a bit over 4 inches and still be okay in terms of time alignment. The spectrogram shows a misalignment of the woofer to midrange of about 1/4 wavelength, which puts you within polarity swapping distance of being aligned.
 
Chris, it's going to take me a bit of cogitating to understand and comprehend it.

I see the spectrogram, what would happen if the Hz was on the bottom and the time was along the vertical axis? I keep thinking all those peaks are like distortion or ringing on an FFT. But the axis are different.

Pano, now that I think about it, I'm sure that I used the 3M spray coating, their undercoating. It was made to dampen and lessen road noise etc.

Now, does anyone know what Paul Klipsch used? Maybe he used something that hardened too much and rang with everything else, even though it wasn't supposed to?

I haven't thought about using that on the underside of my platter. or I could just use the sorbothane on the top .
 
Don't cut up your Cornwalls, please. Sell them if you must, but find some way to at least store the enclosures.

With horns, there is no real need to mount them in enclosures on a flat baffle. You can time-align them by physically moving them in relation to each other and the woofer. If the distance becomes unwieldy, invert the phase of one or more drivers to get back to a reasonable physical layout.

So, do some impulse testing. See Stereophile's test reports to see what that is about.
Build new woofer cabinets of identical volume as your original Cornwalls. That way you don't need to modify the crossover in any way.

They probably would be wider than the originals, so that your mid and High Frequency (HF) horns will still sit at a reasonable height. Then mount the horns in free air on top of the woofer boxes. Align them forward (less desirable) or back as required.

Virtually all classic Klipsch speakers have a time alignment profile whereby the tweeter leads the mid, and the mid leads the woofer. So that's your starting point; moving those drivers back from the woofer baffle plane as appropriate.
 
Rubberized Automotive Undercoating is a viable damping compound. You can buy good quality stuff in spray bombs for $10 a can from virtually anywhere that sells automotive supplies.

Be sure to buy product that says "paintable" on the can. There are two types, one paintable and one not. The "non-paintable" type contains mostly bitumen, which stinks, in an oily base similar to petroleum oil that you would use in your car's engine. The paintable product is tiny bits of rubber suspended in a water-based or other base paint. Apply in layers as required. It works quite well, you will notice a difference using the "rap test" (knocking on the dried painted item with your knuckles).

It can be top-coated with almost any reasonably appropriate paint. It leaves a stippled texture.
 
Last edited:
@Johnny2Bad and others.

Don't worry I have no intention of breaking down the Cornwalls. I agree that they are too nice to ruin. Yes they are, I like them too. They sound pretty darn good the way they are.

Folks say they like the EMI 15C speaker better then the stock woofer. Why and EMI and not something like one of the JBL 15 woofers? That or say something from our friends...the McCauley's? Heck, why stop there...what might ETON have in a 15incher that would settle nicely in this cabinet?

Oh, you are talking if I make some new cab and do some other horns? Use the paintable stuff. I got it. Maybe I should just to a simple test of each speaker first.

Document the starting point of what I have. Maybe there are issues that I don't know about.

Okay, I've got to start with something simple.

I can't do anything earth shattering, but what should I start with. AND That doesn't require a PhD in physics or EEM.

Maybe something that a simple tech can do?

Cheers,
 
Okay, to help you read the plot, here is the same thing with reversed axes, and zoomed in a bit to show the time offset (i.e., vertical axis...in milliseconds) of the tweeter below the midrange at higher frequencies (i.e., to the right of the graph-->higher frequencies), clearly shown from 5 kHz on up, but also showing output to 1500 Hz and the midrange continuing to output HF energy all the way to 20 kHz, albeit at much reduced SPL. This is due to the fact that Klipsch did not use a low pass filter on the midrange high end in order to curtail its output electrically above the crossover region. I'm told that more recent versions of the Cornwall crossover (within the last 10-20 years or so) actually do use low pass crossover filters now.

IF YOU REMOVE YOUR TWEETER FROM THE CABINET (STILL ATTACHED ELECTRICALLY TO THE CROSSOVER) AND REST IT ON TOP OF THE LOUDSPEAKER, MOVING IT BACK 10.25 INCHES TO ALIGN WITH THE MIDRANGE HORN/DRIVER BUT STILL AIMED IN THE SAME DIRECTION, YOU WILL EXPERIENCE GREATLY INCREASED SOUNDSTAGE AND IMAGING, THUS TIME ALIGNING THE LOUDSPEAKER.

Chris

Cornwall (corrected EQ) Reversed Axes Spectrogram.jpg
 
Last edited:
MOVING IT BACK 10.25 INCHES TO ALIGN WITH THE MIDRANGE HORN/DRIVER BUT STILL AIMED IN THE SAME DIRECTION, YOU WILL EXPERIENCE GREATLY INCREASED SOUNDSTAGE AND IMAGING, THUS TIME ALIGNING THE LOUDSPEAKER.
Hi Chris, Got it, even without yelling. Tweeter & mid horn/driver are aligned at 10.25 inches deep. Still forward firing; (i.e., not facing rearward or to the sides).

I'll pull the tweeters out tonight and see how it sounds.

When I look at that spectrogram, when I see the large red/orange dispersion (for lack of a better word) is this a smear at the given frequency?

Looking at it I assume the woofer is along the 766m and from about 40Hz through about 440Hz? or higher up to about 1500Hz.

So I see the tweeter at 5kHz along the 0 time.

So if I read this the mid horn is off by 766ms @5.47kHz. 1/5.47kHz = 0.0002, 2ms is the time for the wave length. 0.0002/5470 = wave length = 3.342 ^ -8 meters or 334mm. If I didn't screw up the units.

Thanks for the info.

Going to head on over to the Lone Star Audio Fest, always the first week in May, Embassy Suites at North Central and LBJ.

Cheers,
 
Didn't know you were local. Can't make it to LSAF this year. Too much happening at the homestead that we have to take care of this weekend. I usually like to attend, but it seems that everything always happens this time of year...Oh well.

I'm on the Klipsch forum as Chris A in case you want to tap into that forum for more information on Cornwalls, etc. over there. Nothing that I wrote here is new except perhaps the spectrogram of the Cornwall's output. That in particular came out especially clearly in showing the time alignment and crossover issues of the Cornwall II. I use a pair of them for surrounds presently, but hope to replace them with K-402-MEHs as soon as I can get the horns produced.

Chris
 
When I look at that spectrogram, when I see the large red/orange dispersion (for lack of a better word) is this a smear at the given frequency?...Looking at it I assume the woofer is along the 766m and
from about 40Hz through about 440Hz? or higher up to about 1500Hz.

The Cornwall and Heresy are unique in the Klipsch Heritage line in that the midranges are the lagging element in terms of time alignment (in all other cases, its the horn-loaded bass bins that lag the most, followed by the midranges). So you see the effects of that lagging SPL from about 700 Hz up to about 7 kHz. The Cornwall would actually be easier to time align if the midrange horn could move forward by 10.25 inches, but the horn mouth needs to be clamped to a baffle in order to avoid horn ringing. The later model ABS horn material rings a little less than the aluminum midrange horns, but still need the mouth clamped.

I believe the reason why so many people have gotten into applying sticky goop coatings on the back of midrange horns is due to the old 511 and 811 Altec multi-cell horns which really can't be clamped to a baffle effectively to prevent ringing. They tend to ring like bells because of that, I'm told. I'm not a multicell midrange horn enthusiast.

I think you've get the numbers read off the spectrogram chart correctly, as you've mentioned above.

One word about DSP time alignment: its a lot more work to get it dialed in correctly and requires more understanding of the measurements and the effects of the controls in order to get in all under control, but the payoff is larger in that you can correct for more issues like equalization and phase, etc. I don't recommend going that direction until you're comfortable with all of those elements (i.e., taking acoustic measurements with a calibrated microphone, interpreting the measurements, making corrections in the DSP crossover, etc.). That's why the physical alignment of tweeter and midrange is the preferred approach until you're ready to jump in the deep end.

Chris
 
I believe the reason why so many people have gotten into applying sticky goop coatings on the back of midrange horns is due to the old 511 and 811 Altec multi-cell horns....
The 511 and 811 aren't multi-cell horns. They are known as sectoral horns. The mutli-cell horns are the 805, 1005, 1505, 803, 1803, etc. Originally they were tar filled, then later just coated with Aquaplas instead, which helped damp them but not as much as the tar fill. Done right, it works well. Clamping the mouth solidly to a wooden frame also reduces the ringing. Should work as well for the Klipsch horns.
 
Status
Not open for further replies.