"Into The Abyss", Tweeters in low places

[xmax]

I have generally been against running tweeters below 2500hz a point were most SB tweeters start to misbehave.
But now that I have been playing with wave guides and a new breed of tweeters that seemingly play nicely
down to 2K and even LOWER. A well know member of thee audio community told me "never
expect a 1" dome to go under 2K" and I have lived by it, but the times they are a changing.

So over the next few weeks I will be testing larger dome tweeters and maybe even some AMTs
to see what I can get away with, with and without wave guides I will also be experimenting with
minimal part XO's, that these tweeters and wave guides allow.

Some of the "victims" on hand.

SB29SDAC-C004

Wavecor TW030WA11

SEAS Prestige 27TBCD/GB-DXT

Satori TW29R Ring Dome Tweeter

Scanspeak D2604/833000

Peerless XT25SC40-04

Peerless DA25TX00-08

DAYTON RST28F-4

And some smaller tweets that might be able to go lower with the use of a wave guide.

SB29RDNC-C000-4 (and it's beryllium cousin asap)

SB21RDC-C000-4

To match or not match levels? There has been some concern, so I will do both.
Because I often will be comparing another tweeter to the main unit to give
some sort of reference.

The test rig will be a B&K 4007 omni and a Focusrite Scarlett 2, the tweets and WGs will be mounted
to a 19x19 inch baffle counter sunk. Tweeters driven by a hypex UcD 700, with a analog active XO.

What are your thoughts on standardizing the levels and mic distance, since this test is looking
mainly at the tweeters ability to play low (and the resulting THD) I am thinking of keeping the
mic close like 6" to help avoid reflections and background noise. and running the tweets around 100dB.

Speak now or forever hold your peace.

[dcibel]

To match or not match levels? There has been some concern, so I will do both.
Because I often will be comparing another tweeter to the main unit to give
some sort of reference.

If you intend to only measure a single SPL, then yes by all means level match is the "correct" method. A comparison simply cannot be made accurately if the passband SPL varies between tests. Just as you would not want to create subjective opinions by listening to two different speakers at two different volumes, you would not objectively compare how one tweeter performs at 89dB to another at 92dB, that is not useful information. I'd recommend setting the SPL match to something like 92dB at 1m (regaldless of mic distance).

What are your thoughts on standardizing the levels and mic distance, since this test is looking
mainly at the tweeters ability to play low (and the resulting THD) I am thinking of keeping the
mic close like 6" to help avoid reflections and background noise. and running the tweets around 100dB.

For a distortion test I would say the response from the speaker is secondary and perhaps not important at all, so a near mic of 4-6" is just fine as long as the SPL at the mic is within its limits where it doesn't become overloaded. Ideally, it would be nice to normalize the frequency response such that the harmonic information is all relative to the fundamental, so you don't have to consider the frequency response to determine how many dB down each harmonic is.

What AMT's did you have in mind ?

[guitar maestro]

That's exactly what I was gonna say but dcibel beat me to it.

xmax, I'm a big fan of horn-loading AMT's, as I've found out they respond very well. And of course that's why I bought the Beyma TPL-150H's off of you as well. Did you see the results/graphs that I posted with the Hygeia/Airborne RT-5002 with a Dayton H110 horn?

[Billet]

I'm also not a fan of low crossed tweeters. I believe that they are just physically too small to make the required SPL levels at lower frequencies. Displacement increases, which theoretically causes distortion. Problem is, I don't know what is considered a lower frequency. Maybe 2000 Hz is fine? 1500? 2500? I don't know. This should be interesting.

I like your idea of a simple filter, probably first order I guess. I think this really helps see what happens when a tweeter rolls off most naturally.

[BOBinGA]

Test with and without a crossover. One thing that does not get tested much is the effect of a crossover on tweeter HD profile.

[dcibel]

Test with and without a crossover. One thing that does not get tested much is the effect of a crossover on tweeter HD profile.

You can simply apply the transfer function of the crossover to the fundamental as well as the harmonics for a good idea of the result. This method will be fairly accurate up to the point where the distortion is caused by over-excursion, wither by exceeding xmax or hitting mechanical limits. Within the driver's linear range, the harmonic profile will be proportional to the fundamental amplitude for the most part. Personally, I'd rather see a measurement to about 500Hz without any filter in place.

[Pete Schumacher]

Xmax, add the SB26STAC to the list. In a guide I'd be stunned if you couldn't run it to well below 2KHz with the proper 2nd order HP filter.

[isaeagle4031]

Xmax, add the SB26STAC to the list. In a guide I'd be stunned if you couldn't run it to well below 2KHz with the proper 2nd order HP filter.

Pete, the 26stac easily plays to 1.8k without a guide. Crossed 3rd order electrical in a series circuit in my Armonicos. I would suspect in a guide, 1.2k would not be an issue.

[Wolf]

I think you all know what song these words are to....

Chorus:

Yeah- I got tweeters going low places
Sometimes waveguides help
Them match the traces
In the proper way
'Least that's what I say

Well I'm not big on smiling faces
Type of EQ that
Jacks the treble and basses
Oh I got tweeters going low places

------------------------------

At first they didn't sound right
The bass was too light
And the treble was too hot for sure

Turns out initially wrong
Helped me write this song
And experience helps 'em sound pure

And when I get 'em done
I'll shine like the sun
They are voiced and they measure spot on

I'll be really happy with this pair for a while
...But that won't last long!

---------------------------------(Chorus)

You know I build speaks
From dog woofs to mouse squeaks
and you know that this bug has me bad

First it was to save money
That's what I told my honey
But more expensive drivers were to be had

I shelled out for a few
If I only had knew -----------------------(poetic justice on that one, give me a break!)
How far my addiction had gone

Now I just can't help it I sit back in my chair
....and listen to song after song!

Heeheehee,
Wolf

[xmax]

"even after I knew"?

[xmax]

Xmax, add the SB26STAC to the list. In a guide I'd be stunned if you couldn't run it to well below 2KHz with the proper 2nd order HP filter.

Ok I will get some, I see they have a little extra air up top, is that why you want to use them with a WG? If you send me the files I will CNC one, since WG help extend the
the tweeters low freq. ability they will be used in these tests, but the tweeters will also be measured without the WG to see the difference.

[xmax]

Here is a prime example of how a WG can extend the low end of a tweeter, this was done before the "official" standardized
test will begin, but fun stuff, this is the tiny SB21RDC-C0004 ring dome with and without a passive LR2 at 4K with a bored out WG16-25-B

[xmax]

How about if I use the same gain setting always so I get a -10 reading @ 2K, =100dB it seems to me if you are looking for a low
safe XO that would be a fair and equal test. I.E. adjust volume to get 100db= -10dB?

[Wolf]

"even after I knew"?

Fits, but the meaning is different.
Wolf

[tom_s]

Wolf - I never would have guessed you were a Garth fan!