• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Tekton style like tweeter array good idea?

cavedriver

Addicted to Fun and Learning
Joined
Jul 10, 2020
Messages
745
Likes
844
Location
Maryland, USA
I'd just like to see some actual directivity sonograms and not this retro stuff from Stereophile.

In theory a coaxial with different vertical and horizontal radiation could be useful, not to say this is the way to go about it. The quasi coax with the ribbon and midranges created by the german guy measure extremely well.
I think that's what we've been asking for here, and afaik there is no such data available, which is why we're saying we'd like to see one on a Klippel. Obviously, if you find any data please share! :)
 

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,464
Likes
5,442
Location
Somerville, MA
I think that's what we've been asking for here, and afaik there is no such data available, which is why we're saying we'd like to see one on a Klippel.
Maybe Tekton would be amenable to sending one if we were a bit nicer
 

Duke

Major Contributor
Audio Company
Forum Donor
Joined
Apr 22, 2016
Messages
1,665
Likes
4,093
Location
Princeton, Texas
I would not call these Tektons arrays "coaxial", obviously the 2+1 and 14+1 arrays are not even coaxial from a geometric perspective and even in the 6+1 array the top and bottom tweeters are used for a different frequency range the the other 4 ones on the left and right. So we can expect more evenly radiation on the horizontal than on the vertical axis, which is what we see in the 3 stereophile measurements:

2+1: https://www.stereophile.com/content/tekton-design-enzo-xl-loudspeaker-measurements
6+1: https://www.stereophile.com/content/tekton-design-impact-monitor-loudspeaker-measurements
14+1: https://www.stereophile.com/content/tekton-moab-be-loudspeaker-measurements

This has little in common with a real coaxial or fullrange driver.
The Impact Monitor's response looks the most interesting, the horizontal off-axis being imo quite good.

The off-axis measurements are normalized which makes it a little bit harder to interpret them. The apparent roughness off-axis in the top end may actually be smoother than the on-axis response, but because it's normalized to the on-axis response, it shows up in the curves as roughness.

The big dips in the vertical off-axis look to me like comb filtering, which looks worse than it actually is because the off-axis energy is still there, just locally cancelling as those wavelengths pass through one another largely out-of-phase at those particular angles. So the net vertical off-axis energy is going to be smoother than any one of those vertical off-axis curve, and imo the net off-axis energy matters far more than the curve along any one angle.

I don't know why Tekton distributed the energy to the drivers in the circular array the way they did, but maybe it helped them reach an SPL or power-handling target. Anyway imo it makes sense to prioritize horizontal plane smoothness over vertical plane smoothness... and retaining vertical plane symmetry is arguably not a bad thing.
 

Salt

Addicted to Fun and Learning
Forum Donor
Joined
Aug 7, 2023
Messages
761
Likes
449
Location
DE
Maybe Tekton would be amenable to sending one if we were a bit nicer
If they had measured data certainly would have published them asap, not?
But back to Your question: if a certain directivity is the aim, just take a rectangular 'Brett' (AMT, ribbon, electrostat ....) and the width and hights, together with the dimensions of cabinet front plate will determine dispersion pattern. This can be predicted via simple calculation or simulation.
No rocket science or quantum knowledge needed.
 

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,464
Likes
5,442
Location
Somerville, MA
If they had measured data certainly would have published them asap, not?
But back to Your question: if a certain directivity is the aim, just take a rectangular 'Brett' (AMT, ribbon, electrostat ....) and the width and hights, together with the dimensions of cabinet front plate will determine dispersion pattern. This can be predicted via simple calculation or simulation.
No rocket science or quantum knowledge needed.
Any wideband diaphragm of that size is going to go into breakup eventually as frequency goes up, which creates spatial irregularities. At least in the case of Tekton, there are no breakup related directive inconsistency for the most part since tweeters are used.

I just want to see what they accomplished using whatever measurement technique they deem sufficient for home speaker design.

One could argue the spatial resolution of the NFS allows scrutiny of speakers which is beyond what is needed for a consumer decision. Kind of like assigning a score to distinguish between 200 different DAC devices which are audibly identical. Nevertheless as a designer I would love to see what they ended up with - it is somewhat novel idea, although perhaps harkening back to the 70s and 80s, where there was a lot of experimentation with speaker radiation from DCM, Ohm, McIntosh, etc.
 

Salt

Addicted to Fun and Learning
Forum Donor
Joined
Aug 7, 2023
Messages
761
Likes
449
Location
DE
Any wideband diaphragm of that size is going to go into breakup eventually as frequency goes up, which creates spatial irregularities.
Nope.
And directivity is calculable.
 

Duke

Major Contributor
Audio Company
Forum Donor
Joined
Apr 22, 2016
Messages
1,665
Likes
4,093
Location
Princeton, Texas
if a certain directivity is the aim, just take a rectangular 'Brett' (AMT, ribbon, electrostat ....) and the width and hights, together with the dimensions of cabinet front plate will determine dispersion pattern.

Depending on what that target "certain directivity" is, and how much of the spectrum it applies to, various manufacturers use various techniques in pursuit of the directivity and other characteristics they prioritize. Shaded arrays (that is, arrays where not all of the elements get exactly the same signal) are one such technique.
 
OP
R

Randolf

Member
Joined
Jan 9, 2024
Messages
46
Likes
67
Location
Germany
The off-axis measurements are normalized which makes it a little bit harder to interpret them. The apparent roughness off-axis in the top end may actually be smoother than the on-axis response, but because it's normalized to the on-axis response, it shows up in the curves as roughness.
Good point, for a single real coaxial design one would expect very similar horizontal and vertical radiation. When you look at the Stereophile data of 2 coaxial designs


at first glance the horizontal vs. vertical radiation look very different, but once you take the different scales of the diagrams into account they look much more similar as one would expect.
 

ROOSKIE

Major Contributor
Joined
Feb 27, 2020
Messages
1,944
Likes
3,553
Location
Minneapolis
In case you are interested in DIY sets using the Wavecor TW030WA12 you can have a look https://www.lautsprechershop.de/intro/suche_kit_en.htm , just enter TW030WA12 in the "Search used speaker chassis:" and you will see all sorts of kits, e.g.

https://www.lautsprechershop.de/hifi/rhone_en.htm third order high-pass filter at 1000Hz
https://www.lautsprechershop.de/hifi/cursa_en.htm first order high-pass filter at 1650Hz

So it is somewhat questionable why one need an array of these tweeters at all to achieve a low crossover point.
Those benefit from horn loading and that is a beefy 30mm. So in the '2-12 perfect Set' example it might be overkill.
Sensitivity is a factor, IDK what woofers those are but if 2 of them have high sensitivity then extra tweeters make some sense.

I have used (with limited experience)some typical 1" tweeters with low fs and low HD, (individual & stereo pairs)below 1500hrz and basically they tend to sound hard/grainy at higher SPL. The exception was in my car where a 1200hrz HP worked with a 1" dome but I was 3feet or so from each driver, even there I ultimately ended up using 1500hrz active @24db slope.
I have some 30mm Wavecore TW030WA05's that I am hoping to cross around 1200hrz in a pair of speakers. Also some Peerless 32mm Corundrum tweeters. We will see.
I really think the limits are not well represented by Harmonic distortion alone, especially with so much energy in content around 750-1500hrz. I need to start doing multitone testing for my stuff. I have not yet done any at all, my hobby needs to grow a bit after a year or so of no great level of attention.

You can do the calculation of maximum theoretical SPL at each frequency but typical 1" tweets have very limited Xmax and every 3db of SPL gain is a big excursion increase.
That low measured 86db/1m HD of some tweeters mentioned here is happening at a very low excursion for the tweeter, no where near Xmax.
@ctrl shows has a nice example of a simulation with 6 vs 1 tweeter in post #130 & #139.
That excellent example is based off measured Harmonic Distortion charts but I would like to see Multitone affects.
@Rick Sykora in post #129 shows the rising distortion of a tweeter, but that is SINGLE tone, I'd imagine multitone (and some real content) hits much higher distortion at 1000hrz.

@Randolf Your post #112 shows a cool 'numbers game' example based off of Xmax but I think you want to well avoid Xmax so maybe add 25-50% more tweeters for each correlation between number of tweeters needed and frequency response goal. Of course that chart has a robust 120db SPL goal so that might be adjusted as well based on needs.

I don't know. I'm kind of rambling but I think in my opinion @leat in my Tekton Impact monitors, the 6+1 tweeter array is about right for those designs. Now whether the design is a good one or not is clearly controversial, especially some of the designs with even weirder arrays. They are kind of cool to me but I'm a speaker lover and after many years as such unique designs have a place in my space. Would I buy these as a 1st choice, no.

Does anyone have detailed directivity measurements of the so called coax array? I'd be curious how well Tekton got it to work. I find it intriguing that he made it asymmetrical.
Can get an idea from the Stereophile measurements for the Impact monitor.

@ROOSKIE can you help?
@617 Rick mentioned me as I have Tekton Impact Monitors with a 6+1 array.
The Stereophile measurements are the best available right now. They are normalized so more or less depict the directivity.
I took some basic measurements. Everything I did was simple and mainly to confirm they measured as intended and all drivers were operational and in spec as I purchased them used. Just simple Impedance curves, HD, and a few on and off axis horizonal quick measurements & then a little in room measurement from the listening position. That is it. I bought these 15months ago somewhat spontaneously & out of curiosity and really haven't had a lot of time since to investigate them. (they are hooked up right now though and I have been using them the last 2 months). I have not had loads of extra time the last year or so.

I would be interesting to see a pair on the Klippel. Mine are used and I believe are old, maybe even from 2018 when they came out. (serial number are under 0025, & I'm sure he must be far beyond that now) The last thing we need is this pair being submitted.

One thing to note from Stereophile is JA stated "Nevertheless, throughout the measuring, the Tekton's low frequencies sounded rather exaggerated." I have not had this experience, they do not sound exaggerated here, so who knows ---->subjectivity right?

The Impact monitor speakers measure very well IMHO, given the unique design. We have seen so much worse from other unique products and poorly performing very conventional looking products.

Subjectively, I like the sound but I do not 'love' them. I tend to prefer a wider dispersion through the upper mids and into the treble range. The bass also doesn't dig quite as usefully deep as the size would suggest. The vertical dispersion is wonky but I have no way of knowing how that is affecting my subjective experience. They are decent.
 
Last edited:

gnarly

Major Contributor
Joined
Jun 15, 2021
Messages
1,068
Likes
1,514
There is an issue with magnets being so tightly packed in an array like that.
Yes, for sure. It was easy to feel the magnetic field interactions, positioning the 3" faital's in a ring around the central single Wavecore tweeter...(posted in #30).
Twas one of the reasons I went with rings of 5 drivers, instead of 6 ( which would have beneficially tight packed further without changing ring diameter. )
Magnetic interactions felt quite a bit less, 5 vs 6.

All that said, I really have no clue how the field interactions might or might not effect things.

The big issue I have with ring arrays, or arrays in general, is not so much the frequency response (mag and phase) or the directivity reponse.
It's more that they all produce multiple-sources of the same frequency, and that generally only some of the frequencies coherently combine. Just get's back to geometry and center-to-center driver spacing vs wave lengths produced.

Most line arrays have the same problem. At some frequency, the hf/vhf drivers can't achieve the summations that allow them to behave like a line source,
and the drivers start producing multiple paths of the same frequencies. Gives that whooshy, phasey sound, you might have heard at live concerts ...especially when windy.

Measurement wise, one telling property of arrays, is the drop in coherence seen with dual channel measurements. Coherence essentially measures the "sameness" of the reference stimulus signal with the captured measurement signal.

Multiple arrivals from multiple drivers making the same freq, show a drop in coherence. Measurement programs can't correlate reference to measurement, across multiple arrivals. (reflections, diffractions, cause same drop in coherence)
Make for a bit of blur, with softer transients etc, i think.
 
OP
R

Randolf

Member
Joined
Jan 9, 2024
Messages
46
Likes
67
Location
Germany
@Randolf Your post #112 shows a cool 'numbers game' example based off of Xmax but I think you want to well avoid Xmax so maybe add 25-50% more tweeters for each correlation between number of tweeters needed and frequency response goal. Of course that chart has a robust 120db SPL goal so that might be adjusted as well based on needs.
For sure this is just a adhoc rough estimation, but to my own surprise it matches rather well with the known/guessed Tekton crossover frequencies. Staying below the full Xmax and going for a lower SPL both will change the result, doing both simultaneously may result in a similar graph. BTW in Stereophile Impact Monitor (seven 1" (25mm) soft-dome tweeters, two 6.5"-cone (165mm) bass/midrange units) I noticed the crossover freqeuncy is mentioned twice:


"The 6.5" woofers handle frequencies below 1kHz, and the seven-tweeter polycell array handles frequencies above 1kHz."
"The impedance traces suggest that the crossover frequency between the woofers and the tweeter array lies around 1kHz."

To my understanding the first one comes from Eric Alexander himself in the interview and the second one from John Atkinson.
 

Sal1950

Grand Contributor
The Chicago Crusher
Forum Donor
Joined
Mar 1, 2016
Messages
14,499
Likes
17,399
Location
Central Fl
I read some plasma tweeters use argon gas as part of the generation process.
Dr Alan Hill's Plasmatrionics speaker used helium gas and a laser to ionize the gas.
A real "Star Wars" approach that never enjoyed any real success.
But ya gotta admit, they're way up there in the KOOL factor.
Even better than the Tektons tweeter swarms. ;)
614hillplasma.insystem.jpg

614hillplasma.inside.jpg
 

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,720
Likes
7,507
Location
Stow, Ohio USA
For sure this is just a adhoc rough estimation, but to my own surprise it matches rather well with the known/guessed Tekton crossover frequencies. Staying below the full Xmax and going for a lower SPL both will change the result, doing both simultaneously may result in a similar graph. BTW in Stereophile Impact Monitor (seven 1" (25mm) soft-dome tweeters, two 6.5"-cone (165mm) bass/midrange units) I noticed the crossover freqeuncy is mentioned twice:


"The 6.5" woofers handle frequencies below 1kHz, and the seven-tweeter polycell array handles frequencies above 1kHz."
"The impedance traces suggest that the crossover frequency between the woofers and the tweeter array lies around 1kHz."

To my understanding the first one comes from Eric Alexander himself in the interview and the second one from John Atkinson.

Given how elusive EA is, 1 kHz could be intentionally deceptive. Or could refer to an acoustic crossover point. The electric one for the array was modeled on diyaudio and clearly rolls off well above 1000 Hz. In the end, it is one number in a complex crossover design for the array. The bass portion looks rather conventional so adds little value to understanding the array.

As to your original question, at most the array is an interesting gimmick. The Stereophile reviews and measurements do not reveal any major advantage. For that matter, one Stereophile reviewer owned Tektons and moved on. @ROOSKIE owns and says they “decent”. So nothing to indicate anything that the array contributes that makes them sound way better than more conventional speakers. The Tekton speakers are not awful and their higher sensitivity is going to make them stand out more than most. However, can find higher sensitivity speakers from JBL and others. So find little evidence here for anyone to go out of their way to buy Tektons.
 
OP
R

Randolf

Member
Joined
Jan 9, 2024
Messages
46
Likes
67
Location
Germany
Given how elusive EA is, 1 kHz could be intentionally deceptive. Or could refer to an acoustic crossover point. The electric one for the array was modeled on diyaudio and clearly rolls off well above 1000 Hz. In the end, it is one number in a complex crossover design for the array. The bass portion looks rather conventional so adds little value to understanding the array.
Yes, when Eric Alexander gives statements like

"...especially with the model like the Ulfberht where it plays down into the 200 Hertz region"
"...those tweeters dig down to believe this about 400 Hertz"
"...all combined to do the mid-range down to 400 act technically these devices start kicking in at 270 280 they start contributing to the frequency response at 270 Hertz"
"...for example the Moab and that one's a very popular speaker we've got these tweeter contributing to the frequency response at 270 Hertz and not only are they doing the mid-range that they're doing a section of what I would define clearly as mid bass"

(all from his videos on the Tekton website)

I find this highly misleading, but since these statements are not precise and use terms like "plays down into", "dig down", "down to", "kicking in at" or "contributing to the frequency response at" they are not completely wrong but I would say completely irrelevant since I guess the vast majority of the output at that frequencies comes from the woofer instead and the tweeter array contributes at -X db with X being a pretty big number. The 2 more or less confirmed crossover frequencies of 772 and 1000 Hz do not really match with the above statements but at least sound somehow feasible and match with impedance measurements. I guess Tekton owners can easily proof that by playing a 200, 270 or 400 Hz sine wave and check which drivers contributes most at this frequencies.
 
Last edited:

cavedriver

Addicted to Fun and Learning
Joined
Jul 10, 2020
Messages
745
Likes
844
Location
Maryland, USA
just thinking, if someone's dead set on doing this idea as a designer, why not use a ring of BMR's? They're Fs is much lower, they are happy to be mounted without faceplates, allowing them to be packed closer together, and if you brought the lower crossover point down the lobing would be reduced. With a tweeter like the Scanspeak D3004/6040, which has a faceplate only 62 mm across, you could even get within the ~3" you would want to make a D'Appolito with the crossover at ~2 kHz. Note that the 2" BMR does have a higher moving mass, as in 1.0 g vs. 0.25 g for the BE tweeter, so if we're talking about the masses of violin strings the BMR is too heavy for Eric's patent perhaps (haha). Btw, the TEBM35C10 2" BMR is on sale at PE for just $8 right now so Eric can turn the M Lore's into a six-driver array on a budget. Eric A., you owe me for my unsolicited design services...

2" BMR, with 62 mm tweeter, in a support frame, should get under 3". Note that this array would look very much like the Monitor Hyphn array and perhaps some of the past McIntosh arrays, which both use 2" drivers, and flat or nearly flat surfaces in the case of the Hyphn.
297-216_HR_0.default.jpg
 
Last edited:
OP
R

Randolf

Member
Joined
Jan 9, 2024
Messages
46
Likes
67
Location
Germany
Maybe one example how Eric Alexanders interprets the Stereophile measurement (https://www.stereophile.com/content/tekton-design-impact-monitor-loudspeaker-measurements) from 07/2018 of the Impact Monitor since it is about the "polar response" of the tweeter array and I find it misleading:

In his video
he first talks about narrow vs. wide polar response, at 140sec he claims

"...if you go to stereophile review you will discover that this array is in fact the best averaged polar response on the 30 degree window in the history of stereophile magazine measurements up until that time and for the bookshelf stand category...".

In the Stereophile review we read

"This speaker offers a superbly even on-axis output in the midrange and treble, with only a couple of small dips visible in the mid-treble....The speaker's horizontal dispersion is commendably even below the cursor position at 3.8kHz. The tweeter array becomes gradually more directional as the frequency increases, but in a well-controlled manner up to 15kHz, above which the speaker's output drops very rapidly to the sides of the central tweeter axis....The Impact Monitor's vertical radiation pattern suggests that the speaker needs to be listened to within a narrow window centered on the central tweeter axis if the midrange balance is not to sound colored."

To me this is misleading again. Eric Alexander suggest by first talking about wide polar response and than referring to "Fig 3 on central tweeter axis at 50" averaged across 30° horizontal window" a wide polar response. However Stereophile comes to the more or less opposite conclusion by referring to "Fig 5 vertical response family at 50", normalized to response on central tweeter axis".
 

p.trianon

Member
Joined
Apr 9, 2024
Messages
18
Likes
11
Yes I would also really like to see a Klippel NFS measurement of the 14+1 and 6+1 array, but I guess for very well known reasons nobody currently likes to measure any Tektons at all;).
I would send some for evaluation but I doubt that will happen at this point in time.
 

cavedriver

Addicted to Fun and Learning
Joined
Jul 10, 2020
Messages
745
Likes
844
Location
Maryland, USA
Rick mentioned me as I have Tekton Impact Monitors with a 6+1 array.
The Stereophile measurements are the best available right now. They are normalized so more or less depict the directivity.
I took some basic measurements. Everything I did was simple and mainly to confirm they measured as intended and all drivers were operational and in spec as I purchased them used. Just simple Impedance curves, HD, and a few on and off axis horizonal quick measurements & then a little in room measurement from the listening position. That is it. I bought these 15months ago somewhat spontaneously & out of curiosity and really haven't had a lot of time since to investigate them. (they are hooked up right now though and I have been using them the last 2 months). I have not had loads of extra time the last year or so.
Any chance you would open yours up and take some pictures or make drawings of the crossover? Curious what EA is actually doing inside these. For example, are all 6 getting the same signal or is he truly adjusting for position in the array. He suggests he's doing something different depending on the driver position but I think he's bs'ing.
 

Duke

Major Contributor
Audio Company
Forum Donor
Joined
Apr 22, 2016
Messages
1,665
Likes
4,093
Location
Princeton, Texas
Curious what EA is actually doing inside these. For example, are all 6 getting the same signal or is he truly adjusting for position in the array. He suggests he's doing something different depending on the driver position but I think he's bs'ing.

Stereophile's measurements of the Impact Monitors show very different off-axis behavior for the array in the horizontal and vertical planes. The horizontal off-axis behavior is smooth which to me suggests a good blending of the drivers in the horizontal plane, but the vertical off-axis behavior shows lobing effects. My guess is that the middle three drivers in a vertical line receive the same signal or nearly so, while the two on either side are rolled off with a slope that results in good pattern blending with the center drivers.

 
OP
R

Randolf

Member
Joined
Jan 9, 2024
Messages
46
Likes
67
Location
Germany
Any chance you would open yours up and take some pictures or make drawings of the crossover? Curious what EA is actually doing inside these. For example, are all 6 getting the same signal or is he truly adjusting for position in the array. He suggests he's doing something different depending on the driver position but I think he's bs'ing.
To me the crossover diagram from the diyAudio forum I referenced in my initial thread post looks plausible (it just lacks the coil parameters). It also matches with the Stereophile vertical vs. horizontal measurement. Please check from legal / copyright perspective before posting anything if it is ok to post such details to avoid any trouble. I am neither sure how this is in the USA nor in Europe.
 
Top Bottom