Best Cheap Dayton AMT Tweeter?

[Ben_Tech]

I assume the dynamic compression that Pallas is talking about is specific to the AMT tweeter used in the GoldenEar speaker and may not necessarily apply to the Dayton / Airborne AMT tweeters. Also I assume this issue even if present in the Dayton / Airborne would not apply in a small room situation as volume levels tend to be reasonable.

 

[Pallas]

I think it's highly unlikely. It's not thermal compression, not enough power input for such a limited bandwidth (..see John K's research into this).

I think you're simply not understanding the data, which are differences in FR between a sweep at 70 dB and a sweep at 90 or 95 dB (depending on graph).

I simply don't buy any explanation for the NRC linearity-with-level results except the obvious one: garbage tweeter.

Diffraction is just not a sensible explanation for why the FR is so different when swept at 70dB and when swept 20dB louder.

Worthless garbage tweeter (along with maybe gutter-grade crossover components) actually fits the data.

I should note that I've seen this sort of thing from diffraction without a grill and normal tweeters as well.)

Please post the data you have of level-dependent compression at the bottom of the tweeter's passband caused by grill diffraction.

 

[ScottG]

I assume the dynamic compression that Pallas is talking about is specific to the AMT tweeter used in the GoldenEar speaker and may not necessarily apply to the Dayton / Airborne AMT tweeters. Also I assume this issue even if present in the Dayton / Airborne would not apply in a small room situation as volume levels tend to be reasonable.

If correct, yes - there is no guarantee that the Dayton's will be similar, in fact there is no guarantee that the same driver (GoldenEar) from a different production run will be the same.

AMT's from other sources have sort of had a "hit and miss" quality with regard to production quality and resulting performance. Even Mundorf had some production problems at an earlier point in their development.

 

[ScottG]

I think you're simply not understanding the data,

..Please post the data you have of level-dependent compression at the bottom of the tweeter's passband caused by grill diffraction.

No, I understand the data.

-and you could be correct, I just think the data (and the conditions underlying that data) - supports an alternative explanation.

(..and of course if you are correct, this doesn't mean that Dayton drivers are necessarily flawed as well.)


I've not had that data for over decade (..hint: it was on a Windows 98 machine) - it was when I was experimenting with an "Avalon'esq" baffle (..no felt like Avalon uses). Above and below the dominate diffraction peak the soft dome had this problem.. (which it didn't when on a large test baffle). What it did *not* have - was a problem well above this area, which is unlike the Triton One (..but the dome center was not so well coupled to the baffle as the Triton One's AMT is.)

Hmm, makes me wonder if Earl's "hole" on his OS profile waveguide ever changes with very low amplitude like 70 db vs a 1 watt 1 meter response..

 

[Ben_Tech]

I noticed user milosz has some nice things to say about the Dayton AMT, although it's the bigger one that requires a separate enclosure to curtail its backwave.

Here are the pertinent quotes:

http://www.diyaudio.com/forums/multi-way/265046-anyone-using-fountek-neocd-2-0-5-ribbons.html

The sound was quite different from the Fountek ribbon. There was more detail in the mids / lower treble, although I think the Fountek has a little better upper treble "bite" and extension. Even so, the AMT sounded more realistic in this application than the Fountek ribbon, mostly I think because it could be crossed over lower. Also, I think the AMT is a bit more "dynamic" in the lower treble.

The AMT sounds different again, more dynamic and "bodied" but with some of the delicacy of a ribbon. So far with my design here I am liking the AMT

http://www.diyaudio.com/forums/multi-way/264096-build-2-way-biamp-integrated-source.html

the AMTPRO-4 has a very 'dynamic' sound and works quite well down to the 1500 Hz crossover point I used here.

I am REALLY liking this Dayton AMTPRO-4. Once decently balanced with the Satori woofer, I am appreciating it's very transparent sound- really really good.

Anyone think the smaller Dayton AMTs with sealed backs would sound as good? They sure are easier to integrate.

 

[multitask]

bah..ribbons. 4 out of 5 beaten by a $29 dome tweeter.
Zaph|Audio

"I often hear people babble on about the shimmer, sparkle and air of ribbon tweeters. These tests show where these "special effects" are coming from - distortion."

 

[ScottG]

"..mostly I think because it could be crossed over lower.."

Anyone think the smaller Dayton AMTs with sealed backs would sound as good?

No, because they can't cross-over as low as the large AMT 4.


Again, most of the sound is all about the Alpair without a tweeter (and crossover) that can extend into the upper midrange.

Here is an alternative:

-4 driver vertical array of these:

https://www.madisoundspeakerstore.c...-paper-cone/?gclid=CP3b3Y2gpsUCFQoKaQodfQEALg


-they should have a more "electrostatic-like" sound, and can be crossed over low without problem. (..on the other hand they will also need an enclosure.)

 

[Ben_Tech]

"..mostly I think because it could be crossed over lower.."
No, because they can't cross-over as low as the large AMT 4.

Right, but within their range...... Milosz needed a tweeter to extend quite low because of the woofer he was using. He said the woofer has problems playing above 2K due to coloration of the sound.

Just wondering if anyone has compared the sound of the different AMTs.

The best I could find was this:

Finally, I compared the AMTPRO-4 to the Dayton Audio AMT Mini-8 tweeter. These had all the extension and refinement of the AMTPRO-4, but the only thing lacking here was that the Mini-8 could not be run in a dipole fashion. And I would never have imagined the Mini-8 was lacking in any way, unless I had not just listened to the dipole configuration of the AMTPRO-4.

[Review] DIY supertweeters for Electrostat speakers

(He goes on to say that "in comparison to the Mini-8s the AMTPRO-4s rendered a considerable increase in dimensionality and realism" but I think that only applies to the dipole setup he was talking about.)

So I thought it reasonable to assume that within their range the tweeters in between the mini-8 and the pro-4 would sound quite good as they cost twice, three times what the mini-8 costs.

 

[ScottG]

So I thought it reasonable to assume that within their range the tweeters in between the mini-8 and the pro-4 would sound quite good as they cost twice, three times what the mini-8 costs.

I think its a reasonable assumption, but it still requires proper integration with your Alpair.

Note however that the drivers are still fairly directive (horizontally, they are obviously very directive vertically).


If you can deal with higher non-linear effects (which can be reduced via the high-pass filter) there is also Hi-Vi_RT 1.3:

It has significantly wider horizontal dispersion and more extension.

http://www.zaphaudio.com/temp/Hi-Vi_RT1.3-FR-offaxis-horiz-0-15-30-45-60.gif

http://www.zaphaudio.com/temp/Hi-Vi_RT1.3-HD.gif

-put a 1st order electrical high-pass near 20 kHz, and "blend" the low pass filter on the Alpair to match it.

Forum Member: Wolf_Teeth can tell you more about this driver's subjective qualities.

Nice thing is that it's discontinued and as a result is on sale:

https://solen.ca/pub/index.php?cata...eau1=1&niveau2=1&niveau3=3&s1=2&s2=1&s3=2&s4=

 

[5th element]

??-????????

Wants a 4th order acoustic 4kHz xover ideally imo. You could possibly squeeze 3k out of it but that'd be pushing it a little.

 

[ScottG]

??-????????

Wants a 4th order acoustic 4kHz xover ideally imo. You could possibly squeeze 3k out of it but that'd be pushing it a little.

That's more of an ideal (for the AMT, not the mid-driver), the problem is that the Alpair won't work well with that off-axis.

Just at 30 degrees at 3 kHz the Alpair has almost 6-7 db of loss when compared to the 0 degree axis.

The Alpair is going to require a non-standard crossover - one with a lot of overlap between the drivers (..to strive for any sort of horizontal polar uniformity).


BTW, nice link - we can see different voltages to get a better sense of linearity problems at higher input levels. While there are some deviations, they aren't what the GoldenEar Triton is showing (between 90 and 95 db). Could be a different tweeter of course, but it's moot given the poster is requesting information on the Dayton drivers.

-call me crazy, but for the price I like the Hi Vi driver better (at least objectively when focusing on "planar'esq" tweeters at a lower price-point).

 

[Ben_Tech]

Thanks 5th for the link I never would have found that.

Google translate:

The high-frequency transmitter Hale from the American company Dayton Audio.

Pros:
- High sensitivity (95-95,5 dB)
- After a simple flat frequency response correction (+/- 0.5 dB at most in the range 5,5kGts-18kHz)
- Low harmonic distortion after 4,5kGts

Cons:
- Application range limited below 4,5kGts below which greatly decreases the frequency response and distortion grow

The amplitude-frequency characteristics of the axle mounting "flush" with the voltage 2.83V, 4B 5.6V. AFC modes are for uncorrected and corrected

I guess this fellow uses some kind of correction circuit to flatten the response peak... if I'm reading it right....

Also thanks Scott for the Hi Vi info I sent Wolf a PM asking about it (and his opinion on the AMTs since he's used these).

Interesting that you think a first order crossover at such a high frequency would be best.... should be a low component count I would guess.

Once I decide on the tweeter I'll fire up the mic and REW and get some data to use in PCD. I'm no expert on crossovers but I'll run it by you all in a separate build thread once I get that far.

If anyone else wants to pipe in on the mid priced AMTs or if the more expensive and difficult to implement pro-4 is the only way to go, please do so!

 

[OlavHaal]

Good results.....

Now that the Dayton Air Motion Transformer (AMT) tweeters have been out for a while, has anyone listened to the various sub-$100 USD models and chosen a favorite?

Mini-8 $30 USD - crossed at 5K

AMT2-4 $70 USD - crossed at 3500

AMT3-4 $100 USD - crossed at 3K

Obviously the mini-8 is much lower priced than the others. Are the others worth the premium..... especially given they would all be padded down to around 87 dB in the crossover to a typical midwoofer.

Anyway, the good reviews on the treble of the B652-AIR have me wanting to go AMT for my next design.

Hi to all on this tread.....

Read the TNT-review - and ordered 2 set with AMT - mini8 (one pair for me, and one pair for friend). And the result on my integration on my FHXL-Alpair10.2 is MUCH better than expected. I did stuff my DIY-build FHXL a little to much, so the treble maybe was a little to low/disclosed. Instead of removing some stuffing I ordred these and with 1 uf( app 12000 Hz) they is giving life and sparkle to all music. But the best result is with 24/96 stuff from B&W music-club, I-Trax + etc! I will try 1,5 and 2,2 later this week to see if there is better results with those = 9-11000 HZ, or maybe down to 6500 Hz.
The joy is in the searching - and I will maybe try them on my Alp12 with my homemade SIT(LÀmp) and my Luminaria(SIT)preamp?

Best to all

Olav

 

[wolf_teeth]

That's more of an ideal (for the AMT, not the mid-driver), the problem is that the Alpair won't work well with that off-axis.

Just at 30 degrees at 3 kHz the Alpair has almost 6-7 db of loss when compared to the 0 degree axis.

The Alpair is going to require a non-standard crossover - one with a lot of overlap between the drivers (..to strive for any sort of horizontal polar uniformity).


BTW, nice link - we can see different voltages to get a better sense of linearity problems at higher input levels. While there are some deviations, they aren't what the GoldenEar Triton is showing (between 90 and 95 db). Could be a different tweeter of course, but it's moot given the poster is requesting information on the Dayton drivers.

-call me crazy, but for the price I like the Hi Vi driver better (at least objectively when focusing on "planar'esq" tweeters at a lower price-point).

The AMT's from Airborne aren't bad at all, and I've heard the Pro and Mini-8 from Dayton. I think I prefer the Hygaeia/Airbornes, but that may very well be implementations. The 5002 needs response shaping, has a sensitivity of 100dB, and really is only good to about 3.4k. I used it in the Nephila. The 4001 similar to the Dayton is a known quantity that Paul Carmody used in his Spitfires, and sounds quite nice too. The little sibling of both is only good to about 7k, and almost not worth it except in terms of a supertweeter for off-axis response. I used those planars in the Fenghuang.

BUT- the planar RT1.3WE will likely complement the MA driver better than the others. IMO.
Later,
Wolf

 

[ScottG]

..BUT- the planar RT1.3WE will likely complement the MA driver better than the others. IMO.

Later,
Wolf

..I think that's what Ben needed to "hear".


https://www.madisoundspeakerstore.com/ribbon-tweeters/hi-vi-rt1.3we-isodynamic-tweeter/

 

[5th element]

While there are some deviations, they aren't what the GoldenEar Triton is showing (between 90 and 95 db). Could be a different tweeter of course, but it's moot given the poster is requesting information on the Dayton drivers.

I am not sure, it depends on how the measurements are taken, if they are done with a very quick stimulus then I'd expect their to be less deviations between the different drive levels. If however the measurements are done after a 'warm-up' period as it were, then I'd expect their to be more of a difference. Clearly this kind of thing doesn't affect the distortion at high frequencies. Still it does give some more food for thought. The Dayton doesn't look half bad if crossed over appropriately and it certainly looks better than a number of other ribbon or planar drivers.

 

[ScottG]

I am not sure, it depends on how the measurements are taken, if they are done with a very quick stimulus then I'd expect their to be less deviations between the different drive levels. If however the measurements are done after a 'warm-up' period as it were, then I'd expect their to be more of a difference. Clearly this kind of thing doesn't affect the distortion at high frequencies. Still it does give some more food for thought. The Dayton doesn't look half bad if crossed over appropriately and it certainly looks better than a number of other ribbon or planar drivers.

It's a standard "sweep" without gating.

Considering the crossover, both its electrical "slope", and its resistive components - they could do the test over and over and it's still highly unlikely to warm-up the VC of the tweeter (..or any driver for that matter). But I'd seriously doubt that NRC spends it time on 20 sweeps (..and even then I'd doubt under these conditions that 20 sweeps would cause a problem, there will be some "cool-off" in between each test - and an AMT has a lot of VC surface area that can cool quickly). Note: IF they somehow pushed a tweeter that hard, they'd likely do it to all the speakers they test - and it would show up that way. Additionally, thermal effects are directly "tied" to inductive problems that would show-up in the non-linear effects as a substantial "jump" in the problem area well beyond the normal rise for the voltage increase.

Thermal compression is a common misconception (even among "pro.s"):

VC Heating

IMO, the real problem is diffraction near the driver - specifically grouped "sources" with phase problems creating a slight pressure reduction as amplitude increases. This effect is perhaps enhanced visually because of the 2 meter testing (..and also a bit "smoothed").



Note that Be Scan Speak tweeter in the Magico S5 has this problem:

SoundStageNetwork.com | SoundStage.com | NRC Measurements: Magico S5 Loudspeakers

Is this likely a thermal effect? (Hint: I know what Scan Speak engineers would tell you what to do to yourself if you think it's either thermal or physical compression.)

So why is it happening? (..and remember that this one doesn't have a grill when tested.)

-Because those particular $30,000 loudspeakers have a tweeter that's set to far into the baffle.

 

[Greebster]

The 60 odd degree phase shift through the mid/tweeter crossover point is a serious audible problem in my book. Takes a bit more than being recessed a bit to mess up the phase that much imo.

If the phase wasn't an issue would think they tried crossing the tweeter too low... and recessed to much

 

[5th element]

The 60 odd degree phase shift through the mid/tweeter crossover point is a serious audible problem in my book. Takes a bit more than being recessed a bit to mess up the phase that much imo.

If the phase wasn't an issue would think they tried crossing the tweeter too low... and recessed to much

The phase shift is electrical and it has absolutely nothing to do with the acoustic phase and nothing to do with the overall sound of the loudspeaker. The reason why the electrical phase has been included is because it indicates how difficult the loudspeaker is to drive, but providing the amplifier is capable then it has no bearing on how the speaker is going to sound.

The tweeter looks like it's crossed over at around 2kHz and the distortion chart indicates that it's operating within a range that is comfortable with.

Magico's Be tweeter looks like it's heavily based on the scanspeak Be dome. It's certainly surprising considering Scans heritage etc, but the standard Be dome does not use ferrofluid in the gap and this could be what happens without it. The Focus Audio Prestige FP90 BE Loudspeakers also exhibit a similar issue with compression and they too use a version of the scan Be tweeter.

I do not know what Scott is talking about with regards to the recessed baffle as the tweeter in the S5 is simply flush mounted on an ordinary flat baffle.

 

[ScottG]

..Magico's Be tweeter looks like it's heavily based on the scanspeak Be dome. It's certainly surprising considering Scans heritage etc, but the standard Be dome does not use ferrofluid in the gap and this could be what happens without it. The Focus Audio Prestige FP90 BE Loudspeakers also exhibit a similar issue with compression and they too use a version of the scan Be tweeter.

I do not know what Scott is talking about with regards to the recessed baffle as the tweeter in the S5 is simply flush mounted on an ordinary flat baffle.

The tweeter is sourced from Scan Speak, as is the mid-range I believe (..though using Magico's diaphragm), as are the bass drivers. The drivers (not the tweeter) in their more expensive line are Morel - and actually are inferior. Only the mid-range is really "custom".

My guess is the tweeter is recessed by about 2 mm's. (..again, the behavior is somewhat exaggerated due to the 2 meter measurement). The Focus Audio Prestige FP90 BE Loudspeakers have nearly the same problem - actually with a bit of a deeper recess.

Note: that the recessed tweeter may be on purpose - to achieve a certain measured response at the typical 1/watt 1 meter measurement (..neglecting the pressure loss at higher levels within a very limited "on-axis" range).