To Heil AMT or not to Heil at $140, that is the question?

I've found that the ESS AMT-1 enables excellent configurations for 2-way loudspeakers using a DSP crossover with first order (IIR) filters without the type of phase/group delay growth that you see with typical higher order IIR filters, especially when looking at the combined step response. The AMT-1 transient response is second to none that I've measured or listened to. I would assume that DSP with FIR filtering will take this combined performance to an even higher level.

I can recommend that the AMT-1s can be used down to their stated 800 crossover point (or even slightly lower) without issues of significant distortion or loss of polar control...and the sound is superb.

Agreed. I am running mine with MiniDSP sitting on top of a sealed HiVi 8.8+
Crossed at 800Hz .. a little notch at 5K. Sounds spectacular.
 
Ok, have to ask what type of music you guys listen to and at what spl crossing @800hz. My experience agrees with the testing that shows increased distortion w/increased spl at frequencies below 1.5khz using a third order filter. Toss on a MatchBox20 cd and let her rip. Post back your impression(s).
 
I'm using these on top of Belle bass bins for surrounds in a 5.1 hi-fi array (multichannel music, movies). Running them at 100+ dB/1m with music doesn't have very high precedence. The 4th harmonic and above that I measure at 100 dB/1m is >50 dB down from the fundamental from 533 to 1180 Hz--whereupon it drops below the noise floor of the room. That's good enough for my needs, especially considering all the other things that these drivers do so well and at their price point.

If you're listening at over 100 dB at your listening position, I can see where you could justify crossing them above 1180 Hz. I'd add a third amplifier and bring in the Belle's midrange horn/driver to accomplish that. As it stands, I find little reason to do so.


By the way, my AMT-1s actually have several response corrections:

attachment.php



Chris
 

Attachments

  • AMT-1 PEQs.JPG
    AMT-1 PEQs.JPG
    25.2 KB · Views: 1,688
Last edited:
Ok, have to ask what type of music you guys listen to and at what spl crossing @800hz. My experience agrees with the testing that shows increased distortion w/increased spl at frequencies below 1.5khz using a third order filter. Toss on a MatchBox20 cd and let her rip. Post back your impression(s).

Mostly classic rock. System is in my office. Does not get super loud but I do crank it from time to time. I'm running 80wpc from a Kenwood surround receiver with direct inputs. Source is a PC through an old M-Audio Audiophile 24/96 card. A little EQ in the MiniDSP and we are off to the races. I previously had a B&G Neo3/Neo8 combo (with passive crossover) sitting on top of the Hivi woofer. I loved that sound .. but the AMT is cleaner and more open sounding. I would consider a small mid-woofer if moving to a 3-way setup.
 
Thanks for the feedback fellas. @Chris, you claim K4 being 50db down @say 600hz. This flies in the face of a couple tests I've seen posted over the years, one of which I linked to earlier. Do you have any supporting data?
I cross my pair @1240hz LR4. Any lower, I can start to hear the distortion with higher spl,s. They sound fine if they aren't pushed @800-1100 below which, btw, one starts to hear the resonance peak. Girl/guitar, Lyle Lovett, long hair sound just fine. RadioHead not so much. It's not just about the spl as it's about the musical content. To my ears, the more going on results in the vocals getting a distorted sound. Around 1180hz 4th order they sound clean even when pushed.
 
Hello Puppet

I just purchased my drivers and I just did a couple of sine measurements to see what my new drivers were doing distortion wise. This is them with the test signal going through my 24db L/R crossover set at about 850Hz. I can't see any rise in distortion in the measurements. They look fine the the distortion level seems constant without a LF rise. Don't let the 100Db throw you that is not a 1 meter measurement about a foot away so in close. One is THD the other has it from 2nd to 5th.

Rob:)
 

Attachments

  • THD @ 100dB.jpg
    THD @ 100dB.jpg
    82.7 KB · Views: 701
  • 2nd-5th Distortion.jpg
    2nd-5th Distortion.jpg
    105.6 KB · Views: 709
Last edited:
The first plot is at 99 dB at 1m (Belle bass bin + AMT-1, crossed 1st order at 650 Hz):


attachment.php



and the second plot at a more reasonable level of 84 dB for my application (the higher-order distortion harmonics are basically below the noise floor of the room):


attachment.php



Chris
 

Attachments

  • Belle bass bin AMT1 at 1m, 99dB.jpg
    Belle bass bin AMT1 at 1m, 99dB.jpg
    176.4 KB · Views: 1,074
  • Belle bass bin AMT1 at 1m, 84dB.jpg
    Belle bass bin AMT1 at 1m, 84dB.jpg
    163.2 KB · Views: 1,029
Last edited:
Here's a plot of another 2-way (Cornwall 15" woofer bass bin, AMT-1) crossed at 800 Hz using 1st order filters, 90dB at 1m on-axis, but having the additional issues of a direct-radiating bass reflex design (much higher modulation distortion and significant group delay growth at the port frequencies):

attachment.php


Chris
 

Attachments

  • Cornwall bass bin + AMT1 90 dB at 1m.jpg
    Cornwall bass bin + AMT1 90 dB at 1m.jpg
    174.9 KB · Views: 1,031
I hope this isn't out of bounds here but you guys have some info important to my work.

I've been following this thred as I am presently finishing up development and drafting a patent for a free swinging ribbon design that has considerably lower distortion and better reliability than typical practical sized free swing ribbons.
In my cross hairs are the AMTs and the small planer offerings on the market now ;) Although I admit I will not be able to compete with the prices quoted here ha.

Anyway the unit under development now is 15 mm wide by 75 mm tall diaphragm. Crossed over with 2nd order LR for -6 db at 1 kHz the unit shows under 0.4% THD from 1 kHz up, at 95 db (referenced at level region) at 1 meter.

Soo reason for this response. If anyone is willing I would love to see distortion measurements of the AMTs (or any other types) under same conditions for compare. Yes I understand the problems with such comparisons

Distort at 95 and 100 db shown at 1/6 oct smooth at 1 meter . Blu 1st, red 2nd, purple 3rd, grn 4th, lite blu 5th

Horizontal dispersion out to 60 degres in 10 deg increments shown with no smoothing, 3ms gate, 18 inches mic distance
 

Attachments

  • 16 mm tight gap 2slit3inh rake .3 .4 .2.jpg
    16 mm tight gap 2slit3inh rake .3 .4 .2.jpg
    127.2 KB · Views: 160
  • 16 mm tight gap 2slit3inh rake .5 .8 .6.jpg
    16 mm tight gap 2slit3inh rake .5 .8 .6.jpg
    129.4 KB · Views: 170
  • 16mm.jpg
    16mm.jpg
    110.3 KB · Views: 155
Last edited:
..that very impressive for such a small radiator. :cool:

Yes it is. She can move without coming apart. Understandably I've been accused of fudging the data.

I look forward to getting the patent filed so I can let this thing out of the bag.

Using a standard flat , corrugated, or embossed ribbon diaphragm of the same size/ mass in the same magnet assembly yields distortion figures around 3-5 % at these volume levels, so were talking an order of magnitude reduction. But perhaps more importantly is the units ability to survive the abuse of these lower crossover points. Killed 2 birds with one stone on this one.

Have another unit using the same technology but ribbon 17 mm wide by 100 mm long that cuts these distortion figures in half
 

ICG

Disabled Account
Joined 2007
Anyway the unit under development now is 15 mm wide by 75 mm tall diaphragm. Crossed over with 2nd order LR for -6 db at 1 kHz the unit shows under 0.4% THD from 1 kHz up, at 95 db (referenced at level region) at 1 meter.

Soo reason for this response. If anyone is willing I would love to see distortion measurements of the AMTs (or any other types) under same conditions for compare. Yes I understand the problems with such comparisons

Distort at 95 and 100 db shown at 1/6 oct smooth at 1 meter . Blu 1st, red 2nd, purple 3rd, grn 4th, lite blu 5th

Well, what about IMD/THD?
 
I look forward to getting the patent filed so I can let this thing out of the bag.

-make sure you look carefully through other patents (all the way back), both US and Europe. And also carefully describe only those portions of technical advance, and if there is more than one area get additional patents for each, because you never know what part of the design someone else is going to "like". ;)
 

ICG

Disabled Account
Joined 2007
-make sure you look carefully through other patents (all the way back), both US and Europe. And also carefully describe only those portions of technical advance, and if there is more than one area get additional patents for each, because you never know what part of the design someone else is going to "like". ;)

That is the best and at the same time probably unfortunately the worst suggestion at the same time. Great because you can probably keep the most innovative patents if others are "like.." and you can probably withstand a lot of 'patent war pressure'.

Unfortunately, patents cost a lot of money, the more you register, the more you'll pay for it. Even moreso if it comes to court. Patent lawsuits can be very costly, especally if not only one but several patents are in question, probably even from different manufacturers. It might very well be, you'll lose just because you've just got not enough money to fight each to the end. Don't get me wrong, I don't want to prevent you from registering any patents but you should think about it from the monetary side, the risk and what's realistic.

The reason I asked for IMD/THD is because exotic drivers with outstanding measurements often got some drawbacks. In case of a ribbon the critical point is the excursion. If you've got a kinda stroke of genius in regards of the suspension, it might work great in one regard but might also be susceptible to IMD or other non-linear distortion - which might be worse than the lniear distortion or increase vastly above a certain sound pressure level or bring crictical burst decay delay with it.
You don't have to explain technical details but radiation pattern and IMD/THD would be very important to get an impression of the driver, much more important than the response because -unlike the response - these can't be fixed by a crossover or a DSP.
 
I'm assuming this from a "DIY" perspective, with likely a one-page w/ CAD drawing and description/effect. I think small entity filing is around $800 US for something like this (in the US).

-defending it is something else altogether. ..but it can be useful for injunctions.
 

ICG

Disabled Account
Joined 2007
I'm assuming this from a "DIY" perspective, with likely a one-page w/ CAD drawing and description/effect. I think small entity filing is around $800 US for something like this (in the US).

-defending it is something else altogether. ..but it can be useful for injunctions.

Well, you'd practically have to produce it yourself since the small entity filing and its benefits are lost once it's licensed. However, patents can (and actually are) filed internationally. You might find yourself in a situation where US juristdiction does not apply but charges and patent claims still do.
 
Well, what about IMD/THD?

Good question, I do not know how to measure IMD. Any advice would be appreciated.

My assumptions here have been based on surface area. Meaning if the better 1 inch domes can get to 1.5 khz with reasonable perf in this area, then a diaphragm with more area shouldn't have a big issue, if its behaving anyway.

The data shown was for the small ribbon, chosen as an extreme to highlight the order of magnitude decease in THD between standard free swinging ribbon diaphragms and mine.

I have some experience with a few patents in past. Last one filed was for a ribbon in fact. US9668057B1

Patent working on now is finishing up after months of writing, re writing on and on. Patent search after search after... ugg! No stranger to this Ha. So far we see no other patents we may be stepping on. Last one filed was same way, approved on first swing.

Im under no delusions about protection and am painfully aware of the fact that if a giant wants to take I will not be able to afford justice, but biz adviser says do it anyway so there goes another 5 grand Ha.

I do produce myself and qualify as micro entity so less expensive to pat but still a lot.


"In case of a ribbon the critical point is the excursion."

This is certainly one of the critical issues and actually its the one that ultimately convinced me to go ahead with the patent. Yes the distortion numbers are exciting BUT it was the fact that the thing can actually move with far more control, reliability, and making no funny noises ;), than standard free swinging ribbons. The darned things can go much lower than even 1 kHz with surprising reliability but Im sure at some point were well into IMD trouble.

"You don't have to explain technical details but radiation pattern and IMD/THD would be very important to get an impression of the driver, much more important than the response because -unlike the response - these can't be fixed by a crossover or a DSP."

Two things I had to do a LOT of work on to get things right were off axis FR and a clean waterfall plot. Initially the components that make this work caused a sharp wiggle around 5K with a long tail in the CSD, AND an off axis rise in response above about 15K. Both have been taken care of. I can post the CSD if your interested. The horizontal off axis uniformity can be seen in the 0-60 degree response plot posted earlier


BTW as a side note, earlier in this thred there is some talk about a 5K wiggle in the AMTs. Something was mentioned about it being some leakage through the frame or something. I dont know of course BUT I suspect from my own work that may not be it. I suspect it could be in the diaphragm itself. When looking at what the ribbons are doing I often do a waterfall at low volume with a small mic at just a few mm from diaphragm and move the mic up and down the ribbons length to see things. Played with a lot of strange designs over the years and just a guess of course but I suspect the trouble is in those small folds and the mechanical properties of such a structure.
 
Last edited:
Looks very nice. Maybe later you can develop a larger version to match or exceed the clean output of the Stage Accompany or Alcons prosound ribbons in the minimum size possible.

Well those are quite different animals. Actually their planer magnetic rather than free swinging ribbons. They look like nice units.
I did experiment for a while with similar designs and for sure once everything is done to a high standard you can achieve a high sensitivity and volume levels from such a design.
However I never got the planer to sound quite as good as the free swinging ribbon in as small a diaphragm. As well I can make a small fee swing ribbon with a very low main resonance way outside its crossover point. In the end for "pro" sound I would go the way these guys did, but for top class in home sound I believe the free swing ribbon has the edge once excursion control, reliability, and distortion issues are taken care of. Its not going to do 120 + db for sure but the little 3 inch long version has no trouble cursing at 95 and 106 peaks are no problem above 1 khz. I suspect it can handle 110 peaks.

I have made larger ones and they can take more but much of my work has been to get the smaller free swing ribbons to behave at lower cross points, and above about 100 mm length the vertical dispersion in average home listening situation is too limited IMO. The 75 mm long version seems about the best overall to accommodate most peoples head heights when seated and the narrow diaphragm has much better horizontal dispersion than any other type of driver I know of.

Heres the CSD at close mike of a unit from a few months ago. Newer units have things down to zero within about 1.5ms all the way down to 1 k. Not sure I can see the CSD of the Stage A. on their site and different measurements systems and conditions make it impossible to compare anyway.
 

Attachments

  • 15mm ribCSD.jpg
    15mm ribCSD.jpg
    237.4 KB · Views: 574
Last edited: