Heil Air motion transformer for Cornwalls

[Khornukopia]

24 minutes ago, consistent said:

I have been so impressed with the AMT mod with the Cornwall and with Chris's measurements and observations ...

 

The AMT tweeter has a good sound.

[ClaudeJ1]

46 minutes ago, Khornukopia said:

 

The AMT tweeter has a good sound.

I have 3 pairs. I think I agree with you!!

 

[Chris A]

8 hours ago, consistent said:

Love to know what I have actually done from a technical point.

The obvious things that are affected when using different (higher order) crossover filters is that the width of the crossover interference band (in Hz) is smaller and steeper sloped, but also that the phase shift at the crossover region is also greater (a downside of the tradeoff).  This phase shift can be partially mitigated by moving the AMT-1 back on top of the bass bin to take out some of the overall phase shift/time lead induced by the electrical filters, leaving a local disruption in phase response around the interference band.  This is all clearly seen in REW measurements. 

 

Some things that are not often identified from your account include the effect of greater horn loading of the AMT-1 by moving it back on top of the bass bin, thus effectively lengthening the bottom section of the AMT-1/bass bin "horn".  @Rudy81 very recently reported on this effect in his listening room (and I'll let him explain the subjective effects of that one).  Apparently, the trough-type of horn formed by the magnet and its structural elements as well as the polymer case that encloses the magnet assembly is a bit undersized to fully horn-load the AMT-1's diaphragm down below about 2 kHz--to 800 Hz or slightly lower (610 Hz in my measurements--with significant horizontal and vertical polar coverage widening below 1 kHz).  So providing a little more diaphragm support by moving it back away from the front edge of the bass bin provides better horn loading on the front of the AMT-1 (but not the rearward-going backwave). 

 

The sides of the magnet structure can also be extended another 6-8 inches (15-20 cm) to increase the lower frequency horn loading of the diaphragm, thus increasing the on-axis SPL below 2 kHz, but also lowering the harmonic distortion of the AMT-1 above ~95-100 dB (1m) on-axis.  This indicates that the present magnet size/shape is actually undersized to fully support the horn loading effect across the lower frequency range of the AMT-1, but that this undersized horn was probably deliberately chosen in order to trade for a smaller AMT-1 profile (and lower weight/cost driver overall), but also to achieve flatter on-axis SPL below 2 kHz, with the knowledge that this also limits the SPL output of the driver below 2 kHz to SPL values under 100 dB (@1m). 

 

Other effects of moving the AMT-1 back on the top of the bass bin include decreasing the strength (SPL) of the backwave below 2 kHz and widening the effective polar coverage of the backwave.  This is more associated with the subjective listening effects of the driver than the direct-arrivals from the front of the horn.

 

These are a few of the effects that can be measured or subjectively heard when you make the changes that you indicated.  It's a lot more complicated than meets the eye initially, and I'm sure that more measurable and subjective effects can be described than those listed above.

 

Chris

[Rudy81]

Thanks to @Chris A I became aware of the AMT driver while discussing REW measurements.  Out of curiosity, I ordered a pair and have been playing around with them.  I have been running all sorts of measurements on them to see what they can do when mated with my bass bins.  Some waveguide prototypes indicated an increase in lower end performance.  This is a very simple build with just a pair of 'wing's added to the sides.  Adding a top to the driver audibly killed the spacious, airy sound of the driver.

 

As soon as I get time I will formally build the waveguides.  I have also decided to stack a pair of drivers on each speaker. In theory, this should have a net benefit vs. a single driver. 

 

If you guys are interested, I can post my measurements and results as I go along. 

 

So far, I have found the many reviews I have read to be correct.  This is an incredible driver with superb upper end resonse.  It does lag a little bit in the 1-2khz region, right where I wish to crossover.  However, my tests indicate that with the addition of the waveguides, I can improve the lower end response.  As @Chris A will attest, I am no expert on this stuff and have relied quite a bit on Chris to help me along the way. Most of you are better versed to understand much of this, so if I make errors, please point them out.

 

I tried two types of wave guides.  The best performer was the one with the front 'wings' only.  I also measured the effect of sticking an AMT inside an Oris horn.  Results were predictably awful.

 

 

[Khornukopia]

1 hour ago, Rudy81 said:

If you guys are interested, I can post my measurements and results as I go along. 

 

Please do.

[Rudy81]

It will be a few days before I can get back to the important things like acoustics. For now, work calls.

[ClaudeJ1]

26 minutes ago, Rudy81 said:

It will be a few days before I can get back to the important things like acoustics. For now, work calls.

I thought about building an OB speaker using the "X" on those Heils. I have 3 pairs and I'm certainly headed in that direction also.

[Rudy81]

Sounds like a good opportunity to cooperate and compare notes.  When I ran the first prototypes, I found the X setup did not perform quite as well as the front waveguide only idea. However, before I go in for a full build, I want to run those tests again with a more critical eye on response.  Compared to my other builds, this will be a piece of cake, but I want to optimize the waveguide.   Let us know what you find out in your experimentation.

 

I found that the X configuration has a -2dB dip centered around 850hz.  Otherwise, it is the same as just the front waveguide design.

[Khornukopia]

2 hours ago, Rudy81 said:

I tried two types of wave guides.  The best performer was the one with the front 'wings' only.  I also measured the effect of sticking an AMT inside an Oris horn.  Results were predictably awful.

 

You got me starting to wonder about the idea of finding an existing horn or waveguide with suitable dimensions, and sawing off the small end, so that it might work as an extension on the AMT.

 

 

 

[Rudy81]

17 minutes ago, Khornukopia said:

 

You got me starting to wonder about the idea of finding an existing horn or waveguide with suitable dimensions, and sawing off the small end, so that it might work as an extension on the AMT.

 

 

 

 

I had also considered that, but haven't found anything that would work.  The horizontal angles of the driver itself measure 45 degrees plus or minus one or two.  I have not measured the angle at the top and bottom. That would be a time saver and may even work better, but needs to be a pretty big horn to start with.  The driver is 6" tall 6.75" wide.

[Khornukopia]

13 minutes ago, Rudy81 said:

The driver is 6" tall 6.75" wide.

 

Yes, but the area that you applied your tape is 4.25" wide.

[Rudy81]

14 minutes ago, Khornukopia said:

 

Yes, but the area that you applied your tape is 4.25" wide.

 

I see what you are trying to find. Hawthorne Audio, no longer around I believe, built a horn for the Heil AMT.

 

[Khornukopia]

6 minutes ago, Rudy81 said:

I see what you are trying to find. Hawthorne Audio, no longer around I believe, built a horn for the Heil AMT.

 

That looks good. I shall learn from their design.

[Rudy81]

The Beyma TPL 150-H also uses an AMT in a horn.

https://www.usspeaker.com/beyma tpl150h-1.htm

 

[Khornukopia]

The Hawthorne and the Beyma tweeters have horns that were molded for their respective AMTs. They look nice. Since I already have a pair of ESS AMTs and some spare horns to choose from, I will adapt some. The Selenium HM-4750-SLF looks like a good prospect.

 

 

 

 

 

[consistent]

Thanks Chris and others, some very interesting ideas coming through, thanks guys. I remember seeing a recent add for an open baffle speaker using the AMT centred around two open baffle 15” Eminence drivers. It was Pure Audio Project Trio...wonder how that would sound?

[Rudy81]

3 hours ago, Khornukopia said:

The Hawthorne and the Beyma tweeters have horns that were molded for their respective AMTs. They look nice. Since I already have a pair of ESS AMTs and some spare horns to choose from, I will adapt some. The Selenium HM-4750-SLF looks like a good prospect.

 

 

 

 

 

 

Time to start cuttin'. I won't be able to follow you since I will be stacking two drivers per speakers, to will need to go the DIY route. Good luck. Do you have the ability to take measurements after you mount the horn?

[Rudy81]

In this video, Mr. Hawthorne indicates he crosses the AMT at 350Hz!!!????

 

[Chris A]

It's conceivable using the horn shown, but I don't know what kind of AMT he's using--it's big, however, very big (width and length).  It might be able to reach a bit lower crossover point than an AMT-1 due planar-looking AMT that I see.  It would be nice to know who produced it and what the model name is.

 

That's a very splashy (sound like Nils Lofgren) recording that was used for the video.   I think that any listening judgments that were gleaned from that audition were purely incidental. I instead recommend a large string orchestra recording of high quality (i.e., untouched from a mastering standpoint).  Then you can hear the naturalness and presence of the massed strings to compare with a live performance sound with that AMT/horn combination.

 

Chris

[Chris A]

I found some blurbs on the Wayback Machine of the AMT used in this loudspeaker:

 

 

Quote

Introducing the Hawthorne Rainier Open Baffle Reference Speaker [$15000 per pair]

The Rainier is the statement for the Hawthorne design philosophy, but with several unique elements that make it a true reference far beyond the reaches of conventional open baffle designs.

 

Air Motion Transformer

...The Silver Iris Reference AMT is different because rather than having a usable range of 1.8khz-20khz as with conventional AMTs, this patented technology allows the Hawthorne Silver Iris Reference AMT to extend 500hz-40khz. I am waiting on measurements for the AMT in its waveguide, but here is an open-air measurement of the AMT without waveguide used above 1khz (not showing the full extension, this will be provided soon). The AMT has usable output much lower with a Fs of 235hz, which is an incredibly wide-bandwidth AMT!

This clarification is critical because the rainiers are substantially higher performance than this waterfall depicts. If you look at the subtle blips above 3khz those are completely eliminated in the wave guide and the decay is even faster. Think of it this way, the driver extends well below 500hz with less than 1ms Decay time, and ultra low THD. Anyway for comparison purposes I've posted the "bad" version of the Rainier graph compared to the Magico Q5. Try not to laugh at how much better the Rainier is, kryptonite included...

 

...Notice the range between 1khz and 10khz has a decay of 2.5msec whereas the Rainier is 1.3msec, twice as fast. Add the wave guide and you cut the decay in half once again.

 

Here's the cool part:

Imagine that speed from lower midrange up through extended treble. The Hawthorne AMT exhibits those same fast characteristics from 500hz on up with no crossover points. That's about as perfect a speaker as you can get. 

Active Crossover Package

Core Audio Technology offers a fully digital active crossover package for this speaker. This eliminates all A/D and D/A conversion from the signal path along with all passive components in the crossover. The system utilizes a computer for crossover processing, incorporating steep linear phase crossover slopes to eliminate phase distortion between drivers. All drivers are time aligned and then the room is corrected using the multi-phase correction algorithms in Dirac Live. The signal is then sent to a multichannel digital interface via Ethernet, and finally to fully digital amplifiers for amplification per driver. This is an active 3-way design and the package includes custom music server, digital interface, 4 channels of fully digital amplification, 2 channels of DAC/Amp 300wpc amplification for the bass, necessary wiring harnesses, and, of course, the Rainiers. It is a simple system, easy to setup and use, and will bring the Rainiers to new levels of performance well beyond the $200,000 price range.

 

Chris