The Neil Davis Heil AMT

[crzy8z]

Hello, I too am an Amt enthusiast in search of a stronger magnetic field
These are prototypes I built in the mid-1990's, patterned from the Audio Amateur how-to, and scaled up in size:









I still have the drivers except for the bass units, which had some design flaws. I loved the sound of the mids, but they are very inefficient (and heavy).

I hope it's ok to put the pics in directly, not sure of the forum rules.

[Keith Taylor]

Hi Greg, and welcome to the forum. That's a seriously big AMT you have there. The trick seems to be to develop enough force to get the pleats moving sufficiently. Compared to products in the market place you seem to have rather generous pleat widths, where most designs really cram the diaphragm into the the available space.

What pleat depth are you using on the mid AMT? You may find the thread called "Electrostatic AMT?" of interest. In there we concluded that the force that can be generated is really crucial to the sucess or failure of an AMT; or put another way, what F number (depth/width ratio) can be supported.

Keith

[Steve Dunlap]

Originally Posted by paulfx

Now I think I see what you are saying - round cross-section wires are better than angled metal pole pieces in front of the diaphragm, and they could go closer in to shift the diffraction upwards in frequency. Everything else at the front of the speaker would be flush like a dome, even including a felt ring, I mean rectangle.
So, if I understand the questions, would it be possible to flow flux perpendicular to the plane of the AMT diaphragm using power, taking care with the electronics to power-up up nicely, and having only a coil section coming around the front?
I wonder if there is an authority on electromagnetics at diyAudio who could stop in and elucidate the forces. Mr. Emag Expert, are you out there?

I just came across this thread. Is anyone still interested in electromagnets for speakers? I have done this, and may be able to answer some of your questions.

[crzy8z]

Originally Posted by Keith Taylor

Hi Greg, and welcome to the forum. That's a seriously big AMT you have there. The trick seems to be to develop enough force to get the pleats moving sufficiently. Compared to products in the market place you seem to have rather generous pleat widths, where most designs really cram the diaphragm into the the available space.

What pleat depth are you using on the mid AMT? You may find the thread called "Electrostatic AMT?" of interest. In there we concluded that the force that can be generated is really crucial to the sucess or failure of an AMT; or put another way, what F number (depth/width ratio) can be supported.

Keith

I would say the pleats are 3/8 inch depth (actual foil strip size) and about 3/32 inch width. The width varies due to the somewhat sloppy construction. Radiating area is about 2.25 x 10 inches. DC resistance 4-5 ohms.
The diaphragm sits in a 1/2 inch gap formed by two rows of round magnetron magnets sandwiched between strips of mild steel. The strips are 1/8 x 1/2 inch stock :



What took me totally by surprise here was that in spite of the large size, the low frequency range of these things is only 700 Hz, below which the response drops like a stone. Near-field response on center axis is a curiosity, with a straight but rapidly rising tilt to about 8 kHz, and a smooth gentle roll-off above that.
But if you go off center axis a few degrees horizontally or vertically, the response flattens out nicely, as good as most anything I’ve seen. With a 0.25 mh coil and 80 mfd cap, the minus 3 db corners are about 700 Hz and 10 kHz, and extremely flat in between, when measured just a few degrees off the vertical.

So the lack of low freq range precluded their effective use in that particular system, and the project went on the back burner. But their sound had been so good, especially considering the crude construction, that I always hoped I could someday construct a better set of them.

Any ideas on how to extend the lower range?

[Keith Taylor]

Greg, you appear to have a depth/width ratio of a modest 2:1. Trying to increase this by adding more pleats may not help the sensitivity, as with a 1/2 inch gap you may not have enough force to move the air mass when bounded by (say) 4:1 pleats.

The 700Hz LF cut off is a bit of a mystery. If it were dipole loss you would expect it to start higher up and be 12 dB/octave. I am wondering if you have a good seal between the frame and the magnets, or indeed, between the diaphragm and the frame. Something like beads of silicon sealer could be a quick fix but will make the frame hard to remove.

HF performance should be related to the pleat depth. Lambda (the wavelength) is 43mm at 8kHz, so Lambda/4 is 10.75mm which equates with your 3/8" pleat depth. One theory has it that comb filter like artefacts should appear above this frequency, but practice seems to suggest otherwise? The cavities formed by the 1/2" deep pole pieces in front of, and behind the diaphragm could have significant acoustic implications. A comprehensive treatise on pleated diaphragm loudspeakers is long overdue. Part of the reason that one has not appeared is that the knowledge is likely to reside in more than one discipline such as acoustics and fluid dynamics for starters.

Keith

[crzy8z]

Originally Posted by Keith Taylor

Greg, you appear to have a depth/width ratio of a modest 2:1. Trying to increase this by adding more pleats may not help the sensitivity, as with a 1/2 inch gap you may not have enough force to move the air mass when bounded by (say) 4:1 pleats.

The 700Hz LF cut off is a bit of a mystery. ..........................

I'm not sure of the calculation there, do you use the actual dimensions when the diaphragm is laid out flat, or the width when folded? Here I've propped open a pleat for a better look; the actual foil is 3/8 with 1/8 spacing between strips, but slightly less when folded:

Originally Posted by Keith Taylor

.........
One theory has it that comb filter like artefacts should appear above this frequency, but practice seems to suggest otherwise? The cavities formed by the 1/2" deep pole pieces in front of, and behind the diaphragm could have significant acoustic implications. ....................

I'm seeing a smooth HF rolloff. I did not detect any nastiness at the extreme top. If there are any diffractive effects, they seem to be helping. However, I have acquired some Earthworks QTC-1 mics recently, which should be more accurate than the Audix TR-40's I was using, and will retest and check for ripples in response.

Even though the LF range is not extended, the power handling seems to be much higher than the standard Heil dipole. A simple 6 db/octave network bandpass, and no sign of distress, no having to resort to trickery in the network, as with the AMT-1 (another topic for discussion )

The radiation pattern is different as well, leaning just a bit toward an electrostatic.

In any event, I would settle for having more efficiency at this point, and keep the present frequency response. How much more would the latest magnets offer over the ceramic/ferrite ones I have now?

[geraldfryjr]

very cool. i've always wanted to try and build a pair myself. jer

[srinath]

Very nice Neil and everyone.

I had a few questions. But since I have no clue about AMT's I am going to start at the bottom.
1. They are a tweeter I guess ? They compare to a emit ? I am thinking along those lines cos they too have a diaphragm.
2. The basic idea is magnetic noth and south poles staring at each other in a tight space with a electrically energized diaphragm in the middle. It vibrates and makes sound. How would it work with a thinner wall steel tubing. The reason I ask that, it will be a lot easier to work with a thinner walled steel piece than the thick wall one.
3. The steel tube, if you had left one wall a little longer instead of cutting it off @ 3, say it was 3.5 inches for that front side and 3 inches for the back side it would have made it easier to mount. No need for the extra angle piece on the sides. Would that work.

This is super cool, I am a self proclaimed master with working on steel and metal in general. The saws and other tools needed, I also have at my disposal as well as a lot of thin walled steel as well as thick walled. I also have powdercoating guys I regularly work with so after its all done I can get it to look good too.
I guess I better start cutting something huh.
Cool.
Srinath.

[srinath]

Originally Posted by paulfx

Before in this thread I mentioned one corner of the diaphragm came unstuck, this despite choosing the extra-sticky 3M tape with some care. I seems the warpage works on the adhesive over time and slowly pulls it apart.

1/4 inch dowell to the rescue - I carved a notch on the diaphragm side and ground the other until it fit with enough extra length to prop up the side ends, almost to the corners, gently but firmly against the sticky tape bezel. A half-drop of silicone adhesive on each end keeps it situated.

It probably came apart from vibration - - this driver went loud when I tested it. Glad I didn't seal it up right away. Adhesive just isn't enough for the ESS stock replacement. Also my first listening impressions and tests above are with a corner of the diaphragm flapping in the breeze.

Next time (if ever there is a next time), I'll tap in 4 holes in the front of the outer pole during the carving and grinding and hold it on with 4 nylon screws and clips. Maybe 4 notches and a long clamped thingo down each side holding the diaphragm snug against a sticky gasket.

Tape alone is not enough.

Hey man,
I am in charlotte NC, I have a very close biker friend in Cary.

Now I am really intrigued by the AMT (and almost any other exotic speaker style) there is.
That neodymium magnet pair being held north to south 2-3mm apart, not a problem. I can visualise a 100 ways to do it, simplest and cheapest of which is what is called a 2 inch channel and spacer plates. That stuff costs next to lothing at my local metal yard (living in nascar country has its advantages I would say) You can also get a larger channel as well as a smaller channel wiht square punched holes, so is plate, tubing and a 100 other steel structures.
Of course neodymiun magnets - I have bought n50's from kjmagnetics.com and in the flat bar style you can get n52's even.

OK here is where I have no clue what I am doing. OK so I can put these magnetic poles 2-3 mm apart and hold them without clanging together. Fine. I also can fit a diaphragm in the middle and make it sit in that spot too. However what I dont know about is, how will it sound. How big/wide will it have to be to produce what frequency. How much can the magnetic strength vary from location to location, because a bar magnet still has stronger corners, then come edges and then the center. And at the ned of it, how do I cool it, how do I put it in a cabinet and how do I point it for the sound to come out.

Then there is another question I have.
Steel will magnetise itself. How to prevent that. It may be better yto use aluminum, which itself cannot be magnetised, and while its more expensive it is even more versatile, more precise and more useable than steel and it will get magnetic permeation and may be at the same time temper the effects of the poles and edges.

What would be your thoughts on that ?

Cool.
Srinath.

[Steve M]

Hi,

I just built one of these DIY AMT tweeters, it sounds very nice! IMO better than most of the exotic dome tweeters I've tried and almost as good as a Raven R-2.

Just have a question about the pole piece, I made mine out of slotted 10mm steel threaded rod (see pics). Also, tried it without the pole piece and filled the gap with five small block ferrite magnets. Both ways sound good, with the pole piece in place it sounds smooth and sweet, with the the pole piece removed and block magnets in place its sounds more powerful and dynamic (maybe a little brash and forward?). My question is, theoretically do you even need a steel/iron pole piece?? why not just fill the gap with magnets??

Also, the amount of wadding/stuffing inside the tube makes a big difference, put too much in there and the whole tweeter becomes very tame/dull sounding (too much absorption).

Regards,

Steve M.