Double Stack ESS AMT-1 with Wings--Possible Kit for Heritage

[Chris A]

1 hour ago, Tarheel TJ said:

Chris, I am definitely interested in seeing what you can do with a Heil/MEH design.  My first thought when I saw this was "how can I turn this into an MEH arrangement?"  It seems like it would have incredible potential!

Yes, I agree.  The design of the AMT is different than a typical ribbon driver which suffers from the higher pressures that are found in the throat of a horn being modulated by off-axis (i.e., multiple entry port) woofers.  This is the reason why you don't see ribbon drivers being used in horn-loaded loudspeakers: they can't take the modulated pressures.  But the AMT pleated design is relatively immune to these bass frequency pressure modulations, which permits the horn-loaded AMT designs that you see in this thread, and also has been the reason why the AMT-1 itself is as efficient/sensitive as it is--it's already horn loaded down to ~2 kHz. I'm not sure that you could run at disco loudness levels with MEH-loaded AMT-1s without starting to affect the AMT diaphragm pleats but for home hi-fi duty these will work fine.

 

If using the AMT in a full-range MEH configuration (which is what I'd always try to do), the top and bottom sides of the horn will be necessary but these top and bottom sides can't be parallel to the horn's central axis if standing waves are to be avoided, so using a loudspeaker top panel to form the bottom side of the horn can't really be used if you enclose the top with a parallel horn wall to the top of the loudspeaker. 

 

Chris

[Thaddeus Smith]

On 1/12/2020 at 8:21 AM, Thaddeus Smith said:

 

Not yet - I got an automatic "your order confirmed" email just over a week ago, and no news since. Not going with the minidsp.. Going to be using a Yamaha sp2060 instead.

 

So their order notification system just sucks. My two heavy boxes of drivers arrived this morning.

[Pete H]

17 hours ago, Rudy81 said:

Chris A, if you will be there giving lessons, I'm attending!

So what you're saying is, the rest of us aren't worth hanging out with, but Chris is.    LMAO and we knew he would be a draw card, and if that's what gets you there, that will be great!

 

You will be bringing that new top end that you're lovin right?  Maybe the BB's too?  Waiting on mine to arrive, but will be coupling them up with JWC's mini punches with dual 12's for testing purposes.  

[Rudy81]

1 hour ago, Pete H said:

So what you're saying is, the rest of us aren't worth hanging out with, but Chris is.    LMAO and we knew he would be a draw card, and if that's what gets you there, that will be great!

 

You will be bringing that new top end that you're lovin right?  Maybe the BB's too?  Waiting on mine to arrive, but will be coupling them up with JWC's mini punches with dual 12's for testing purposes.  

 

Just a wee bit sensitive aren't we? LOL.

 

I can bring my new toys.  There is no way I can bring my bass bin. Took two of us to move each one upstairs.  Just barely cleared the width of the stairs....it was nuts moving them. 

 

I am about to run sweeps from 1khz down to 400hz with just the AMTs.  Hopefully, this will give you guys an idea of what you may be able to do.  Plan is to just run the AMT HF so as not to include anything from my LF.  For consistency, I will use L/R 24dB slopes.  I am curious to see just what they can do at the various crossover points.

[Rudy81]

23 hours ago, Chris A said:

If using the AMT in a full-range MEH configuration (which is what I'd always try to do), the top and bottom sides of the horn will be necessary but these top and bottom sides can't be parallel to the horn's central axis if standing waves are to be avoided, so using a loudspeaker top panel to form the bottom side of the horn can't really be used if you enclose the top with a parallel horn wall to the top of the loudspeaker. 

 

Chris

 

When I did my first prototype work, I started with a full, four sided, horn with my current sides and tops.  The SPL results showed some cancellation issues in that configuration.  I then took the bottom part off, followed by the top.  The only configuration that worked well were just the side wings.  I use the same angle on the top and bottom as the sides....that was no bueno.

 

A good horn design for this  setup will require someone with good technical know how and a way to manufacture said horn.  I think that  if we all keep working on this, we may come up with something truly worthwhile for a two way system.

[Rudy81]

For those who may be interested in what the different xover possibilities are, here are xover plots from 300Hz-1kHz on my waveguide setup.  Over time, I'm sure better designs can improve on this.  I have yet to decide where to cross my setup and fortunately can select a broad range of xover points.  These plots were the Stacked AMTs only with my bass bin muted, L/R 24dB slope crossover.

 

 

 

 

[Chris A]

One of the things that I found with my Khorn clones (Shinall KKSs) in the past couple of days is that the Khorn design apparently likes to be crossed at 250 Hz instead of 400 Hz from the bass bin to the midrange, as detailed in the following post that I wrote to another forum member in a PM yesterday:

 

Quote

Well, it's a longer story than usual on the lower crossover frequency [250 Hz used in the Shinall KKSs with DSP crossover].  Perhaps if I post an SPL response of the bass bin and midrange response that I measured in-room, you might see why I lowered the crossover frequency 250 Hz:

 

 

The orange trace is the bass bin SPL response (after EQ).  You might see the sharp drop outs in SPL at 277 Hz and at 332 Hz, while the midrange (blue trace) has one drop out at 826 Hz (which is probably not a real drop out).  By using shallow slope crossover filters at 250 Hz rather than 400 Hz, I get a desirable overlap of response at the LF crossover point that fills in the two drop outs on the bass bin. 

 

It turns out that the Atlas PD-5VH midrange driver (a.k.a., the Klipsch K-55 driver after it passes Klipsch receiving inspection and gets a Klipsch icon glued to it) does very well down to about 250 Hz so that's where I chose to cross it in order to compensate for those bass bin drop outs with smoother midrange SPL response.

 

Note that those drop outs are so narrow that humans probably can't hear them, but I chose to crossover lower in order to get closer to a much better phase response out of the bass bin at 250 Hz rather than 400 Hz so as to minimize a crossover interference band cancellation issues:

 

 

That trailing off of the bass bin's phase response at 250 Hz is a problem in the crossover region, where is causes a drop out in the spectrogram plot at the 332 Hz drop out point (which you can still see in the "peak energy time" curve dogleg to the right.  That's what I'm trying to minimize using the lower crossover frequency.  It sounds great, so I think I'll stay with it.  

 

 

When I get a chance, I'll try to clear the computer room a bit so that I can take more measurements off-axis to see what happens at the bass bin--midrange crossover point from a polar coverage viewpoint.  There's just so much stuff in the room right now that I can't move the microphone around in an arc motion to take those measurements unobstructed right now.  I store all my unused audio gear in that room that can't take the Texas summer heat (including the Shinalls and a pair of Cornwalls), there's a futon laid out as a double bed, and there's a desk and filing cabinet in there presently in that 13 1/2' by 11 1/2' room, so it's pretty tight right now.

 

So the bottom line here is that, while all other Klipsch Heritage loudspeakers than the Klipschorn could easily be fitted with stacked/winged AMT-1s and work quite well, I'm not currently sure how this will work with the Khorn itself due to the issues that I detailed above.  Note that:

1. My tests were on a clone of the Khorn, not on the Klipsch-made product, so some differences in the output of the horn bass bin around 300-400 Hz could be evident, making the use of the stacked-winged AMT-1s more doable.
2. The wing size of the stacked/winged AMT-1 assembly will to some degree control the lower frequency limit where the AMT-1 assembly can be operated.  If the wings are larger/longer than the ones used by Rudy, some of that 10 dB drop off from 500-->400 Hz might be avoided.  Since Rudy hasn't indicated that he wanted to cross as low as 400 Hz with his direct radiating bass bins, I understand why he doesn't feel any need to make his AMT-1 assembly wings longer. 
3. The third party horn-loaded AMT used in the Hawthorn Rainier dipole loudspeaker in the fourth post of this thread has been mentioned to be able to be used to below 400 Hz with DSP crossovers. While I believe that some "massaging" of the stacked/winged AMT-1 assembly configuration discussed here will get the user to that lower limit of frequency response, I do think it is prudent to mention that further development of the stacked/winged assembly is necessary for Khorn use (replacing the entire top of the the Khorn--midrange and tweeter). 
4. As I get the chance to assemble my own AMT-1 stacked/winged assembly for my own Khorn clone use, I will report back any developments (either positive or negative) that are encountered. 
5. All the Klipsch Heritage bass bins besides the Khorn are easily pushed to 600 Hz or even higher to 1 kHz without loss of on-axis SPL in order to cross with the stacked/winged AMT-1 assembly.  There is no risk for all other users of the current stacked/winged AMT-1s to replace their midranges and tweeters in stock Klipsch Heritage--with the caveat that this is only recommended if using DSP crossovers.  For those die-hard passive crossover users, the risk is on them to come up with viable passive crossover configurations to accomplish that. 
6. I do not recommend trying to use this thread to discuss passive crossover designs to use with stacked/winged AMT-1s. Another thread could be started for those discussions in order to keep this thread on track for DSP crossover users.  I personally have no use for passive crossovers with the AMT-1s since the AMT-1s do need EQ to get their response flat to 2-3 dB, making it a 20-minute job with a DSP crossover and REW to measure the response.  I've found that passive crossover discussions seem to expand to consume all available oxygen in the room, and I don't wish to be fielding those sort of questions in this thread (as the OP). 

[Rudy81]

Chris, I have not been working on the 400hz region simply to expedite getting something up and running that can be used on a daily basis.  Once I get this working for daily use, I do plan on exploring other, better options.  I would love to be able to cross at 400hz and stay out of the my bass bin resonance point. For now though, I am looking at a 700hz crossover to give the AMT the majority of the sound spectrum. 

 

I am more positive every day on what we may be able to achieve with the drivers.  Clearly, a type of horn is needed to support the lower frequencies, but it is very doable!

 

One I get the system set and tweaked at 700hz I'll post the results.

[Rudy81]

Quick and dirty EQ and 700Hz xover setup. Think I'll start here for some listening and see how it goes.  If this works, I'll live with it a bit and hopefully as time goes on we can get our heads together and work on really lowering the xover points.  Although, I think as it is you could cross at 500 hz, at least on a graph....not sure how it would sound.

 

If you guys think I'm missing something, please let me know. I will move on to sealing up the wing/AMT interface area....pics to follow.

 

 

[Pete H]

1 hour ago, Rudy81 said:

I can bring my new toys.  There is no way I can bring my bass bin.

There will more than a few BB's to use.  Super MWM's, Mini Punches and KPT415's, possibly some La Scalas, so should be choices. 

[Rudy81]

I had one more thing to try in my testing plan and that was to move the wings forward so that the AMT drivers would be flush with the front edge of the bass bin.  This leaves the wings floating in space, but places the AMT where it normally goes on the ESS speakers.  Since active crossing allows me to manipulate the delay, it is easy to move the wings forward. I surmised that with the wings back on the bass bin, the top of the bass bin truncates the sound moving downward from the drivers.

 

After initial setup and EQ, I moved the wings forward and reset the timing to match the 6" forward position. 

 

Listening alone makes this an obvious choice.  Sound opens up even more and I got a fuller soundstage....the type I originally heard when l first listened to individual AMTs on each speaker.

 

Here are the plot results for the wings forward placement.  I think this will be the preferred position for further evaluation.  @Thaddeus Smith, when are you coming over? I need another set of ears to critique the setup.

 

 

 

 

[No.4]

47 minutes ago, Rudy81 said:

I had one more thing to try in my testing plan and that was to move the wings forward so that the AMT drivers would be flush with the front edge of the bass bin.  This leaves the wings floating in space, but places the AMT where it normally goes on the ESS speakers.  Since active crossing allows me to manipulate the delay, it is easy to move the wings forward. I surmised that with the wings back on the bass bin, the top of the bass bin truncates the sound moving downward from the drivers.

 

After initial setup and EQ, I moved the wings forward and reset the timing to match the 6" forward position. 

 

Listening alone makes this an obvious choice.  Sound opens up even more and I got a fuller soundstage....the type I originally heard when l first listened to individual AMTs on each speaker.

 

I am glad you tried this. After reading your post about how you originally had a top and bottom with the wings but did not like it, moving them forward was the first thing that came to mind. I am glad you are finally listening to music!

[Rudy81]

14 minutes ago, No.4 said:

I am glad you tried this. After reading your post about how you originally had a top and bottom with the wings but did not like it, moving them forward was the first thing that came to mind. I am glad you are finally listening to music!

 

It wasn't that I didn't like it. It had worse performance in the area of interest between 800hz and 2khz. I never even listened to that setup, I was simply going by what REW noted. I will try and find the plot on that so you can see.  Edit: I am posting the plot comparing the options I considered.  The focus has been on the lower end of the response, and I found the wings only to be the better choice.  Also, the full horn is really tall with the top side......it is a large structure, and I know since my Oris are darn large as well.

 

Yes, I am all in now on listening.  Looking forward to your results on the extended wings!

[Rudy81]

This is how that experiment went.

 

 

[Thaddeus Smith]

1 hour ago, Rudy81 said:

@Thaddeus Smith, when are you coming over? I need another set of ears to critique the setup.

 

I'm tied up the next couple of weekends, but will try to get something on the calendar after that. Shoot me a pm with your availability schedule and we can go from there.

[Rudy81]

1 minute ago, Thaddeus Smith said:

 

I'm tied up the next couple of weekends, but will try to get something on the calendar after that. Shoot me a pm with your availability schedule and we can go from there.

 

My schedule changes each month, so just let me know when you are bored and feel like a drive. If I'm in town we can get together.

[Rudy81]

@Chris A, looking at the GD plot a few posts up, what do you think is causing the slight hump around 1.2kHz?  No name xover is set to 700hz. I think if I could flatten that 0.3ms bump, I could straighten out the vertical spectrogram.  Figured I'd get this as near perfect as possible.

[rplace]

3 hours ago, Rudy81 said:

Quick and dirty EQ and 700Hz xover setup

That is some good looking graphs for quick and dirty.

[Chris A]

That hump in response is where the first 1/4 wavelength of the AMT-1 is transitioning from the plastic case of the AMT-1 to the wings that you've built.  There is a change in the rate of phase lag (which is what you're looking at) that is due to the acoustic center moving more rapidly away from the listening position side of the AMT-1.  The small amount of group delay growth (looks like 0.3 ms or less) is not really audible, however.

 

But note that the AMT-1 is a symmetric device from front to back.  That implies that either a rear set of wings or some sort of foam plug on the rear side to support and adsorb the rear wave might load the diaphragm a bit more to keep the acoustic center a bit more "centered" as the first 1/4 wavelength progresses outside of the rear plastic case as the frequencies decrease.  This is the likely cause of that group delay growth and phase lag.  Your choice.  Since there are no electrical filters to increase the phase through the 1-2 kHz band, the likely culprit is the unloading of the backwave below 2 kHz.

 

Chris

[Rudy81]

47 minutes ago, Chris A said:

That hump in response is where the first 1/4 wavelength of the AMT-1 is transitioning from the plastic case of the AMT-1 to the wings that you've built.  There is a change in the rate of phase lag (which is what you're looking at) that is due to the acoustic center moving more rapidly away from the listening position side of the AMT-1.  The small amount of group delay growth (looks like 0.3 ms or less) is not really audible, however.

 

But note that the AMT-1 is a symmetric device from front to back.  That implies that either a rear set of wings or some sort of foam plug on the rear side to support and adsorb the rear wave might load the diaphragm a bit more to keep the acoustic center a bit more "centered" as the first 1/4 wavelength progresses outside of the rear plastic case as the frequencies decrease.  This is the likely cause of that group delay growth and phase lag.  Your choice.  Since there are no electrical filters to increase the phase through the 1-2 kHz band, the likely culprit is the unloading of the backwave below 2 kHz.

 

Chris

 

Thank you for the explanation.  I will have to add to my list of things to try out a set of 'mirror' back wings.  I think it will be very easy to add permanently if it works.