ETF Subwoofer Demo

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Ex-Moderator
Joined 2011
Anyone heard the OB subwoofer demo at the ETF?

etf-chamber.jpg


etf-system-1.jpg


from: race for the bottom – lab jc
 
the issue here for me is how to use vintage drivers and make deep deep bass... the people i work with want to know more of how to do that. to hold their own and humiliate the modern age at their own game. it might sound weird to you, but not at all to them. the eminence drivers are very alike altec or lansing drivers from the late 40s and early fifties... light cones, powerful motors, underhung voicecoils with lots of flux in the gap. the very opposite of how people do the first few octaves today.

the interesting thing about the Heil AMT is that it really is a velocity transducer, and NOT directly a pressure transducer. it has more in common with the Manger driver or a true ribbon tweeter... than it does a piston driver. the heart of the matter is the velocity of air near the cone in the chamber, vs. the velocity at the mouth of the chamber. it is very similar, in a way, to what happens in a horn. low velocity and high pressure at the cone, high velocity and low pressure at the mouth. this is why it is called an air motion *transformer*. unlike the linkwitz dipoles or pass' hybrid "slobs", the excursion in the Heil AMT is reduced. yes, that is correct. as anyone at etf will tell you, infra sonic air pumping barely moved the cones... but WOW! what came out of the slots.

nothing comes free, and the cost of the Heil idea for bass is in the number of drivers. it took 4 pair of drivers to do what can be done with 2 each 15" volt radial woofers (the heavy cone long throw type). but the efficiency and distortion is MUCH improved with the Heil arrangement. the Heil thing ain't free! but, we had a 300 watt amplifier for a room roughly 30 feet by 90 feet... i doubt we needed more than 50 watts to play the beastie boys painfully loud and clean. i doubt the peaks hit 80 watts at the loudest? 8 each 3015LFs in series parallel is a lot of driver.

there is NO advantage to using an overhung long throw woofer in an AMT! in fact, that's basically idiotic. you won't use the drivers capabilities in any sort of clever way.

better to go for high BL in the gap, underhung and 16 ohms if possible. the kind of driver meant to be used in a horn is the one you want. vintage drivers! 8 altec 515Bs in a baffle like mine would be terrifying. 2 tesla in the gap and an X max of 3 mm. high Vas. low Fs. the kind of driver that normally barely makes 50 Hz in a reflex box the size of a van. remember, this has NOTHING to do with a piston driver. this is velocity transformer.
 
The eminence drivers are very alike altec or lansing drivers from the late 40s and early fifties... light cones, powerful motors, underhung voicecoils with lots of flux in the gap. the very opposite of how people do the first few octaves today.

The Kappalites have underhung motors?

The BL doesn't seem to be anything special. The 3015LF has a BL of 17 Tm. In comparison, the PA 380 has a BL of almost 26 Tm.
 
i made a mistake... the 3015LFs are in fact overhung. i have no comment other than to confirm i messed up. no more comment at all. not here to debate or defend. in fact, you all will be doing me a favor by simply ignoring or dismissing my post as nonsense. it was my mistake to respond and now i will pay for it...

thank you.
 
Last edited:
We all make mistakes.
The fact is that the ETF subwoofer sounded great, even with overhung woofers.
You have a point to stress the advantage of using underhung woofers in an AMT application because that is where they are expected to perform to their best because of the limited cone excursion.
Interesting remains to discuss how much "worse" the overhung woofers would perform; I guess that these also work in a pretty linear way when not challenged to approach Xmax.
 
I would like to bet that this was impressive in terms of bass. Multiple 15" drivers and a HUGE baffle help to make the pathlength very long, and reduce/minimize the OB short circuit losses that happen at low frequency.

The only issue I see here is a duct resonance from the slot loading:
etf-chamber.jpg


The way the drivers are mounted, opposed, to a slot or chamber that then opens to the "free air" causes a short transmission line to form. This line has a resonance that happens at the frequency corresponding to a length that is four times the "depth" of the slot. Let's say the slot is 20 inches deep (so that I can do the calculation in my head). 4 times 20 = 80 inches, which is about 2 meters. f=c/L and c=350 (rounding up here folks) meters/second, so 2/350=175 per second, or 175 Hz.

Fine, we have some resonance. So what? Well because the aspect ratio of the "slot" is high (it's much deeper than it is wide) the Q of this resonance will be high. You can either (A) avoid this region by crossing over an octave lower, e.g. around 100Hz, or (B) you can try to EQ the resonance to flat. With approach (B) it can be tricky to get the EQ just right. Approach (A) limits the bandwidth, which otherwise with a planar OB could reach as high as the driver can operate. (A) also leave the resonance untreated, and unless the LP filter is steep it will not be sufficiently suppressed. Even acoustic energy coming from a completely different source (the midbass horn) can stimulate the slot's resonance if the Q is high enough, since damping is zero.

Is there another approach? YES: use a wider slot. That will reduce the Q of the resonance and make it easier to work with. While the narrow slot does provide some loading of the driver (reducing Fs and increasing Qts) the resonance is not desirable if its Q is high.

I can suggest making the slot at least twice as wide as it is deep, the wider the better. You will still get the force cancellation, which is one of the benefits of this type of driver arrangement. With a Q more like 1.3-1.4, you can easily incorporate that into a crossover having a corner frequency equal to the resonance frequency by adding a low Q (e.g. Q=0.5) 2nd order low pass filter to get an overall response that is approximately a 4th order Butterworth. This could even be done using passive components only, since a second order Q=0.5 filter is the same as two first order filters of the same corner frequency. Just use two RC filters in series on the amp input, with values spread by 10x to avoid interaction.
 
Last edited:
frugal-phile™
Joined 2001
Paid Member
Fine, we have some resonance. So what? Well because the aspect ratio of the "slot" is high (it's much deeper than it is wide) the Q of this resonance will be high. You can either (A) avoid this region by crossing over an octave lower, e.g. around 100Hz

Charlie,

I followed the discussion of this speaker syatem on the JoeList and IIRC, this woofer array was crossed over at 40 Hz. I can’t recall how fast, but i think it was steeper than i would have guessed.

dave
 
Hi Charlie,
Would this be any better ?

It's a little better. The "slot" is wider, so the Q will be lower. I was thinking more along the lines of making the width at least 2x the depth. This looks like width=depth... it's probably OK. You should be able to see the resonance (around 150Hz I would guess) in a nearfield measurement.

These long slot loaded "bass line array" kind of constructions will also have a weak resonance in the up/down direction within the "slot" cavity. All of these resonances are OK if the Q of the resonance is kept low. The most effective way to do that at low frequencies is to make them much less deep than any other (lateral) dimension, or in other words the "slot" should be wide and relatively shallow. Of course the slot must be at least as deep as the driver frame is wide, so the only dimension you can vary is the slot width.

Also, regarding your image and the last post - if you are going to use a bunch of opposed drivers why not arrange them so that there is some degree of even order distortion cancellation? This does NOT do that:
pp-dipole.gif

Above: illustration of the wrong way to mount the drivers!

Instead, of mounting the two drivers pointing in opposite directions and magnets together, one of the drivers should be mounted backwards so that both drivers "point in the same direction". A pair of drivers operating in this type of system should push air into the slot at the same time. When mounting is done with one driver reversed, one of the driver cones is moving towards the magnet while the other is moving away from it. When the compliance vs excursion varies non-symmetrically (and other parameters, too) then the distortion generated by this will be mostly cancelled out when one driver is reversed. It's worth it when only a simple change to the driver mounting is all it takes. It would look like the attached image.
 

Attachments

  • distortion-cancelling driver mounting.png
    distortion-cancelling driver mounting.png
    18.3 KB · Views: 141
Last edited:
i don't want to debate anything about this design. and will simply leave the discussion if it becomes a ******* contest. there was nothing special about the eminence driver other than it is lightweight. x max is nothing special, and in this design it doesn't need to be. excursion requirements are reduced in the heil AMT... that makes the approach well suited to "vintage drivers" many of which have low fS, but very little excursion possible... and a high VAS. except in a long horn, it is really difficult to take advantage of the low fS and powerful motor in say, a Western electric TA-7181 or Altec 515A. x max is 3/16". but in a heil arrangement, that is more than enough. the cost is, of course, the need for many drivers, or to build a custom driver with many diaphragms and connecting linkages. but the reduction in excursion, the increase in efficiency at low frequencies, and lowering of fS are all positive gains. for pro audio and large scale high end, it's an interesting line of inquiry. i want to point out that the heil approach is the reciprocal of the open baffle bass speaker of sigfried linkwitz, and as modified by nelson pass... pass actually worked under heil at ESS in the 70s. his first audio job, i believe. the OB bass needs as much excursion as possible to function well. 2" x max is not too little. 1/8" is enough in a heil bass driver. all of mine barely move to make the low frequencies they do... and they are all designed to work in large club sized spaces.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.