Heil with a JBL woofer

[Robh3606]

Not sure which direction I am going to go either full up passive or an active with passive compensation on the Heil to help flatten the response. Have to be very careful with the impedance of the Heil driver combining with a passive network. Took a couple of measurements to get started of the raw driver, got an odd notch in response, repeated the measurement and it didn't go away. Have to try off axis but varying distance had no effect nor did temporarily using a baffle to keep the back-wave out of the measurement.. I did a quick 18Db network with a 5 K notch it's not bad at all. With the losses in the network it's close to the 91/92 sensitivity of either a 128H-1 or a 2214H.

Rob

[Mr. Widget]

That actually looks shockingly good. I wouldn't be too worried about the narrow notch at 6800Hz. I bet it would be hard to notice it.

I assume that is on axis. Please post more!


Widget

[toddalin]

I get the same/similar notch. This is my point where I said that you need an RTA for the best placement.

In my system, the depth and bandwidth of the notch is directly related to where the Heil sits in relation to the 2251J cabinet that it sits on. The closer the heil is to the front of the cabinet, the deeper and tighter the notch. I get the best integration with the Heil as rearward as possible. Perhaps this is the best time alignment with the 2251, but to my eye, it looks like it would sit behind the audio plane of the 2251 in this position.

I had attributed this to the interfacing/interference with the 2251J, but am now thinking that this has to do with a reflection, or lack thereof, from the cabinet that it is sitting on.

BTW, Widget is correct that it is difficult to actually hear the notch. OTOH, it makes it harder to hear tones centered at the notch (e.g., some percussion). These are told by listening to other speakers with other notches/peaks (e.g., the L200/300s)

[Earl K]

Very Cool!

It's great to finally see some pro measurements of these large Heils.

Woody's broadcast post was exceedingly attractive ( but alas, I need to sell off gear > not buy more ).

Still, it's all a bit haunting due to long cherished memories of listening to my aunts large AMT's .

I agree that some le14's should be a nice match-up.

[toddalin]

OK I did some testing and included some photos. My problem is an interaction with the midrange speaker. When I turn off the midrange, the tweeter spectrum is pretty much unchanged moving from the front to the back.

I am not getting the dip that Rob notes because I am using this in a 3-way and the tweeter is crossed over over to take advantage of this dip to produce 2nd order slope with a 1st order cap.

Tweeter forward:



Forward spectrum:


Tweeter rearward:



Rearward spectrum:


Spectrum for the tweeter without midrange:

[Robh3606]

So today I tried an Active approach with EQ. The first was trying to use the ESS Active and it just didn't work as the slopes were too shallow. I had to add poles which I suspected it just was not worth it using the ESS I would have been better off just going passive.

Round 2 Active was using one on my JBL M552 24dB L/R and this worked out to be the easiest solution as it was almost identical to the passive network solution with a modified lower Q notch filter.

Hello Widget

Yes that's on axis I will post step response and waterfall.

Hello Earl

I looked around for something like a CLIO measurement and didn't find one? Odd not like these are new kids on the block!

Hello Toddalin

Thanks for confirming the notch glad to know it's not unique too my driver. What is the resolution on the RTA are those 3db or10db notches?

[toddalin]

S

Hello Toddalin

Thanks for confirming the notch glad to know it's not unique too my driver. What is the resolution on the RTA are those 3db or10db notches?

It's 2.5 dB between the dotted lines. The full scale is set at 15 dB.

[Wallacefl]

I wonder how the Heil would respond in a sealed midrange style enclosure ? ESS used to make a model 1B Bookshelf that looks interesting.

[Earl K]

I wonder how the Heil would respond in a sealed midrange style enclosure ? ESS used to make a model 1B Bookshelf that looks interesting.

This is a recent response measurement ( done by JavadS. ) of the big Heil ( as a monopole & also as a regular dipole )



One can see that ( with very minimal response penalty ) one could encase the back of the AMT within a mid-range enclosure ( sealed sono-tube for example ).

[toddalin]

I've devised a "band aid" (pun intended) for the 5kHz suck-out.

Take an engineering scale and stand it vertically with one of the three points touching the Heil right at its centerpoint on the forward face. Voila!

Now to determine the optimum size and make a stand to make something like this work.

Edit: Ordered a length of triangular 1" clear acrylic. This is the same dimensions of an engineering scale, though the scales do have a very specialized shape. We'll see.

[Robh3606]

Finally got a change to build up the network. I updated the network added a .33 Ohm series resistor with the inductor and changed the value to .075Mh. The added resistance altered the Q of the notch and broadened it and shifted it down a bit. It measures well in an actual vs. predicted comparison.

Rob

[Mr. Widget]

That is remarkably close for the predicted vs. measured!

I don’t believe in listening to speakers via YouTube and I also don’t believe in listening to them via plots. That said, do they sound a bit bright? Have you had a chance to audition them?


Widget

[Wallacefl]

OK...I have not posted on the site very often but have been a JBL owner for years. Right now I have some 123a black cones from a L-88 and should like to try the ESS Heils as a mid-tweeter from 1200 hz. The old S70 Alpha has a crossover for the 123a that may be ok as as a start. As to your 18db for the tweeter...how was this derived as the values of components makes acquisition hard? I have no measuring equipment and generally copy others and/or use their guidance. Appreciate your input....

[Earl K]

OK...I have not posted on the site very often but have been a JBL owner for years. Right now I have some 123a black cones from a L-88 and should like to try the ESS Heils as a mid-tweeter from 1200 hz. The old S70 Alpha has a crossover for the 123a that may be ok as as a start. As to your 18db for the tweeter...how was this derived as the values of components makes acquisition hard? I have no measuring equipment and generally copy others and/or use their guidance. Appreciate your input....

What's needed ( from someone owning those woofers ) is a 1 meter, on-axis response measurement ( of the JBL 123 > black cone ) working in a large enclosure ( say 4cu' ) .

No custom crossover can be designed without employing a good response measurement ( & still be considered even half accurate ).

Those are the hard facts.

The HiPass can easily be 2-pole ( 12 db per octave ) with the choice of a higher crossover point ( say > 1500 hz ).

Values ( of the passive crossover parts ) can be massaged to what's available to purchase from the regular suppliers.

[Wallacefl]

Anyone ever copy/replicate the crossover Dennis Murphy for a custom pair of Salk speakers?
I believe it was at about 1600 hz.