PDA

View Full Version : 1kHz suck out from 2206H in SR4722X cabinet, ideas?



sebackman
01-11-2014, 10:26 AM
Dear all,

I have put together small two way with a 2206H and a 2450SL/2332 in a SR4722X cabinet.



These basreflex Cabinets are original JBL and made for the 2206H.

When I measure I get a huge dip att 1Khz, -15db dip.

Any idea what that can be? Resonance in the cabinet?

Regards
//RoB

ivica
01-11-2014, 12:14 PM
Dear all,

I have put together small two way with a 2206H and a 2450SL/2332 in a SR4722X cabinet.



These basreflex Cabinets are original JBL and made for the 2206H.

When I measure I get a huge dip att 1Khz, -15db dip.

Any idea what that can be? Resonance in the cabinet?

Regards
//RoB

Try to "invert" cables connecting 2450SL driver

Regards
Ivica

sebackman
01-11-2014, 12:35 PM
Hi,

It's only the 2206H. No filters, no driver. MLS measuring Does not mater if the gate is long or short so it is not from the room.

Can't see how the cabinet could interfere so much. Both boxes behave the same.

Can it be reflektions from the recessed baffle or the sides? Should not JBL have seen that?

Thx
//RoB

more10
01-12-2014, 05:57 AM
Half a wavelength at 1000 Hz is only 17 cm. Could be reflections from the floor if cabinet is situated directly on the floor.

It could also be the horn working as a kind of helmholz. Try stuffing it with rags.

sebackman
01-12-2014, 11:29 AM
Hi,
Thank for good input.

The box is wired for active XO. Only the 2206 is connected. Horn and driver is not connected at all beyond the rear terminal. I have checked the measuring gear on other speakers and all works as it should.

After measuring again today it appears that it is reflections from the cabinet sides that stretch out on the sides of the 2206. Se picture above.

When measuring the dip is worst in right in front of the speaker. If I move the mike to either side, it reduces significantly. The measuring distance does not seem to matter much. The dip is apparent from 30cm and out to approx. 2m.

I use MLS and can gate down to 10ms or less which would take out any reflections from floor or ceiling. Dip is still there. The box sits on a speaker stand half way to the ceiling (280cm total height).

The only thing I can think about is that for whatever reason this cabinet must give strong reflections from the inside sides on each side of the driver and the bottom that interferes with the direct sound at just 1kHz. Weird.

I tried to put some wooden pieces on the sides of the 2206 to change the geometry and see if that changed the results. It did. It must be reflections and just bad luck that it is below the XO point 1,1Khz… Above would not matter…

Now the question is;

Do I do a new baffle on top of the old to move the 2206 forward and reduce the overhang from the sides/bottom? Drawback is that it adds weight..

Or do I make a 20mm thick MDF distance ring just below the 2206 to move it out from the existing baffle and hence reduce reflections?

Or do I mount 1" dampening foam on the areas where the reflections come from?
What do you think?

It is probably a badly designed JBL PA cabinet as they are original for the 2206H. JBL should have seen this when they made the box I guess. Original it runs with QASR/2417 and maybe the combination with those drivers and the dip in the box design worked. A bit hard to believe but I have no chance testing now since I rebuilt for 2332/2450SL.

I have not seen such narrow dip comming from refelctions before.

Best regards
//RoB

more10
01-12-2014, 12:13 PM
You are probably better at engineering cabinets than the original designers. ;)

Add some kind of lightweight material as wedges towards the edges.

frank23
01-12-2014, 12:26 PM
Interesting find that the sides interfere so much. Must be quite some worst case distances coming together here. Wedges would help I'd say. Or filling the space between the driver and the sides with a dense foam would also help I think.

Earl K
01-12-2014, 12:52 PM
Keep in mind that cabinets like this ( SR4722 ) are meant to be used ( & are measured ) with a metal ( protective ) screen in place ( that is sometimes foam-backed ) .

I agree ( the deep notch ) is most likely ( caused by reflections ) from the high edges of the cabinet .

I would remeasure , with ( 1/2" thick ) acoustic foam that covers the full baffle-board ( plus a metal screen that has 60 to 80% transparency ) . I'm pretty sure your notch will be quite minor after this is done .

:)

sebackman
01-12-2014, 01:13 PM
Hi,

Thank you.

I do have the original metal mesh fronts so I will try to do a measure with them in place. I don’t think that only will reduce the dip enough. There is no foam on the inside.

What kind of foam would you recommend? Would ordinary open cell foam of the same type used in furniture do it? Or are there special types with better function?

Best regards
//Robert

Earl K
01-12-2014, 02:08 PM
- Furniture foam is usually "closed-cell" type / so no / it's not usable .

- Here's the actual foam product ( that I've used for decades ) for my custom SR cabinets ( click the pic ) . It is definitely "open-cell" .

http://www.tchweb.com/tchresource/images/product-images/509-4805/509-4805905-l.gif (http://www.tchweb.com/tchstore/product/509-4805/c905/1-2--thick--Acoustic-Charcoal-Foam.html)

- An added benefit ( of laying in some foam behind a metal grill ) is that if the foam is "lightly compressed" by the metal screen , the metal screen will be damped a bit ( & help stop ringing at higher output levels ) .

- Also, it's a pretty good look .

- I imagine you can find acoustic grill-foam ( in Germany ) from someone like Thomann . (http://www.thomann.de/gb/index.html)

- Here ya go ;

http://a2.images4.thomann.de/pics/prod/182269.jpg (http://www.thomann.de/gb/adam_hall_frontschaum_019512.htm)

:)

more10
01-12-2014, 05:33 PM
divinycell (https://www.google.se/search?q=divinycell), bonocell (https://www.google.se/search?q=bonocell) or polyurethane (http://www.biltema.se/sv/Bat/Kemikalier/Ovrigt/Polyuretanskiva-252138/)

sebackman
01-13-2014, 10:26 AM
I just bought a couple of Auralex Studiofoam sheets. It's time to see what happens.

http://www.audioheritage.org/vbulletin/images/smilies/cryingsmiley.gif

61191

sebackman
01-13-2014, 01:02 PM
Ok, tested the foam sheets and it clearly showed that these are reflections. However I can't get rid of all of it.

I will try to make some MDF distance rings with the router to move th 2206 out of the box to change the geometry.

Best regrds
//RoB

more10
01-15-2014, 08:16 AM
Pictures with foam sheets? Foam will not direct all energy away from the edges.

badman
01-15-2014, 02:17 PM
Same problem here- and I've been working out figuring a similar solution. I keep wanting to just cut the lips off and be done but the cab is fiberglassed.... will be tough to do it well. If anyone has seen these before, I've got a thread in the mktplace, I'd be interested to get some feedback on how a 5525 evolved into a coated 2206h w/2426h and what seems to be a DDS eng90 waveguide. I'm missing the R and L values, which appear to have been intentionally obscured. http://www.audioheritage.org/vbulletin/showthread.php?34693-Sweet-Craigslist-Deal


61210

61209

sebackman
02-11-2015, 10:27 AM
Hi,

I just realised that I newer reported back on this thread. At the end of the day I ended up with making a mounting ring to move the 2206 out from the cabinets and to cover the sides in the box with foam. This made a real difference when listening on axis. Less so off axis.

Attached are some pictures to show the results.

Kind regards

//RoB

64541
64542
64543
64544
64545

sebackman
02-11-2015, 10:28 AM
And the finished speaker loks like this

64546

Horn Fanatic
02-11-2015, 12:35 PM
Half a wavelength at 1000 Hz is only 17 cm. Could be reflections from the floor if cabinet is situated directly on the floor.

It could also be the horn working as a kind of helmholz. Try stuffing it with rags.

It is impossible for a horn to act as a "kind of Helmholtz". There is no such thing as a "kind of Helmholtz". Either it is, or it isn't. If stuffing the horn with rags cures the problem, doing so would render the horn useless under normal operation. The suck out could be caused by a number of factors, beginning with the room dimensions. It could also be caused by phase cancellation due to the box dimensions, or inadequate bracing. Perhaps the culprit is the lousy crossover networks JBL uses, or a standing wave occurring inside the enclosure. Perhaps the horn is resonating due to the cheap material it is constructed from. Try damping the back of the horn with that gooey black material with the adhesive back. I think it can be found at Parts Express. Even if it doesn't solve the problem, it couldn't hurt. The frequency response for the 2206 is virtually flat from 100 Hz to 2KHz, which makes it an awesome driver to reproduce the vocal range. I'll suggest the suck out could be caused by a design flaw with the enclosure. I'll even go as far to suggest the tuning could be improved. Even a box which has no parallel surfaces, will resonate, not just from the material, but standing waves as well. That pearl was imparted to me by an acoustical engineer who worked in the acoustics division at Boeing.

H.F.

BTW - In my 40+ years of speaker manufacturing and design, I have never found one legitimate book, document, or article that advocated the use of foam inside an enclosure. I suspect those who began the use of it many years ago were DIY types who knew nothing of loud speaker enclosure design.

sebackman
02-11-2015, 01:28 PM
Hi Horn Fanatic,

-Totally agree with all of you suggestions.

The horns are really made of cast aluminium and not too shabby. These are also dampened as you suggest.
http://www.audioheritage.org/vbulletin/showthread.php?35243-Dampening-of-ringing-in-horns-baked-horns-)&highlight=baked

“Great minds think alike” :)

The cabinets are internally extra braced and the shape is trapezoid so I don’t think the internals would create the dip. At least not at the level I use to measure. There are no passive filters in the box as I use active XO.

I think it comes down a “design flaw” in the cabinet where the sides on each site of the 2206 cone produces a standing wave that cancels the direct sound at 1kHz- ish on axis. If I move off axis the dip reduces.

These are original JBL SR4722X cabinets made for the 2206 that I had lying around but they were made for PA and not HiFi. I guess that a flat front cabinet had not produced these curves. The ports are changed to reflect HiFi use.

So maybe “design flaw” is a bit harsh as the reassessed baffle is there to protect the drivers when transporting a PA speaker.

Anyway, by introducing the spacers and putting foam on the cabinet sides the dip was reduced significantly. The acoustic foam is applied on the outside of the cabinet, please see the pictures. Internal dampening is normal material.

The 2450SL/2332 measures just like text book. No problems there what so ever.

Well, now they are “good enough”. …Maybe a 077 can find its way in later, who knows.

Thank you
//RoB

subwoof
02-12-2015, 04:27 PM
That cabinet ( 4722 ) and all of it's variants just didn't work for any project no matter the woofer / horn / driver combo.

Impossible to get any decent bass and yes - the "suck" happens with any model 12 ( and I have many )...

The Vb is just too small and the cabinet ( as you discovered ) is too shallow.

I gave up and used the carpeted ones for the cat's perches in the office.

meow