Zilch's AK Design Collaborative - Econowave Speaker

Hi,
I have crossovers done, no new baffles yet. Check this RTA. I had a hole around 2 or 2.5 kHz, reversing polarity only moved it down to 1.5 or so, and looked worse, so I kept designed polarity and moved the waveguide. This is the best I got so far, about 2 inches out from front of current baffle. Given that the JBL 123A-1 has a shallow depth, does this seem reasonable? I estimate that the voice coil of the Selenium is now approx. 2 3/4" back from the baffle.

This is somewhat critical because if I am to keep this physical arrangement, I won't mount the waveguides in the new baffles. I'll have to keep them on top.

Ideas, suggestions???
:scratch2:

Wow, something looks seriously wrong. The photo showing the Selenium back of the baffle looks to be a bolt on unit, rather than a screw mounted. Is this the correct Selenium and waveguide?

Edit: I have gone back through the past few pages of posts, and see that you ordered parts for the low SPL version of woofers. Is it wired correctly? The RTA FR shows something terribly wrong. The waveguide shouldn't have to go so far ahead of the baffle to make the EconoWave work. Maybe Zilch has some ideas as well.
 
Hi Jack,
I have the "OMF" versions which are bolt on, as you see, and Selenium screw adapters. I did that so I could try them on my 811Bs easily. Do you think the adaptors are screwing up the HF?
Here's the part at PE:
http://www.partsexpress.com/pe/pshowdetl.cfm?Partnumber=264-323&FTR=264-323&CFID=28528646&CFTOKEN=18254967

I can't believe the bolt on driver with Seleniums adapter is causing the upset to the HF. The screw on adapter shown doesn't add significantly to the throat length, so it is probably not the problem. The FR of the D220Ti-OMF should be virtually identical to the screw on version. Your FR says differently. That is why I thought something is wired incorrectly in the crossover. Or perhaps something is wrong with the driver itself. Have you tried the other driver you have?
 
Just checked the wiring Jack, all AOK.

My other driver has a broken diaphragm. Got my replacement yesterday, which is out of round and unusable.

I'm thinking I should return both drivers and get the screw on versions. But if you think that's not the problem, I don't know what else it could be.

I just added my other .47uF (c3) across the existing one to increase the capacitance of the compensation circuit, it indeed gets more HF when doing that.

Hmmm. I'll see if anybody else weighs in, if not, these drivers are going back.

I can't believe the bolt on driver with Seleniums adapter is causing the upset to the HF. The screw on adapter shown doesn't add significantly to the throat length, so it is probably not the problem. The FR of the D220Ti-OMF should be virtually identical to the screw on version. Your FR says differently. That is why I thought something is wired incorrectly in the crossover. Or perhaps something is wrong with the driver itself. Have you tried the other driver you have?
 
Anyone working with the D210 screw on?
I might consider reverting to that. Sounds the same, right?
FYI: I looked at my broken diaphragm with a magnifier, and the lead broke where it connects to the tab. I handled it lightly, so, who knows.
 
I agree with Jack. It's just wrong what you're getting:

attachment.php


http://audiokarma.org/forums/showthread.php?p=1722475#post1722475

This crossover?

http://audiokarma.org/forums/attachment.php?attachmentid=83308&d=1205870935

The MLS certainly came out the same:

http://audiokarma.org/forums/showthread.php?p=1728659#post1728659
 
Just checked the wiring Jack, all AOK.

I just added my other .47uF (c3) across the existing one to increase the capacitance of the compensation circuit, it indeed gets more HF when doing that.

Your high pass starts around the right frequency, but seems to rise much too slowly. Perhaps the problem is with a bad part in the crossover. Why don't you try substituting the other crossovers parts for the ones in this crossover. I would suspect the .6 millihenry choke first, and then the 4.7 microfarad capacitor. The resistors and compensation cap wouldn't have the effect you are seeing. I had a shorted turn in a choke once, which caused some terrible looking readings, but it was an iron core type, not air core. But, who knows? Worth a try.
 
Raw 220Ti Response

OK,
Zilch suggested I get raw driver response, so I put the 220Ti & waveguide on my active system, no compensation.

System comprises:
-CX3400 x-over, set to 500Hz hi pass for test .
-DEQ2496 for pink noise and RTA. EQ and all processing set flat AND bypassed.
-Trends TA10.1 chip HF amp.

Raw response in red as measured by Zilch:
attachment.php

Selenium D220Ti in my active system, 1 meter, on axis, no compensation:
attachment.php

BMS 4550 for reference in my active system, 1 meter, on axis, no compensation:
attachment.php



So there's the hole at 2ksomething which caused me to move the waveguide forward.
The peak looks to be maybe 500Hz lower, compared to Zich's reference data.
I do show a dip near 3K.
Above that the curve is not similar enough, I don't think.
Am I right?

I'll wait for your comments and suggestions, but right now I plan to send both drivers back. Swapping more diaphragms is not appealing to me. I visually inspected the diaphragm you see RTA of, and the surround and dome are not concentric. I don't believe it's rubbing, but I don't have enough experience to be sure. (My replacement diaphragm was rubbing a lot -- out of round -- enough for me to tell easily.)

My feeling about this is that I do not want to climb an Everest of production run problems. I'd rather switch drivers. Is the 210Ti acceptable within the same crossover build? It seems to peak at a lower point. If it's good, I'd give that a whirl, and buy the thread on version to begin with.

If I'm missing something don't hesitate to let me know.
Thanks.
 

Attachments

  • D220TiRaw.jpg
    D220TiRaw.jpg
    65.1 KB · Views: 1,051
  • 4550Raw.jpg
    4550Raw.jpg
    60.6 KB · Views: 910
Hi guys,

I don't post here much, if at all. I am one of the diy-Audio'ers. Anyways, I'd see references to the PT series of w/g's before, but cheers to trying some of them out.

Anyway, I have a pair of BMS 4540ND's on hand. They are screw in CD's and I hadn't figured out what to do with them, but was thinking of using them in some speakers for my shop. Had planned on using some Dayton round waveguides, but then saw this thread. The JBL's are cheap enough to give them a shot.

But, any of you have suggestions for "econo-woofers/box"? I am likely to place them on the wall so I wouldn't need any baffle step to speak of and will get some lift from near boundary placement. I will likely use some not yet chosen subwoofer, so first and second octave isn't necessary, but inexpensive (<$100/ea) would be nice as well as ability to do a good job on the midrange.

I was thinking Selenium 12PW3 in a sealed box. Would also like to keep the box reasonable sized, to not take up too much room in the shop. I have no experience with most pro woofers out there, so some guidance would be nice.

I plan to run these off my magnavox PP 6v6 console pull amp (~11wpc).

I see lots of talk of vintage JBL/Altec woofers, but those aren't econo-woofers, at least at the prices they fetch on epay. It'd be nice if there was some chatter about woofers to mate that was also economy for folks starting from scratch.
 
OK,
Zilch suggested I get raw driver response, so I put the 220Ti & waveguide on my active system, no compensation.

System comprises:
-CX3400 x-over, set to 500Hz hi pass for test .
-DEQ2496 for pink noise and RTA. EQ and all processing set flat AND bypassed.
-Trends TA10.1 chip HF amp.

Selenium D220Ti in my active system, 1 meter, on axis, no compensation:
attachment.php




So there's the hole at 2ksomething which caused me to move the waveguide forward.
The peak looks to be maybe 500Hz lower, compared to Zich's reference data.
I do show a dip near 3K.
Above that the curve is not similar enough, I don't think.
Am I right?

I'll wait for your comments and suggestions, but right now I plan to send both drivers back. Swapping more diaphragms is not appealing to me. I visually inspected the diaphragm you see RTA of, and the surround and dome are not concentric. I don't believe it's rubbing, but I don't have enough experience to be sure. (My replacement diaphragm was rubbing a lot -- out of round -- enough for me to tell easily.)

My feeling about this is that I do not want to climb an Everest of production run problems. I'd rather switch drivers. Is the 210Ti acceptable within the same crossover build? It seems to peak at a lower point. If it's good, I'd give that a whirl, and buy the thread on version to begin with.

If I'm missing something don't hesitate to let me know.
Thanks.

I don't really think that there is a production problem with the Selenium driver. Compare your FR of the D220Ti with mine in post number 33. It is the bottom FR of three. It shows essentially the response of the D220Ti turned up much higher than the woofer, but crossed at 1,200 Hz. Looks pretty much like your uncompensated response, doesn't it?

The FR you show for the D220Ti seems pretty much normal to me. I still think that there is a problem in your crossover that distorts the FR as shown in your post of yesterday, post number 260. I cannot guess what the problem could be, but I feel sure it is not the Selenium driver.

The big hump in the response of the Selenium at around 1.6 KHz was the reason the high pass was moved up to 3 KHz. That tamed the bump, and smoothed the response. Your crossover should yield the same response as Zilch shows in his post number 266, and that I show in the first FR in post number 66..
 
Zilch,

Following GM calculations, Gedlee states:

I'd also like to point out that drivers are not flat and the directivity tends to be wider for a cone that a flat piston. My "rule of thumb" is that a 12" driver is about 90° at 1 kHz.

go to the thread as stated in your post for details.

Furthermore in JBL applications notes(ACC2212-95) for the PT-F95HF, it uses 12" woofer as well.
 
Zilch,

Following GM calculations, Gedlee states:

I'd also like to point out that drivers are not flat and the directivity tends to be wider for a cone that a flat piston. My "rule of thumb" is that a 12" driver is about 90° at 1 kHz.
Indeed, at 1 kHz, not 2 kHz. As I read it, Geddes is disagreeing with GM, saying a 12" is too large for 90° at 2 kHz, i.e., same as what I am saying.


Furthermore in JBL applications notes(ACC2212-95) for the PT-F95HF, it uses 12" woofer as well.
At 1.7 kHz, 3rd order Butterworth:

http://www.jblproservice.com/pdf/AE Series/AC2212,95-WH.pdf

And that's not a 4540 equivalent HF driver, either. When JBL DOES use an equivalent, the 2407H, they push the frequency down to 1.3 kHz, 4th order L/R:

http://www.jblproservice.com/pdf/AE Series/AM4212,95-WH.pdf

I built that crossover and wrote a thread on this, running 4540/2407H below 2 kHz, over in LHF.... :yes:
 
Zilch,

After reading all threads what you are saying is 2Khz and above x-over freq. the 10" woofer is more suitable and below 2Hz eg. 1.7Khz then 12"(what about 15" woofer?) is more to the mark?

Is this correct?
 
Is this correct?
Existing product provides guidance, but is not absolute. None of it is.

I'm saying there's a complex of factors involved, and the only way to know for certain is to measure the polar response of the woofer and HF driver/waveguide in their design alignments. 4430 used 1 kHz with the 15" 2235H at 100°, and JBL claimed they were matched in dispersion at that frequency.

I look at the chart in Dickason, run GM's calc, listen to Geddes's "Rule of thumb," and examine the polar responses of existing products using the waveguide, and knowing from experience that BMS 4540 on this waveguide is best crossed at the factory recommended frequency, or just a bit lower using a steeper slope, and suggest that a 10" driver may be a better match than a 12" one.

It is virtually a certainty that it's not going to be "seamless" with a 15" woofer at 2 kHz, and I don't know of any woofer that size which plays worth a whit up to that frequency, anyway.

4540's a tweeter, a super-tweeter, by some definitions, NOT a general-purpose compression driver....
 
Thanks, guys, for looking.

But I see something different which may account for the hole at 2kx I had before moving the horn. [Unfortunately I deleted that full range in-phase RTA pic with waveguide in standard location. When I am setup to run measurements again, if I still have the driver, I'll re-do that.]

Comparing my RTA (on right) to Jack's, I note the similarities, but see the difference (both uncompensated):
attachment.php

Jack shows no trough at approx 2k5/3k15 which mine shows. As Jack said, my filter is bringing in the HF at the right point, but not with enough amplitude, or much too slowly. To me, the FR suggest why -- this dip in response from the driver. Mine also shows the big hump a bit lower, I don't know if that makes much difference.

Now if you're saying the trough is not significant, well.....alright. I'll accept that. I want my Econowave to happen, and I've been spending all my spare moments in the past 24 hours staring at these curves, et. al., so I might be blowing this all out of proportion. Dunno. Hell, it's 3:30AM!:boring:
But the curves are different. In a crucial range.
:scratch2:
Tomorrow I will build a crossover using entirely new components. I don't have an inductance meter to check the coils, and my multimeter is probably not precise enough to check using resistance, so that's best.

I believe 210s are in my future.:D

Thanks again.
 

Attachments

  • side_by.jpg
    side_by.jpg
    101.2 KB · Views: 773
The inverse phase dip at 1.5 kHz is the acoustic crossover point.

What happens in correct phase, you haven't shown us yet.

I'm running the same woofers as you, but with a different lowpass.

Frankly, I haven't worked that part up, rather merely stuck in what I thought might work and have been enjoying them ever since.

[I'll PM you the details; they're not ready for prime time.... ;) ]
 
Zilch,

After reading all threads what you are saying is 2Khz and above x-over freq. the 10" woofer is more suitable and below 2Hz eg. 1.7Khz then 12"(what about 15" woofer?) is more to the mark?

Is this correct?

Generally, with a constant directivity horn or waveguide, you'll want to cross drivers at the frequency where the drivers' polar responses match.

Let's assume that the waveguide has 90 degrees coverage. The question then, is at what frequency a 15", 12" or 10" exhibit (more or less) the same coverage.

This ideally should be measured in any case, but as a good rule of thumb I'd say a crossover frequency of 900 Hz-1 kHz for a 15", 1.2 kHz for a 12" and 1.4 kHz for a 10". This should help maintain constant directivity as low down as is possible.

I would not cross any two-way higher than this. You'll find a lot of drivers meeting this criteria.
 
Generally, with a constant directivity horn or waveguide, you'll want to cross drivers at the frequency where the drivers' polar responses match.

Let's assume that the waveguide has 90 degrees coverage. The question then, is at what frequency a 15", 12" or 10" exhibit (more or less) the same coverage.

This ideally should be measured in any case, but as a good rule of thumb I'd say a crossover frequency of 900 Hz-1 kHz for a 15", 1.2 kHz for a 12" and 1.4 kHz for a 10". This should help maintain constant directivity as low down as is possible.

I would not cross any two-way higher than this. You'll find a lot of drivers meeting this criteria.

http://www.bmspro.info/index.php?show=item&usbid=10278&id=54358

Put it on the waveguide and dial the frequency through the range of interest on an active crossover while listening for coloration.

Here's one of several 4th-order crossovers I've built to make 2407H/4540nd play nice down that low. I only use them above 2 kHz now:

attachment.php


[NOT "Econo," obviously.... ;) ]
 

Attachments

  • _Temp.jpg
    _Temp.jpg
    42.6 KB · Views: 810
Back
Top Bottom