Zilch's AK Design Collaborative - Econowave Speaker

You are correct, in that higher mass will somewhat affect Qms.

BTW: Higher Qms means less loss in the suspension- which will, in a tiny way, increase efficiency. It's also affected by the spider stiffness- with that Vas reading, it appears that the spider isn't broken in enough yet- that Vas should increase somewhat, once that's done- and the Qms and Fs will also come down a bit at the same time. <snip>

Also BTW: The E120 and 2204 share the same top plate thickness, so they will have the same X-max with the same voice coil. It's just that the magnet is stronger on the E120.

Gordon, Thank you for sweeping some cobwebs out of my old brain with your clear explanations re: mass and stiffness.
So I guess the dominant factor in the Qes difference is that the E120 coil has more turns *in the gap* ?
 
Hi Katalyst,

Here's what WinISD says one will get with that custom woofer ( & it's custom TS parameters ) working within a 2.5 cu' box tuned to 43 hz .

This specific alignment ( woofer within tuned box ) is realized by continuing to use your boxes existing 2 ducts ( 2.25" I.D. x @ 3" long ) but in an effectively ( executed by you ) smaller box.

( Obviously ) you'll need to displace some of the existing box volume ( at least 1.3 cu' ) or build a new box .

You'll know you've arrived at this smaller effective box size by measuring the systems Fb ( made by running DATS with the woofer mounted in the smaller box ).
- Essentially Fb ( box tuning ) is represented by determining the frequency corresponding to bottom of the trough between the 2 impedance peaks.

Katalysts_Custom_2204H_tuning.PNG

You can see I used some of your published TS parameters ( from woofer #2, I think ).

It would be handy if you published all the parameters ( just post a full picture of your DATs file, using the snipping tool found in Windows > png files display more clearly than jpgs in most browsers, btw ).

:)
 
Last edited:
Katalyst,

It appears your woofer is just not feeling the effects of ( interacting with ) the box that its currently placed in.

IOW, it would appear that your woofer is acting as if it's mounted in an OB situation( contrary to what the commonly used Prediction Software says should happen, which predicts it's within some sort of IB ) .

You've already run into this strange situation ( & commented on this behavior ) when you ran the woofer in the box with the back off ( +/or sealed up the ports, +/or changed the port quantities ).
- You commented that the response didn't change ( even with the back of the box left off )

Again, this is behavior that is not predicted by the software that we all use ( so it's new territory for all or most of us ).

FYI, just about all box alignments are made by sticking a woofer in a box where the effective box volume is a fraction of the size of the woofers VAS ( which as a working convention, has been broken in your applied usage ) .

So, you're observation ( a few posts back ) where you suggested this woofer ( seems to want to be placed in a smaller box ) is well worth exploring.

Hopefully the woofer will start acting like the predictions suggest, when the box size is the same ( or smaller than ) it's own VAS figure.

:)
 
Last edited:
Gordon, Thank you for sweeping some cobwebs out of my old brain with your clear explanations re: mass and stiffness.
So I guess the dominant factor in the Qes difference is that the E120 coil has more turns *in the gap* ?
My understanding is that the E120 magnet slug is bigger/taller ( by @ 1/8" I think ) therefore increasing the flux lines available within the gap from 1.2 Tesla to 1.35 T .

This increased flux level working across the same amount of voice-coil winding's will result in a higher BL product ( effectively lowering the Qes figure ) .

:)
 
Last edited:
My understanding is that the E120 magnet slug is bigger/taller ( by @ 1/8" I think ) therefore increasing the flux lines available within the gap from 1.2 Tesla to 1.35 T .

This increased flux level working across the same amount of voice-coil winding's will result in a higher BL product ( effectively lowering the Qes figure ) .

:)

That is correct. And, with the higher Bl product, comes a higher mid-band sensitivity. We're just going to trade some of that off, by adding mass, to get a lower bass rolloff frequency.

I would think that once another 1 to 2 ounces of mass is added to the cone, it should start performing nicely in the original enclosure.

Regards,
Gordon.
 
That is correct. And, with the higher Bl product, comes a higher mid-band sensitivity. We're just going to trade some of that off, by adding mass, to get a lower bass rolloff frequency.

I would think that once another 1 to 2 ounces of mass is added to the cone, it should start performing nicely in the original enclosure.

Regards,
Gordon.

I wouldn't do anything right now that isn't completely reversible ( such as adding more weight to the cone ).

I would first see if the woofer will act as predicted by putting it on a tuned box that has an internal volume equal to or smaller than the woofers measured Vas .

I would also want to see that adding more cone weight is going to encourage this woofer to start reacting like it's working within a tuned enclosure rather than simply acting as if it's OB ( the present status quo ).

This current situation ( OB behavior ) was not predicted by anyone's software /// therefore I think one should be ultra careful in assuming a simple fix has a predictable outcome.

If adding more cone weight can get the Vas up to 4cu' then I think the typical box/woofer alignment rules should kick in for the current 3.8cu' box ( but that's just a guess )....
- Honestly > I believe getting to 4cu' wll require increasing the compliance ( not just adding weight ) and I wouldn't recommend going down that road blindly.

I also wouldn't add more than another 35 grams >> doing so takes the sensitivity down to about 91db ( so what's the point ? ).


:)
 
Last edited:
I believe this is when I "Fated" myself.

Hi,

Ha! In which post # were those fateful words uttered?

Anyways, the job as I see it now is to get your woofer to start interacting ( predictably ) to box volume and that means putting one of those woofers in an enclosure that's smaller than it's measured Vas.

Do you have anything in the 2 cu' size?

:)
 
This is where I take a break from this build. The Ewave section is a total success. The Peerless DFM-2535R00-08 and B52 buyout wave guide sound great. The cross over EarlK helped to create works fine. The woofers do not like the box. To others who read this thread. Use an existing woofer and box combo that plays well to 2k . Add the wave section and you should be well pleased. When the weather is nice later this year I will build boxes for the woofers. The DATS V2 will help in the box design. I will give these woofers what they want. A different box.
 
This is where I take a break from this build. The Ewave section is a total success. The Peerless DFM-2535R00-08 and B52 buyout wave guide sound great. The cross over EarlK helped to create works fine. The woofers do not like the box. To others who read this thread. Use an existing woofer and box combo that plays well to 2k . Add the wave section and you should be well pleased. When the weather is nice later this year I will build boxes for the woofers. The DATS V2 will help in the box design. I will give these woofers what they want. A different box.

Yep, I can certainly understand the need to take a break from all this frustration.

I'll edit my post shortly & display a couple of WinISD predictions ( for what was expected using the current box/woofer combo ) vs what actually happened ( ie; OB-like performance curves ) .

:)

EDIT:

Expected ( predicted );

3.8cu-ft_response_+custom_2204H_.PNG

Reality ( measured );

index.php


Katalyst_Reality_.PNG

 
Last edited:
This is still fun to me. Learning is enjoyable.The set backs have taught me various practices to use . Just because the simulation says it will work in a certain manner is not always true. This is the lesson to take from this.
 
An amazing number of folks will insist that reality is wrong when it doesn't match the model or simulation.
 
I am stubborn. Added Styrofoam bricks to make the box "smaller". Volume at 2.5 cuft. One port open 2.5 inch .75 length. Now we are getting some where.2204 120 2.5 cuft one port.PNG
 
index.php


Pretty sure this will drive you nuts << but >> BTC ( Box Tune Calculator ) says that to get the displayed 35hz box tuning the box must be sized in the neighborhood of 4.18 cu' ( when using a single 2.5" diameter port that's .75" deep )

That suggests those foam blocks ( aren't so solid ) and are actually adding virtual volume ( instead of subtracting it ).

I have to guess that the 2204 simply sat there and yawned ( simply offering no extra bass output for all your effort ) .

:)
 
So the low point between the impedance peaks shows the box port tuning? The fs is 53.16 . Qts 0.5645 . Qes 0.6407 .Qms 4.746. These are with the blocks in. Blocks removed are as follows . Fs 52.49 . Qts 0.5931 .Qes 0.6448 . Qms 7.388. The bricks are not porous. How would they add volume? First picture bricks in . Next bricks out. This is the link I am using to plot ports. http://www.mobileinformationlabs.com/HowTo-1Woofer-Box-CAL Port lenth 1.htm
 

Attachments

  • DSC00004.JPG
    DSC00004.JPG
    69.2 KB · Views: 13
  • DSC00006.JPG
    DSC00006.JPG
    85.6 KB · Views: 13
So the low point between the impedance peaks shows the box port tuning?

Run another DATS impedance sweep with no foam blocks and then post a similar pic as above ( put the cursor at the bottom of the trough > just like above ).

- Keep one port blocked and the other duct removed ( just like above ).

:)
 
This is the link I am using

Inputting your boxes port tuning specs into that online calculator ( then guess-timating the box size to agree with my phone app gives a ) 4.2 cu' box tuning of 35hz ( that's provided by a simple 2.5" diameter hole in a 3/4" deep baffle board ).

Box-Calculator.PNG

Really, I don't know anything about the effects of Styrofoam when its placed in speaker cabinets ( I've never done that experiment ) but impedance sweeps don't lie ( unless one ignores the calibrating procedure within the DATS software ).

:)
 
Last edited:
Yes dats calibrated. Here is the sweep after the blocks were removed. The low point between the peaks is the actual box port tuning?
 

Attachments

  • BlocksRemovedOnePort.PNG
    BlocksRemovedOnePort.PNG
    115.4 KB · Views: 21
Yes dats calibrated. Here is the sweep after the blocks were removed. The low point between the peaks is the actual box port tuning?

Yes the low point between those 2 impedance peaks is essentially the box tuning ( Fb ).

:)
 
Last edited:
Back
Top Bottom