recycling old woofers

[haavind]

I have a handful (4) of old 8" woofer elements from my old Dali AXS8000 and I am looking for something to do with them.
The model number is Vifa C20WJ-17-08. I haven't found any specs for that one but this one is probably pretty close.

I was thinking of getting a cheap DSP + class D amp board from aliexpress, something based on the ADAU1701 maybe, and make some active subs.

One plan is to make a pair of open baffle subs, a Linkwitz LX521 style with the 90 fold, and just use them to assist my mains. If I understand correctly, putting a dipole next to a sealed woofer will sum to a cardioid, and that should help mitigate the backwall cancellation dip I have between 70 and 100hz. Does that seems reasonable?

The AXS8000 was an unsophisticated party speaker, I'm now running a pair of Martin Logan Ascent i, so a step up in fidelity.
Will these elements be suitable for open baffle, and will they play hifi enough to not spoil my "nice" speakers?
..or should I just forget about it and make car-subs instead?

 

[Devnull]

Have you thought about measuring the TS parameters use REW?

 

[DVDdoug]

I haven't found any specs for that one but this one is probably pretty close.

It's probably best not to assume that...

You should avoid a ported speaker if the Thiele-Small parameters are unknown.

Things don't usually go "terribly wrong" with a random driver in a sealed box. And the bass range can be extended with EQ/DSP (it doesn't have to be digital).

Open baffles usually don't make good woofers/subs but some people seem to like them.

If I understand correctly, putting a dipole next to a sealed woofer will sum to a cardioid,

I don't know ANYTHING about that, but multiple subs at different locations can help to smooth-out room modes. Another option would be to build 4 subs and space them around the room if that works with the space & decor in the room. But personally... If I was going "that crazy" I'd want something bigger than 8-inch drivers.

 

[haavind]

Have you thought about measuring the TS parameters use REW?

Not really sure how. Also, not really sure how I would interpret the result. TS is still a little bit mysterious to me. What would a good spec be for this use case?
The speakers have been lying still for, idk, like a decade. Should they be broken in again before measuring?

The specs for C20WJ-19-08 and C20WJ-09-08 are available and pretty much identical, so I assumed mine are the same.

I don't know ANYTHING about that

The idea is that the negative DP backwave cancels some of the positive MP backwave, while the positive DP frontwave adds some to the positive MP frontwave.
A dipole again, is a sum of two monopoles with opposite polarity (the difference of two monopoles if you will).

Open baffles usually don't make good woofers/subs but some people seem to like them.

I could go sealed, (or TL or horn or other weird topology), but I am in general skeptical of using resonance to store and release energy and will sacrifice efficiency to avoid it.
M-frame dipole seems like a simple build, and one that could maybe solve some issues for me.
My mains "can" be moved somewhat, but for practical reasons are best where they are, and that means I have a noticeable trough in SPL between 70 and 100 hz. Pretty sure it is backwall cancellation. I could EQ it up a bit but that would put more sound-power out in other directions and I don't want that. Cardioid bass should fix both the dip and the phase lag. They don't have to go deep-deep, just assist and even out the mains. From between 50-70hz, up to 100-300hz would suffice. For the deep-deeps I'm currently considering Rythmiks.

My room is large, irregular, and leaks bass in all sorts of openings, nooks and crannies, so room modes are not a huuuge problem. My design plan is then to have all the direct sound emanate from one "stage area", and be timing-and-intensity-correct. After that the sounds can decay, or linger in the room as they naturally se fit.

this paper describes my case perfectly

 

[MAB]

Not really sure how. Also, not really sure how I would interpret the result. TS is still a little bit mysterious to me. What would a good spec be for this use case?
The speakers have been lying still for, idk, like a decade. Should they be broken in again before measuring?

The specs for C20WJ-19-08 and C20WJ-09-08 are available and pretty much identical, so I assumed mine are the same.




The idea is that the negative DP backwave cancels some of the positive MP backwave, while the positive DP frontwave adds some to the positive MP frontwave.
A dipole again, is a sum of two monopoles with opposite polarity (the difference of two monopoles if you will).



I could go sealed, (or TL or horn or other weird topology), but I am in general skeptical of using resonance to store and release energy and will sacrifice efficiency to avoid it.
M-frame dipole seems like a simple build, and one that could maybe solve some issues for me.
My mains "can" be moved somewhat, but for practical reasons are best where they are, and that means I have a noticeable trough in SPL between 70 and 100 hz. Pretty sure it is backwall cancellation. I could EQ it up a bit but that would put more sound-power out in other directions and I don't want that. Cardioid bass should fix both the dip and the phase lag. They don't have to go deep-deep, just assist and even out the mains. From between 50-70hz, up to 100-300hz would suffice. For the deep-deeps I'm currently considering Rythmiks.

My room is large, irregular, and leaks bass in all sorts of openings, nooks and crannies, so room modes are not a huuuge problem. My design plan is then to have all the direct sound emanate from one "stage area", and be timing-and-intensity-correct. After that the sounds can decay, or linger in the room as they naturally se fit.

this paper describes my case perfectly

Before even the most basic cabinet, you will need to have the specs for the woofer. And preferably be able to measure them. A DATS will collect the speaker's parameters if using REW and a soundcard is too involved.
I can't find the specs for the C20WJ-17-08 either. But I can illustrate what would happen if you assumed that it was identical to the C20WJ-19-08.

First, I found the actual Vifa spec sheet for C20WJ-19-08:

Below is the Frequency Response in a 125 liter ported cabinet for the woofer assuming the above specs are correct (black trace). I also have the response if the woofer is about 30% lower Q than the spec sheet (red trace), and 30% higher Q (orange trace).

There are plenty of 8" woofer from Vifa and other manufactures with larger difference in Q, for a range of applications that need high or low Q woofers. The resulting frequency response is most likely going to be whacko unless you actually know the woofers parameters, and can use basic box-modeling software. Both are trivial, and will allow you build a good sounding speaker, rather than a random attempt.

The DATS will also allow you to tell if the woofers are blown or damaged or have a voice coil rubbing. And measure the results of your build, and help tune the port.

I see you have some exotic ideas you want to implement. Best to start understanding the basics.

 

[AnalogSteph]

I also have the response if the woofer is about 30% lower Q than the spec sheet (red trace), and 30% higher Q (orange trace).

Are you sure those aren't actually backwards?

 

[haavind]

Finally got my wiring harness and buggy mic soundcard to work

Here's my impedance curves, yellow is free air, purple is 7g added mass
seems generally reasonable to my very untrained eye. The hump at ~250hz seems a bit odd.


And here are my calculated TS parameters:

 

[DavidEdwinAston]

Sorry, I got completely the wrong idea from the thread title, and was going to suggest taking them to the Battersea Dogs home.

 

[Wolf11Man]

Sorry, I got completely the wrong idea from the thread title, and was going to suggest taking them to the Battersea Dogs home.

Maybe not the worst idea what to do with these...

The measured data seem a little odd. I would carefully press them in and out a few times to check limits and loosen up the spider. Then generate a ca. 50 Hz frequency and let them bark for an hour or two, with significant motion, like +- 4mm. You should need only very few watt to have them move quite far, as you stimulate the resonance frequency.

Then measure again, in a silent room. Your curve looks like made in a very noisy surrounding and the speaker picks up the noise.
Fs may be much lower and Qts higher. Maybe post a good picture of the woofer, to show the magnet size and details.
These TSP don't look like something you would have ordered as a "custom" woofer from VIFA.

 

[AnalogSteph]

And don't forget ambient temperature as a variable. The suspension typically gets stiffer when cold, particularly the surround. That would drive fs up and increase sensitivity.

Right now these are looking more like woofers rather than subwoofers (although drivers like this can still be useful if EQ'd electronically, even if you'd prefer 10" or 12" ones). Although impedance is getting a bit high for my taste by about 1.7-1.8 kHz already, clearly no shorting rings. These would be nice for use in 8" monitors with a 1" tweeter in a big ol' waveguide (or even a compression driver) for a sub-2 kHz XO, think Behringer 2031 or Heissmann DXT-MON or the like.

Had a look at the AXS8000 these came out of - twin stacked 8" woofers in a 2-way with a 1" in the smallest of waveguides on top, what a daft construction. (Not the most inert enclosure either, apparently? I guess they were cheap and made quite a racket.) Might have worked OK as a 2.5-way, but ideally should have been a 3-way with an additional midrange crossed over at 600 Hz tops.

 

[haavind]

Sorry, I got completely the wrong idea from the thread title, and was going to suggest taking them to the Battersea Dogs home.

I would never recycle my old wooffer

Even if he barely woofs anymore

The measured data seem a little odd

My measurement rig is not superhigh fidelity. some 50hz mains noise, and a relatively loud pc-fan. I did longer and multiple frequency sweeps to get down to this noise level.
I've wired up a break-in routine for them and will try to isolate better before re-measuring.

I guess they were cheap and made quite a racket

I demo'ed some smaller, similar priced dali 3-ways that played nicer, but the feeling of getting punched in the chest from inside was more profound. I was a young teenager at the time and the heart wants what the heart wants. I still like big bass, I cannot lie.

The AXS8000 was rated to 32hz so i figured electronic eq could make an acceptable sub. They are playing spotify now on +24db bass-eq and they aren't awful full-ranges either

 

[MAB]

Are you sure those aren't actually backwards?

Yes, am fairly certain! Since it's a couple harmonic oscillator, my intuition usually fails me!!!

I didn't change the cabinet, just aligned it to the CWJ20-19-08 specs (VituxCAD already has the parameters in the database! )

I then hacked a low Q version of the same woofer:

Since the actual C20WJ-19-08 is a slightly high-Q woofer for a ported enclosure, the large box' alignment is somewhat fragile. Dropping in a low Q woofer will cause the port starting to hoot at higher and higher frequencies, some low-Q woofers more than others. You could solve this by changing the cabinet and port for sure. But would of course need measurements, which was the point!

You can play around with random lower-Q woofers, most show similar behavior to my hack.

 

[MAB]

View attachment 425448

Finally got my wiring harness and buggy mic soundcard to work

Here's my impedance curves, yellow is free air, purple is 7g added mass
seems generally reasonable to my very untrained eye. The hump at ~250hz seems a bit odd.


And here are my calculated TS parameters:

View attachment 425449

Nice, great start!
These won't be good subs.

Also, you may have used too high Sd (effective cone area). Sd should be around 231 cm^2 based on the similar Vifa units (324 cm^2 is more like a typical 10" driver's Sd).

The hump at 250 Hz is likely a resonance. The fact that it goes away with added mass may be due to the mass damping out the resonance. Here is an Audax woofer with resonances:

Very similar to the data in Audax' spec sheet:

My unit has renounces in the impedance trace around 300 and 500Hz, Audax' spec sheet shows them at ~250 and 420Hz, very close to my sample. The Audax spec sheet shows several dB dips in the frequency response at these same frequencies (I added red arrows). It's likely that the 300 and 520 Hz impedance spikes in my sample are 5th and 3rd order HD of a resonance mode at 1500Hz. Possibly surround. I might measure the frequency response and distortion for the sake of speaker trivia.

 

[haavind]

These won't be good subs.

I see. What metric are you basing that on?

Sd should be around 231 cm^2 based on the similar Vifa units

Yeah you are right. I used the built in calculator in the TS module in REW and punched in 8". 231cm² means effective diameter of less than 6.8", which is, honestly, surprising to me. I went and measured, and indeed they are just barely 7". I'm not sure what the correct measure point is. And my majestic 10"s in my mains are just measly 9"s .

 

[MAB]

I see. What metric are you basing that on?

One limiter is the 4mm xmax. If used as a sub, you likely want a steep high-pass filter below 30Hz to keep it from bottoming out.
If your measurement is correct, the resonance frequency is pretty high, which will limit it a bit more than what the spec sheet suggests.
You might want to measure the other drivers to see if they are all matched.

Yeah you are right. I used the built in calculator in the TS module in REW and punched in 8". 231cm² means effective diameter of less than 6.8", which is, honestly, surprising to me. I went and measured, and indeed they are just barely 7". I'm not sure what the correct measure point is. And my majestic 10"s in my mains are just measly 9"s .

Yeah, it typical, and understandably confusing. Most 8" drivers have Sd around 220 cm^2, here are a bunch of 8" drivers, randomly selected from a few manufactures:

Sd (cm^2)	Manufacturer of random 8" woofer
211​	Dayton
206​	Goldwood
232​	JBL
213​	Wavecor
227​	LaVoce
222​	Ciare
219​	Morel
220​	Tangband

 

[haavind]

pre-breakin (darker colors) vs post-breakin (brighter colors) impedance plot. The general trend is that the resonance peak shifted up (yes, up!) in frequency by 5-10%, and down in magnitude by around 5-9%. I tried to isolate some of the fan-noise but its still noticable, and there is definitively power-noise at 50hz, but overall it seems about right.

Seems like pretty big difference between them?

TS parameters for driver1 (the red one):

 

[Wolf11Man]

I have no idea what one could make wrong with a simple resonance frequency test, so I think this must be what it is.
Anyway, for a sub these seem to be not the best candidates.

If you had wood panels for free and don't mind the time spend, one could build a large, closed square box. May look like a coffee tabl. Now put in a diagonal cross and mount one of these barkers at each side in it's own compartment.
Then get a cheap, strong DSP amp (the BRU5 from Aliexpress is a great value at 40$) and drive two of them in parallel with each channel of it. Then DSP to taste.
You will not have a wall breaking sub, but with some room gain, maybe placed near a corner, it should work quite well.
4x 8" even with only +-5mm excursion, are quite some cone area. As these seem to have a 35mm voice coil, they should be able to take some amp power, which is needed to force them to low frequency.

Attached is a picture of a large and a smaller woofer in standby mode.

 

[MAB]

View attachment 425749
pre-breakin (darker colors) vs post-breakin (brighter colors) impedance plot. The general trend is that the resonance peak shifted up (yes, up!) in frequency by 5-10%, and down in magnitude by around 5-9%. I tried to isolate some of the fan-noise but its still noticable, and there is definitively power-noise at 50hz, but overall it seems about right.

Seems like pretty big difference between them?

TS parameters for driver1 (the red one):

View attachment 425750

What do you suppose is 'breaking in'?
They are used drivers, right?
Even if new, no way changing by more than a small amount.
I don't understand what you are measuring here.

 

[haavind]

What do you suppose is 'breaking in'?

I don't know. I was hoping someone wise here could enlighten me because this was the opposite of what I expected. A softer spring should lower frequency. Less damping could increase frequency somewhat but should also increase the peak. Qes is an electromagnetic parameter right? It appears to have changed, does that seem weird?

The drivers are about ~25 years old, but have not been played for at least 10. My measuring rig might be wrong, or badly calibrated, but it is the same as it was yesterday. I'm not sure if I used properly non-inductive resistors. Dunno what that would do to the result.
I can speculate on a hypothesis, but it is practically unverifiable:
"A stiff surround&spider couples to the cone as an extra mass. It can also have rested in an offset zero-point."
Calculated Mms have indeed gone down, maybe by about the mass of part of the surround.
I don't think it matters practically though. it is what it is (probably).
Maybe I'll play them some more and see if it stabilizes.

Attached is a picture of a large and a smaller woofer in standby mode.

A larger and smaller opposing pair with shifted color phase. cardioid setup <3

Anyway, for a sub these seem to be not the best candidates

Yeah, I am considering alternative uses.
I don't think the drivers themselves are crap. There's a bit of noise (rattling/buzzing/clicking) when I play pure test tones of very low frequency, but with music I can't really tell and the harmonic distortion numbers seems fine. Just lying open and flat on a mattress, with (significant) EQ to taste they are actually enjoyable to listen to. they have stereo imaging and plays fun and dynamic, although lacking the final "airiness".

I'm just thinking aloud here:
I have one roughly matched pair, red and green.
They resonate at 58.3hz +- 0.1 and have a Qts of 0.55 +- 0.1
They should probably do something together.
I tried putting the numbers into an enclosure calculator and it said I needed a 56l sealed cabinet per driver if I wanted a Butterworth sub. A bit on the large side, but manageable.
Two sealed ones would not play deep enough to really be of interest in my main system it seems. -3dB at 70hz.
Vented would be to large for me I think and I don't want vented anyway.

The other two resonate at 52hz and 68hz, with a Qts of 0,49 and 0,627 respectively. They are pretty different and the second one is basically a bit too resonant for a reasonably sized cabinet if I want decent sound quality?

That gives me a few ideas.
For the matched pair.
1) car sub: Any low-end extension would be an upgrade from the current situation, sound quality is of less importance, and especially ULF just drowns in roadnoise anyway. I might also find a way to use the back of the car as a semi-infinite box.
or 2) A pair of single driver dipoles next to each main speaker. The problem frequency is around 80hz, not too low, and according to the cardioid paper I found it doesn't need equal volume to get SBIR benefit. Even a -10dB driver can help, so an 8" paired with a 10" could be perfect.

The unmatched pair (blue and purple) can make a sealed and open baffle pair with the low Q element in the box.
Get a suitable upper-midrange or midrange&treble driver and make a simple wide dispersion mono-speaker for radio and casual listening at the cabin.
I need to think some more about the details, but something like this:

 

[AnalogSteph]

Yes, am fairly certain! Since it's a couple harmonic oscillator, my intuition usually fails me!!!

I didn't change the cabinet, just aligned it to the CWJ20-19-08 specs (VituxCAD already has the parameters in the database! )

Now those results are just like I would have expected. The lower-Q version has a droop above port resonance (much like in a closed box), and port output effectively is marginally right-shifted because the woofer's response is more heavily tilted in the frequency range in question. That's quite a small effect though and nothing like what your first comparative graph was showing. I would investigate what happens using a CB in general, going ported just seems like an unnecessary complication.

pre-breakin (darker colors) vs post-breakin (brighter colors) impedance plot. The general trend is that the resonance peak shifted up (yes, up!) in frequency by 5-10%, and down in magnitude by around 5-9%. I tried to isolate some of the fan-noise but its still noticable, and there is definitively power-noise at 50hz, but overall it seems about right.

Seems like pretty big difference between them?

How well-controlled is your ambient temperature? That's probably the single largest environmental factor influencing fs. There have been cases of "burn-in" where the effect was shown to be entirely due to the driver having warmed up while playing, with fs basically returning to what it was originally after things had cooled back down.

I am still puzzled by all the noise on your measurements. Could you show what your impedance measurement cable looks like? It should be as shown in the LIMP user manual. While pickup of acoustic noise is fundamentally possible, speaker drivers are a fairly lossy affair overall and any happenings on the acoustic side should only reflect weakly on the electrical one. Have you tried increasing output volume / generator output level?