M14s: I may have a pair

[cradeldorf]

That truly is a miracle in a can!!! Those look very much like brand new.

[Alien_Shore]

That truly is a miracle in a can!!! Those look very much like brand new.

Yep, thanks CD - if you have any old cabs that could use a freshen-up, you gotta try that stuff!

Well, the woofers came back from GPA today, so I could finish up the refurb.
Here are the impedance sweeps of the re-coned woofers.

woofer 1:



woofer 2:



I have also now finished up the repairs on the baffles. Here's the woofer tee-nut - Altec should've used the brad-nailed ones in the first place:



The hardwood feet are now stained, drilled, and attached (those nail-on glides might be too big - I'll look at Lowe's next time I'm there to see if they have smaller ones):



The new woofer cones look fabulous...thanks Bill H @ GPA!



And back to listening! The old girls have gotten a nice 30-year upgrade...

[Alien_Shore]

Here are the crazy-looking measurements for the rehabbed Model 14's:



This is using the refurbed stock crossover. These results really make me wonder about this crossover, and/or my measurement technique. It was interesting how much differently I needed to set the XO shelf controls to get the curves to line up close.

[cradeldorf]

21 and 22 on the FS I think that's awesome dude, what are those 12 inchers?

[Alien_Shore]

21 and 22 on the FS I think that's awesome dude, what are those 12 inchers?

Yes sir, the M14 has a 12" driver.

Well, I played around with the measuring rig and crossovers tonight for a while. I was trying to get the flattest response I possibly could from the M14s.

As you can see in the previous post, with the stock XOs there is some strangeness happening between 400 and 2K.
I experimented quite a bit with the shelving controls (min, mid, max, "optimum") for both MF and HF.
Then I broke out the active XO and bi-amped. Here's the best I could achieve:



This is a 1.5KHz crossover point, using butterworth-12.

What was odd to me was that by rolling off the woofer more steeply, say 24 db/oct, the midrange response improved. I have a feeling this is due to some phase issue or cancellation happening, or maybe I just don't know the tricks yet to get an accurate measurement with two drivers. Anyone with more experience care to help out?

[Earl K]

Mike,

You should remeasure the FR ( Frequency Response ) of your woofers ( using the same near-field manner like you did back in July ) to get a baseline for it's midrange response with the new recones . Have you ever run those ts numbers through a box-tuning program ( such as WinISD ) ? The super low Qts is a cause for some concern ( both theoretically and practically, by the looks of your measurements ) .

FYI, using a vertical scale ( as displayed on screen ) of 10db per inch gives a super critical look at whats going on ( when coupled with "no windowing" ) and 1/12 octave smoothing .

I would take many more measurements to try to figure out two important mysteries ;
(1) Where has your midrange response disappeared to ( above 450hz )?
(2) Why is there such a lack of measurable bass ( below 200hz )?( when your earlier nearfiled measurement certainly wasn't lacking )

For instance , the following picture ( a near-field study of your original M14 woofer ) shows a beautifully balanced response ( making it easy to see why Altec felt it un-necessary to stick an inductor into the network to achive a 2K roll-off ) .

.

- I would remeasure one cabinet with the test mic 3ft away & located @ half-way between woofer & horn ( aimed at the speaker ) to figure out the negative contributions of your rooms acoustics . Do this in many locations .

- Also, take some measurements with the test mic 4ft off the floor and further back ( from the speaker ) . Something like 6 to 10ft.

- Also be sure to move the speaker a couple feet forward of where you presently have it ( for a handful of these new measurements ) .

[Earl K]

BTW, what does the response look like with the polarity of the horns flipped ( using the original crossover ) ?
Does the area between 400 - 2000hz change any ?

Also ( so that I'm not misunderstood ) , from what I see, the TS of your rehabbed woofers are nicely in line with the Ts of the originals ( as posted by Richard C. some years back ) .


Still, I'd like to know "where has the bass & upper mid (gone)?"

There are some real unexplained paradoxes within this thread .

[voice of the theater]

Still, I'd like to know "where has the bass & upper mid (gone)?"

There are some real unexplained paradoxes within this thread .

I agree--there's either something wrong with your Model 14's or something wrong with your measuring procedure. If your measured response curve is accurate--you should be hearing a huge hole in the upper midrange and they are also lacking in bass and upper treble/sparkle. If they don't sound that way, maybe they're fine and there's a problem with your measurements. Mine have very present upper midrange and more than ample low bass and h.f. extension/sparkle.

[Alien_Shore]

Mike,
You should remeasure the FR ( Frequency Response ) of your woofers ( using the same near-field manner like you did back in July ) to get a baseline for it's midrange response with the new recones . Have you ever run those ts numbers through a box-tuning program ( such as WinISD ) ? The super low Qts is a cause for some concern ( both theoretically and practically, by the looks of your measurements ) .

FYI, using a vertical scale ( as displayed on screen ) of 10db per inch gives a super critical look at whats going on ( when coupled with "no windowing" ) and 1/12 octave smoothing .

I would take many more measurements to try to figure out two important mysteries ;
(1) Where has your midrange response disappeared to ( above 450hz )?
(2) Why is there such a lack of measurable bass ( below 200hz )?( when your earlier nearfiled measurement certainly wasn't lacking )

Firstly, thank you very much Earl for taking the time to help - I very much appreciate your questions and direction with this.

Here is a NF measurement of the re-coned woofer. It looks similar to the earlier measurement, with the dip of about 3db at 500Hz...but certainly there are some concerning differences, especially with rolloff on both ends. Could all of this be due to a recone? Maybe they need to break in some?



I've not used a box-sizing program yet, that will be a new field of study for me. I will look into that this week.

- I would remeasure one cabinet with the test mic 3ft away & located @ half-way between woofer & horn ( aimed at the speaker ) to figure out the negative contributions of your rooms acoustics . Do this in many locations .

- Also, take some measurements with the test mic 4ft off the floor and further back ( from the speaker ) . Something like 6 to 10ft.

- Also be sure to move the speaker a couple feet forward of where you presently have it ( for a handful of these new measurements )

For all of these new measurements, the cabinet is at least 5 feet from either side wall, 6 feet from the back wall, and 21 feet from the wall opposite the front baffle.

OK, first up is the woofer only @1m, comparing the cabinet on the floor vs. elevated ~17". Mic is aimed on-axis between woofer and horn. Looks like the floor is definitely impacting the response! Although why the floor would contribute to the huge trough at 580Hz I just do not know :



Next, here's just the 902 @1m (cabinet on the floor). Mic is aimed on-axis at 21" height (midway b/w LF and HF):



And the merged graph of those two measurements - this just looks awful to me:



Next I added in the stock XO, and took 3 measurements @1m, mic again aimed on-axis at 21" height. This compares the settings of the stock crossover shelving controls at min/min, mid/mid, and max/max:



This certainly explains why I'm on this quest - I know these speakers should sound better than they do. I suppose with enough EQ to boost this hole up, it's liveable but I want to get this fixed without EQ if possible. I tested both cabinets, and there is very little difference between them, so that rules out inconsistencies with the re-coned woofers or the XOs that I re-capped.

BTW, what does the response look like with the polarity of the horns flipped ( using the original crossover ) ?
Does the area between 400 - 2000hz change any ? Still, I'd like to know "where has the bass & upper mid (gone)?" There are some real unexplained paradoxes within this thread .

Here is the comp driver reverse polarity data. This is with the XO set at "optimum" settings, comparing the 902 at normal / reversed polarity:

I agree--there's either something wrong with your Model 14's or something wrong with your measuring procedure. If your measured response curve is accurate--you should be hearing a huge hole in the upper midrange and they are also lacking in bass and upper treble/sparkle. If they don't sound that way, maybe they're fine and there's a problem with your measurements. Mine have very present upper midrange and more than ample low bass and h.f. extension/sparkle.

Agree with your comments. To be sure, there is certainly something missing with the sound of these. As I have tried out different recordings, different genres, different EQ, it has become increasingly apparent that they can do better. The bass is pretty good, actually - I did find that my earlier measurements were starting the sweep at 100Hz, so by backing that down to 20Hz, the graphs look a bit better, and the tones below 100Hz certainly were causing some noticeable vibrations and reflections from stuff in my room (which by the way is admittedly horrible in terms of acoustics - it doubles as our home gym so there are mirrors all down one side wall, and the ceiling is only 8', and I'm sure with the regular rectangular shape it's creating some wicked standing waves...

Here are the measurements from 2m away. This chart compares on-axis versus 45-degree off-axis:



Again, my sincere thanks for the help guys.

[Earl K]

Nicely formed "V"-Shaped notches ( suckouts ) are sure indications of a hard reflective surface combining ( at 180deg phase flip ) with the main wave .

- In your case ( the 580hz notch ) , it's a good bet it's due to a hard ( cement ??? ) floor below that rug ( giving an almost pure reflective bounce ) .
- 1130/580 = 1.95ft x 12 / 2 = 11.7inch ( which looks about right , as the distance from the center of the woofer to the floor ) .

Other peoples floors ( if more porous , like a typical 1st floor made of wood ) will give different midrange results ( a wider Q with less attenuation in the notch ) .



Speaker components do have a breakin cycle . I would expect that to happen to yours / though I'd be shocked if it ever turned itself back (100%) into this ( time will tell );



I'm not sure what to suggest as a remedy ( since none I can think of are simple ) . Since I own a ton of stuff I'd be inclined to see if lashing up an ( ad-hoc ) MTM setup might restore some of the lost mid . Another 12" ( in a sealed 1.5cu' box / like a 2206H ) would be worth a try .