Cab Design: Avatar 2x12 w/ Greenbacks: Way Too Boomy!

[bob p]

BTW, I saw this piece by Celestion's "Dr. Decibel" that explains their company policy of not publishing Thiele-Small parameters.

http://professional.celestion.com/gu...netdesign.html

To me that says "Our guitar speakers are so non-linear that we can never get the same answer twice when we try measuring the T-S parameters, so we gave up trying" ;-)

Yes -- I think that if you read between the lines, they're saying , "we don't have a statistical process control engineer in our factory, and our unit to unit variance is horrble." LOL.

Seriously, though, you have to pay close attention to how they choose theier words. They say some very interesting things. for example:

... the frequencies at which Thiele Small parameters have significance are mostly below the [guitar's] operating range.

Well, its unfortunate, but the "Celestion Boom" occurrs right there IN the guitar's operating range! Using T-S parameters to eliminate the BOOM would be a very helpful way to approach the problem with surgical precision.

Later on, they go on to say something that's more interesting -- if you read between the lines, it sounds like they're trying not to admit exactly why they don't want to provide their T-S numbers:

Care should be exercised designing ported (or reflex) cabinets for guitar speakers as the increased cone excursion below the tuning frequency can cause speaker damage.

The "speaker damage" relates to a speaker that exceeds its maximum linear displacement (Xmax) when producing low frequency signals. Looking at spec sheets, Xmax forthe EVM-12L is 3.3 mm. Celestion won't disclose Xmax for their speakers. Eminence makes a number of speakers that are patterned after the Greenback. The Eminence Governor has an Xmax of 1.2 mm, and the Legend GB128 has an Xmax of 0.8 mm.

YIKES!!!

Looking at the numbers, the EVM-12L is happy as a clam to live in a T-S array, and the Eminence Greenback-clones will probably blow up if you hit them with any real amount of power below the port tuning frequency. Celestion won't even tell you how to calculate the box tuning frequency, because they just don't want you to go there. Their numbers must be really, really, BAD.

That's too bad. Using the EVM-12L as a guinea pig, I crunched some T-S calculations and found that I could easily cure the boominess problem with surgical precision by porting the cab. My calculations showed that a 6-inch port 6-inches long would tune the box in step-down mode so that rolloff started at about 200 Hz and removed 5 dB of frequency response between 80 and 160 Hz. That would have been a wonderfully precise solution to the boominess problem. Its too bad that a mathematical approach to the problem is likely to blow up my Greenbacks. Sigh.

[Steve Conner]

I wouldn't be at all surprised if the Celestions turned out to have a minuscule or maybe even zero Xmax. You can actually have Xmax=0 If the voice coil is the same thickness as the front plate.

That would be the most likely explanation for being unable to measure the T-S parameters: the mere fact of working up enough cone excursion to measure them would cause enough non-linearity to change them significantly. Even if every speaker were identical, you could get any answer you wanted by changing the drive amplitude. More importantly, a T-S cabinet tuned for one volume level would be out of tune at higher cone excursions.

As far as I know, Xmax isn't a maximum *safe* excursion: it's a maximum linear excursion. Above Xmax, bits of the voice coil start to leave the gap, which changes the value of the B-L product or whatever, and the harmonic distortion starts increasing rapidly. Which just makes even more flabby, throbbing synthetic bass. To be fair, the sound of Celestions is exactly what a large portion of the market wants, and I think Celestion do have more linear guitar speakers in their product line.

P.S. If you start designing T-S cabinets for use with tube amps, remember to take the output impedance of your amp into account. The usual equations assume a solid-state amp with an output impedance as near zero as makes no odds. Tube amps are usually much higher, I have heard figures of up to 18 ohms for an 8 ohm tap.

I found that ported bass cabinets designed for SS amps sound very fat and boomy with my tube amps, as the reduced damping makes them grow humps at their resonant frequencies. (Ported cabs have two bass resonances.) Changing my power tubes to 6550s from 6L6s tightens it up, as tubes with more gm give more effective NFB and lower the output impedance. Also, I have my secret weapon, a tube amp with 2-band parametric EQ :-)

[bob p]

I found that ported bass cabinets designed for SS amps sound very fat and boomy with my tube amps, as the reduced damping makes them grow humps at their resonant frequencies. (Ported cabs have two bass resonances.) Changing my power tubes to 6550s from 6L6s tightens it up, as tubes with more gm give more effective NFB and lower the output impedance. Also, I have my secret weapon, a tube amp with 2-band parametric EQ :-)

Heh. We're always circling back to EQ as being the magic bullet. Somehow I just can't see adding a parametric EQ to an old Tweed Super. LOL.

Yes, I appreciate that we're taking a pair of GBs out of their intended operating environment when we think about T-S porting the cab. I guess in some respects, its not fair to think about T-S parameters with some speakers, so maybe the only options left are to enjoy the boom or to move on to other speakers.

As an interesting side note, I learned something very interesting in all of this -- the Avatar 2x12 cab has almost exactly the same internal cabinet volume as the EV TL-806D (2x12) T-S cabinet. The internal volume is only off by about 3%, which is "neglegible" by T-S standards. So if anyone is interested in rolling-up a T-S cab for a pair of EVM-12L, it would be very easy to obtain an Avatar cab (here in the States), add the port, and drop in the pair of EVM.

Just in case anyone doesn't know this, the original Mesa 1x12 Thiele ported cab that came with the EVM-12L was nothing more than an EV TL-806 (1x12) Thiele cab wrapped in tolex. (I guess that Mesa didn't feel the need to re-invent the wheel.) I haven't heard any horror stories about speaker blowups in that application -- and these setups were hooked up to LOTS of Mesa amps. What really amazes me is that players who don't know better would pull out the EV and put a Celestion in there, which is just asking for trouble. I guess that using step-down tuning, to take fb down from 75 to 53 Hz protected a lot of "inferior" speakers. But that's the beauty of the T-S alignment in a case like this -- by squashing down the bass response, frequency response in the low octave of the guitar is exchanged for frequency response in the low octave of the bass -- a band that we're not planning on using anyway. So throwing away 5 dB of response in that band solves our problem. And the Fb is low enough that it shouldn't hurt our speaker anyway.

Putting all of this into perspective, I've been using that TL-606/EVM-15L for guitar and bass with both SS and valve amps for 30 years now. So far, so good... knock on plywood...

[Ray Ivers]

Bob,

Here is an extremely bare-bones set of T-S parameters for the Celestion G12-M (hopefully it will display properly):



Ray

[bob p]

Thanks for the help Ray! Where in the world did you find those T-S parameters? Did you measure them yourself?

I have to admit, I'm getting a bit confused. AFICT, the G12 "Greenback" that I'm dealing with has an Fs of 75 Hz. What confuses the issue more is that Celestion now has two different versions of the G12-M Greenback:

1. "Classic Series" G12M Greenback (click here for spec sheet)
2. "Heritage Series" G12M Greenback (click here for spec sheet)

FWICT I have the "Classic" series G12M. It reportedly has an Fs of 75 Hz. I notice that your parameters list an Fs of 92. Did you get that data from a spec sheet, or did you measure it yourself?

Thanks again!

[Ray Ivers]

Bob,

Thanks for those spec-sheet links!

The parameters I posted are a screen shot from HT Audio's BassBox Pro program (the database has a ton of different drivers). I don't know if HT Audio got them from Celestion or measured it themselves. FWIW, it's always seemed to me that the various flavors of Greenbacks had higher resonant frequencies than those listed on the labels - I have three '68 55Hz ones here that I sincerely doubt go that low (I haven't measured them, just a gut feeling).

If you'll be using a program to design a cab, you could try plugging in different Fs values and see how much it affects the design dimensions, or maybe do a quick by-eye frequency sweep on a broken-in speaker for a rough indication of the actual production-speaker Fs.

Ray

[mojah]

QTS High=very big box=flat low end

The qts is high for the greenbacks so if they are in a smaller cab they will be boomy. I tried to emulate my 4X12 mesa in a 2X12 with vin30's and it's boomy too. They need to be in a bigger box to get the low end flatter. You could try your stuffing fix. It would to the speaker look like the box is ~15% bigger and deaden any standing waves too. I just got onto the qts ratio from building home speakers. You may end up having to lose the greenies if you want to keep the cab.

[Steve Conner]

You may end up having to lose the greenies if you want to keep the cab.

Yes, Bob, why don't you just put a pair of EVM-12Ls into the Avatar cabinet and add a port to the back panel? You know you want to

[bob p]

The qts is high for the greenbacks so if they are in a smaller cab they will be boomy. I tried to emulate my 4X12 mesa in a 2X12 with vin30's and it's boomy too. They need to be in a bigger box to get the low end flatter. You could try your stuffing fix. It would to the speaker look like the box is ~15% bigger and deaden any standing waves too. I just got onto the qts ratio from building home speakers. You may end up having to lose the greenies if you want to keep the cab.

What kind of value for Qts are you referring to for the Greenbakcs? Are your numbers similar to Ray's? Of interest is the fact that my Fs number is widely different, so if you could post a complete set of TS numbers I'd appreciate it.

Well, I have to admit that I've set that cab aside for a while while I've been playing with a different cab.

I haven't tried the stuffing fix yet, though I have done a little bit to try to hasten the break-in to get rid of that shrillness that came with the new Greenbacks. I hooked the speaker cab up to a stereo receiver and blasted out 24-hours of program material to try to break in the cones a little.

Although dance music and rap seem to be common recommendations for this sort of thing, I don't have any techno or rap CDs, so I opted for just putting a few guitar CDs on auto-repeat. I was hoping that by playing 24 hours of live Zeppelin through them, the Greenbacks would learn what they're supposed to sound like. But seriously, I fed them some Zep, AC/DC, RCHP, Stones and the like for 24 hours and gave them another listen.

The good news? Well the shrillness isn't as bad as it used to be, but its not entirely gone. I'm thinking that if I did this for another 24 hours the speakers would sound noticeably better, but I'm not sure if we (or the neighbors) could stand another 24 hours of blasting sound.

The bad news? The boom is still there, but it seems a bit less objectionable now -- for whatever reason. I guess if break-in isn't the answer then I'll have to try a) stuffing, b) open-backed, or c) new speakers.

[bob p]

Yes, Bob, why don't you just put a pair of EVM-12Ls into the Avatar cabinet and add a port to the back panel? You know you want to

You don't know how close I came to buying a pair of EVM-12L on eBay that were within driving distance for a pick-up. Right now I have a 1x15, two 2x12 cabs, and 3 extra unmounted 12-inch speakers in the living room, and I got the impression that I might wind up DEAD if I added another cab or another pair of speakers to the mess...

Looking at the Avatar cab, I'm thinking that I'd prefer the port to be on the front, but that a 6-inch diameter port is kinda big and might be hard to fit on the baffle. I might be able to get away with cutting a pair of 4 or 5 inch circular ports on the front panel instead. Hmmm...

[Thorny]

Weber's Michigan is designed off an EV - and it has the same cone on it as the Zack Wilde speaker from my understanding. And Weber rates his at 100w. I know he is conservative on his ratings. I have not tried the speaker but a friend has the 15" version. It is really a clean speaker. If in doubt ask Ted Weber. He gives a straight answer. He can be more specific than I can be.

[bob p]

Thanks for the recommendations. I've owned Michigans, Californias, EVs and JBLs. They're all different.

Like Celestion, the T-S parameters for the Weber Michigan are not readily available. Not surprisingly, EVs and the Michigans don't perform identically with high level signals. They are NOT the same speaker.

Part of this might be due to the differences in magnet structure. The EVM 12L uses a 4.9 lb magnet. Maybe if Ted reads this he could give us some information. I have to admit that I don't know the weber specs off hand, but its magnet appears smaller to me. Lets look at photos of the 12 inch EV and the 12 inch Weber:



Part of the sonic differences might also be attributable to differences in the cone. Its interesting to hear your report that the cones are identical. I've compared them side by side and the cones were different. The original EVM speaker has 5 ridges in its paper surround. Look at a blown-up pic of the Michigan and count the rings. It looks like even the current production Michigans have a different surround.

What's left to compare? Voice coils? I'm not willing to dissect a perfectly good pair of speakers to comapre the voice coils. Maybe Ted could give us some more info, like whether they're edge wound or not.

Suffice it to say that while the Weber Michigan might offer some similar tone characteristics to the EVM speakers (so do the Eminence RWB and Tonker, BTW), the Michigan isn't an exact replica of the EVM line, and its not a sonically transparent replacment for the EVM 12L or 15L.

Similarly, the JBL D series speakers aren't easily replaced by Calis. Calis sound a lot rougher than the JBLs when you're hammering them.

I'm not bashing Ted. I think its great that he's making the speakers that he's making. Still, not all speakers are the same and there's a reason that people will pay extra to own the genuine EVs and JBLs. With that in mind, the California, Michigan, Red White & Blue and Tonker are a great way to get into the same ballpark without having to spend a lot of cash, but they are NOT the same speakers.

[mojah]

My Ts

I have the ts for the vin30 but I don't think they are 100% correct I was referring to the one posted above. Small box speakers generally have a qts below .5. They can go in smaller sealed enclosures and sound relatively flat. The higher the qts the larger the enclosure has to be. Now a lot of old design hi fi speakers were put in giant boxes... Remember them? It was for a reason. If you put a higher qts speaker in a smaller enclosure the system qtc rises and you get a pronounced mid bass hump. Depending on the speaker it can be ok or annoying... If it were me I'd just loose those greenbacks and look for some ev's or maybe an eminence delta pro 12a. Talk to Jim Weber he is really knowledgeable about speakers. He also has a technique of spraying fabric softener on the cones to mellow out the top end. I run a 6.3v transformer hooked up to my variac then ac line so 60hz goes through them for 24 hrs to loosen them up. I think weber has/had a calculator for doing that. anyway the v30's qts .56 qes .409 fs 70hz spl 100db

[bob p]

Thanks for the help.

Regarding the Greenbacks, I took a different approach in breaking them in.

I couldn't stand the monotone 60 Hz signal, so I decided not to go the variac route. (I also have a very real fear of the variac dial getting bumped and my speakers going POOF!).

I also couldn't stand the annoying "whoop whoop" of my sweep function generator running through selected frequencies for the speaker. I guess it would be okay if I could do it elsewhere, but I'm stuck with doing break-ins at home and sonic tolerability is an issue.

My compromise has been to do it the low-tech way -- I've just pumped programmed material through a CD player and a stereo for hours on end. I put them in the bedroom at the far end of the house and run them all day long for a few days.

I checked the GBs at 24 hours and the shrillness was markedly improved. They're developing a very sweet character with P-90s, but they are still quite shrill with a strat and there's still room for improvement. They're on-deck to go into "the booth" for another 24-hours as soon as my cab with the Tonker/Governor finishes its 50-hour pounding. If all else fails, I may just have to run the GBs in an open backed cab.

I have to admit, I'm afraid of chemical break-in processes. They're all designed to perform a "controlled aging" of the speaker in some way. All of these methods chemically accelerate speaker wear and end up shortening the cone's life. I'm not sure of the stoichiometry that's required to allow the chemical reaction to only go so far and then stop. Being a chemist, if I were to start a chemical reaction that was intended to degrade the cones, I'd want to be able to apply a second reagent to quench the reaction. Then how I'd get all of this stuff out of the cones is another problem.

I guess I'm conservative in that I like to break-in speakers the old fashioned low tech way -- by playing music through them. Besides, its alot easier to listen to that repeating AC/DC CD that's playing in the distance on autorepeat than it would be to listen to sine waves...Ick!

Regarding those old speaker boxes... you made me think of the huge EV Patrician when you mentioned that!

[Steve Conner]

Jeez! There's really no contest in magnet size there.