I am currently putting together a system for use at small to medium festivals.  At the moment i am designing a bent horn for 50-150hz that will use the 21DS115.  I have completed the prototype and am pleased so far with the sound.  Next I will be designing a 25-50hz box for under the bent horn.  It will most likely use the 152's.

Im probably going to use k20's on the bent horns and maybe the subs, but may stick with the 9001's on subs.

Now that I am moving up to modern amps I am wondering how to best set the long term limiters.  The idea of having 225 peak volts is really exiting and the drivers can certainly take it.  However the average level of this new dirty trap and bass music is very compressed and $300 recones is not sounding fun.  

One of the reasons I am so excited about the newer amp is the ability to have large voltage swings and limit the average power to keep driver heating down.  Should give more confidence when running hard.  

The box i built for the 21DS115 needs 160Vrms to even reach xmax in the passband.  Anything less would be wasting the driver.  For EDM in the past we undersized the amps and hard limited the subs for driver protection.  This I believe hurts the dynamics of the system.  Its always some guesswork to match the drivers and amps.   Hopefully, with the new amps I can find the balance without cooking too many speakers.

I did some goa trance gig a week ago, the leds on the midhigh boxes were basically lit all the time with very little dynamic. Had to reduce a lot to prevent blowing up the HF. 

Very interesting read.  They show some experiments that are pretty eye opening with regard to power compression.

The 21DS115 is not a "normal" 21". 

 Im not sure what you mean "cone area not important" ?  More displacement should equal more SPL?  Am i missing something?

turbo7 wrote: You dont need that much cone area for 50-150hz, you will not really achieve more output and you would probably get a better sound with a 15". The price difference should also be quite significant

Simulations are saying otherwise.  Also, I A/B tested my prototype against 3 X HD215's eq'ed to match as closely as i could get.  The hd's are loaded with EV DL15X's (EBP = 140).  The single 21 had no less "punch" than the 6 X 15's.  Sound quality was very good.  Obviously the the 6 15's were louder but not by much.  And the 21 when not eq'ed went at least 10 cycles lower.  

The 18DS115 looks like a nice woofer.  However, the 21 is a much more efficient driver.  I should have mentioned I plan on using the box as a general purpose sub down to 40 cycles for non EDM gigs.  The extra surface area should be helpful here.  Good point on the possible thermal problems if the cone is not moving enough.  That kind of goes back to my original post..... What long term power are people putting to these things?

Not sure I follow.

Anyone have any real world experience with the 21DS???

Keep in mind the dl15x is a 30+ year old design.  Not a bad woofer, just old.  Well suited for a BP horn.

dl15x  sd= 860 cm2 x .41cm xmax = vd 352.6 cm3,  EBP=140

The hd215 is bandpass horn, so totally different to the folded horn I have the DS in.  The hd215 has lots of efficiency in the 160-220 cycle region.  I wanted to compare the 21 box to  a known box i have used a million times.  I had to eq out the top end of the hd and the low end of the 21 prototype to match them up for subjective sound quality listening.

I would like to sim this.  What are some modern drivers I should look at?  I tried b c sw115-4.  I can get similar response to 21ds with a horn as you describe.  However, it shows a bit less sensitivity ~2db and 3-4db less max spl.   

I guess my design is a bit different from a standard FLH.  One of the design parameters was to have the driver "backwards" to allow better cooling of the VC.  The other was to have symmetrical loading in the throat 

The dark plot is your sim at 2 pi  the other is my design.  The standard FLH will go a bit lower in multiples.  The loading I came up should be much better in singles and the the mouth is not much bigger.  I am curious how the RCF 15 would work in a loading like mine?  Will play around with it tonight.  Funny thing the RCF driver is $100 more than the 21DS here in US.

Here is the plot at 105v.  Note this is the 4 ohm version of the 21DS.  So around 2500 watts.

I do think the volume in front is having a bandpass effect.  The rear volume of the woofer occupies this space too.  And the magnet sits in the "neck" area.  I subtracted the area of the magnet from the "neck" cross section  (S2).    The shape of the woofer plays into the first segment actual shape and expansion.  Not sure how to model it in hornresp.  Should that area be considered a throat chamber or a neg expansion like I have it?  The prototype measures very close to the sim.  

 

http://data-bass.com/data?page=system&id=144&mset=169" rel="nofollow - http://data-bass.com/data?page=system&id=144&mset=169

 

If you aren't familiar with navigating DB if you click on most of the titles or names they are links to more information. There are currently 4 measurements sets with notes for the 21DS115-4. There is a lot of information there.

 

100V into the 21DS115-4 could be dangerous to the driver thermally over the long term with high duty cycle.

Thanks  That's what i was looking for.  I can't really do full power long term tests in my shop due to neighbors.  

You put some time into those measurements!  

How long are the sine sweeps for the "long term" power measurements?

 

The sine sweeps are ascending covering the range of 2Hz - 240Hz (240Hz needed for measurement resolution to 120Hz.) Duration is roughly 22 seconds. It is brutal at the higher voltages.

 

It is very demanding. The reason it is done like this is an effort to produce an accelerated heat buildup. I don't have time or the area for a whole other setup for much longer thermal compression testing over many hours with pink noise. This type of thing is already done with AES testing anyway. Also it will change drastically depending on the enclosure used and signal applied. This type is more abusive over the short term but not as abusive as applying a sine wave at a worst case frequency. At tuning for a vented cab for instance. Most drivers seem to bounce back from being driven to 3-4dB of thermal compression once or twice. I'm not sure that would be the case if heated that much many times. Of course I've burnt up some drivers and also broken some mechanically with this type of test. Using that type of signal also pushes the speaker extremely hard at all of its weak points in the sub bandwidth so it reveals a lot of warts, distortion and other bad distortions or sonic issues. Often driver excursion becomes a limiting factor right at about the time the driver is running into serious thermal issues. Whichever one throws up a flag becomes the limit for that test type.

 

Don't be scared  Pedal to the metal

Seriously though,  If your not putting at least 4Kw peak RMS to a 21SW152 you might as well get a cheaper woofer.

The available wattage calculated from the maximum RMS voltage capability of the amplifier.   Roughly .707 X the Rail voltage.  Or the "continuous program" rating used by most woofer manufacturers.

Hey. Digging old corpses I guess...

 

In B&C, they say they put more wire into the magnetic gap with 21DS115, for motor strength. I still cannot wrap my head around how they did it. The math doesn´t work here (Resistance, wire diameter and length) for me. Looks like it is not possible with Aluminium. Yet they did it somehow.

That is not the topic though.

 

The more I think about it, the more it seems to me that the Al coil must be beefy one. Significantly greater wire diameter and volume than with Copper. And thicker coil body too, obviously. That might mean good things for cooling capabilities, couldn´t it? Greater coil area and volume/heat capacity. And the weight is also there. Mms of 21DS115 is greater than with 21SW115.

 

The 21SW152 doesn´t have much more coil area than 21DS115, and that is not counting for split coil of 21SW152. It might be, that the coil area of 21DS115 is the same as 21SW152, despite significant difference in the coil diameter. That would certainly make Ricci´s tests slightly less "unbelievable".

 

 

Quoted from :  https://data-bass.com/drivers/5af4b3c1e8c90b0004781fa9" rel="nofollow - https://data-bass.com/drivers/5af4b3c1e8c90b0004781fa9

In time, these drivers BOTH rose and dropped in my evaluations. In my usage, it is not enough coil, and 2000+ model is very needed, in order to work its purpose. On the other hand, for peaks, with these kinds of Bl, in 21" size it starts to show "that scary" bass, and it is able to blur my vision when I turn my back to it. So it is not that bad. 

In the meantime,many thongs moved and changed. I have tested many more speakers, Celestion TSQ2460 arrived yesterday, and soon I am deploying a review platform.