JBL DMS-1 or very close?

It was three or four years before the introduction of the M2 that I saw a pair of 2450SL drivers on 2332 horns in an eBay listing and thought, "That could be the start of a DMS-1 build." Now, years later, the project is finally completed, except for some final tweaks such as new internal wiring and SpeakONs to replace the old binding posts.

No surprise that completing the project was only possible with information found on Lansing Heritage and from assistance in various forms provided by such Lansing Heritage stalwarts as Dave M, Guido, Greg Timbers, subwoof (aka overtweet), speakerdave, and sebackman. (Apologies to those whom I may have omitted to mention!)

In brief. . .

— The cabinets are original, early DMS-1 cabs (serial #003), recently acquired from subwoof.

— The woofers are not 1401ND, but 1400Pro cores reconed as 1400ND with aftermarket cone kits acquired by The Speaker Exchange. Are they identical to the original JBL kits? I've never had the opportunity to compare the two, but I can't complain about their performance.
— As mentioned above, the horns are the 2332 used in the DMS-1. The compression drivers are not the 2450SL-A, but the much more common 2450SL, the only difference being, if I understand correctly, that the SL-A diaphragms have both acquaplas treatment and the "diamond" stiffening ribs, while the SL diaphragms lack the latter.

— For speaker controllers, no final conclusions as yet. I also acquired from subwoof one of his SMC24 analogue units, which is presently with a local tech for testing and refurbishing. I started out using the "DMS-1A" preset on an NOS JBL DSC260A, which had sat sealed in its shipping box since it was manufactured in 2002. Lots of drive and energy with this speaker controller, but rather crude sounding. Next I programmed the JBL-published settings for the DSC280 into a dbx DriveRack 4820 that I have used in other speaker projects. The improvement was striking—and not just to my years—and that's what I'm using now. Will see how that compares to the speakers' performance with the refurbished SMC24 analogue controller once I get it back from service.

— Will eventually embark on formal measurements using the REW microphone and software, which may lead to some additional tweaking for my particular room,Attachment 84467Attachment 84468Attachment 84469Attachment 84470Attachment 84467Attachment 84468Attachment 84469Attachment 84470 and will probably replace the internal hook-up wire and binding posts (prefer SpeakONs). There's also the question of stands, as they need to be higher than I've presently got them. So far I like them better than my D2430K/M2 waveguide/2216Nd-1 "hybrids," but on those I still have a ways to go on optimizing the EQ settings, so no final opinions at this point.

Thanks again to all of you who made this project possible! Rgds, Joel.

Those look amazing man! Finding a pair of cabs is cool, but #003 is even cooler. Then again, I'm not sure how many pair JBL made.

Quite surprised at the initial comparison to M2 'clones'... I always assumed there would be a noticed improvement between generations of monitors. I guess eventually once innovative steps become smaller, they're left to working with what they have and making compromises or trade offs in some aspects.

How would you describe the imaging difference between the M2 waveguide and an MTM type arrangement on the DMS-1? Does the bass have a more "effortless" sound due to dual bass drivers?

The main thing is that I don't consider that the crossover and EQ for M2/4367 "hybrids" are fully optimized—yet. The 2216Nd-1 is a significantly different woofer than the 2216Nd used in the real M2, so it's not just a matter of importing the M2 processor settings into the DriveRack and applying them to the speakers I've made. By contrast, I'm using exactly the crossover and EQ that JBL specified for the DMS-1. So, in fairness, I'm not yet comparing apples to apples.

Apart from that, there is, or can be, some advantage to a well-implemented "MTM" design such as the DMS-1. I think of it as creating, or simulating, a point-source design using separate drivers. But maybe that's my illusion influenced by 40 years of playing around with 12" and 15" Tannoys.

And of course, two 14" low-frequency drivers (in this case, covering 30Hz to 1 kHz) don't need to work as hard as a single 15", dual voice coils and whatnot notwithstanding.

Hello Joel

Nice build!

Quote:

---

Originally Posted by dubkarma

The compression drivers are not the 2450SL-A, but the much more common 2450SL, the only difference being, if I understand correctly, that the SL-A diaphragms have both acquaplas treatment and the "diamond" stiffening ribs, while the SL diaphragms lack the latter.

---

I think the first generation of DMS-1 had the smooth SL diaphragm. Don't know if each generation had specific EQ settings though...
Regarding the 1400ND's, as these kits are probably 8 ohms variants you might need to adjust the gain 3dB down there.

Quote:

---

Next I programmed the JBL-published settings for the DSC280 into a dbx DriveRack 4820 that I have used in other speaker projects. The improvement was striking—and not just to my years—and that's what I'm using now.

---

IIRC the DSC280 has its EQ Q values specified as fraction of an octave (bandwidth), so you need to convert that into Q for the DBX.
You also need to make sure the converter you use targets a "constant Q" type of EQ, as used in the DBX, and not a "proportional Q" one.

Another point is the shelving cutoff frequency, which can be quite different in its interpretation from one processor to another (middle, 3dB start, 3dB stop, ...), leading to very different results.

Ultimately, you might want to measure the line output of the DSC and make sure you are getting the same transfer function with the DBX.

Very Nice!

Rob:)

pos,

Many thanks for your comments. I knew the job wasn't quite done!

Once I've got the SMC24 back from the shop, I'll borrow a friend's REW microphone and compare some objective measurements using both the SMC24 and the dbx 4820.

Then it'll be time to go through the points you've made, though, as with other aspects of the project, I'll have to get some help from more technically knowledgeable friends. . .

For now, a quick question about one of those points: what is the rationale of reducing the gain of the LF section by 3 dB?

In any case, many thanks!

— Joel

Quote:

---

Originally Posted by dubkarma

pos,
For now, a quick question about one of those points: what is the rationale of reducing the gain of the LF section by 3 dB?
— Joel

---

My guess is that POS thinks the 1401ND’s are 16ohm run in parallel, which would be 3db less efficient than two 8ohm speakers (1400ND’s) run in parallel. Since the DMS-1’s have a set of binding posts for each speaker, they can be run parallel or each off its own amp channel, depending on your preferences, but make sure you compensate. None of my DMS-1’s are currently running, so it would be tough to test, but I think the only difference between the 1400ND and the 1401ND is the finish on the frames - 1401ND’s have the same, really rough texture in a flat sheen as the cabinets, while the 1400’s have a slightly smoother (but still roughened) texture and also more eggshell like sheen. If anyone really wants to know, I can dig mine out and run impedance sweeps on both, though my 1401’s have completely lost their foam, so it wouldn’t be an apples to apples comparison.

Btw, if anyone has a pair of 1401’s they can spare, let me know - I’m one pair short, and can trade 1400PROs reconed as 1400ND’s.

Quote:

---

Originally Posted by HCSGuy

My guess is that POS thinks the 1401ND’s are 16ohm run in parallel, which would be 3db less efficient than two 8ohm speakers (1400ND’s) run in parallel. Since the DMS-1’s have a set of binding posts for each speaker, they can be run parallel or each off its own amp channel, depending on your preferences, but make sure you compensate.

---

Thanks for your comments! You've got me thinking. First of all, I should admit that I've always assumed that both the 1400ND and the 1401ND are 8-ohm drivers.

One thing I've wondered about is that both the DSC260A (which has a pre-set for the "DMS-1A") and the SMC24 (which is configured exclusively for the DMS-1) have only one pair of outputs for the LF while, as you have pointed out, the speakers themselves each have two pairs of LF inputs.

And, the DMS-1 literature recommends a separate amplifier for each LF input. So how does one accomplish that?

My best guess is that JBL assumed one is using power amplifiers with regular XLR inputs and XLR "pass-through" outputs. So the LF output of the processor (analogue or digital) feeds one power amp and that power amp, with its male XLRs, passes the signal through to the second, presumably identical, amp. In that way each power amp would "see" an eight-ohm load.

But if, as I'm doing, one is driving both woofers in each channel with one amp (I've got jumpers joining the two LF inputs), it will "see" a 4-ohm load. So it would make sense that in such a case, it'd be necessary to reduce the LF gain by 3 dB (maybe a bit more?). Or, perhaps it would amount to the same thing to increase the gain of the HF by 3 dB?

Meanwhile, still waiting for the SMC24 to come back from service. . .

Quote:

---

Originally Posted by dubkarma

Thanks for your comments! You've got me thinking. First of all, I should admit that I've always assumed that both the 1400ND and the 1401ND are 8-ohm drivers.One thing I've wondered about is that both the DSC260A (which has a pre-set for the "DMS-1A") and the SMC24 (which is configured exclusively for the DMS-1) have only one pair of outputs for the LF while, as you have pointed out, the speakers themselves each have two pairs of LF inputs. And, the DMS-1 literature recommends a separate amplifier for each LF input. So how does one accomplish that? My best guess is that JBL assumed one is using power amplifiers with regular XLR inputs and XLR "pass-through" outputs. So the LF output of the processor (analogue or digital) feeds one power amp and that power amp, with its male XLRs, passes the signal through to the second, presumably identical, amp. In that way each power amp would "see" an eight-ohm load. But if, as I'm doing, one is driving both woofers in each channel with one amp (I've got jumpers joining the two LF inputs), it will "see" a 4-ohm load. So it would make sense that in such a case, it'd be necessary to reduce the LF gain by 3 dB (maybe a bit more?). Or, perhaps it would amount to the same thing to increase the gain of the HF by 3 dB?Meanwhile, still waiting for the SMC24 to come back from service. . .

---

Hello, did you manage to compare the SMC24 to the DSC260A and the DBX? Thank you