RS390/ER18RNX/27TDFC dipole

[Mazeroth]

Now that we've moved into our new house (heated, extra large garage!) there's no reason anymore for me not to start building some delicious speakers. Living in the townhouse left me with little room to work and most of my ideas went away after months of not doing anything with them. Now that we're settled in I want to build a pair of full-range speakers for my listening room that will give me the best bang for the buck possible. I'm planning on building a pair of dipoles very similar to the Isiris (physically), only less expensive. What I'm thinking of using is:

(1) Dayton RSS390HF-4
(1) Seas ER18RNX
(1) Seas 27TDFC loaded into an MCM waveguide

I would like to use a DCX2496 to handle the crossover duties which will probably also mitigate the overall cost of this project (I already have the amplification). I'm guessing I'll begin construction of this speaker sometime in November and am still tossing around which mid/tweeter option to go with. For the price and distortion profile I don't think I can do much better than the RS390, so that's pretty much set in stone. I'm really liking the ER18RNX after Zaph's recent evaluation and I can't see why I shouldn't use a waveguide in a design like this.

Right now the overall cost is hovering around $800 for all the parts (I already have the DCX2496 so for me it will only be ~$550). I'm not sure how much more performance I can obtain spending another $100-$200 and if it will even be worth it?

As always, any input is greatly appreciated!

[Paul Ebert]

The only things I might consider are doubling up on the ER18s (MTM) and buying the drivers soon to try to beat the price hike. But, MTM might not work too well with the waveguide.

Sounds like a very nice system. I wish I had the space for dipoles.

[ThomasW]

We've posted about the waveguide before. Jon doesn't think it works well at the low end of the passband. Zaph thinks it's okay for the tweeters he's used.

That said...

Since after the crossover losses are included the tweeter is much more efficient than the single ER18, why bother with the the hassles of using the waveguide?

I agree with Paul, for a larger room and HT output levels, adding a second mid-woofer to create an MTM is a good suggestion.

FWIW there will be a lower cost version of the Isiris, it will still use the BMS/ENG setup (if product is available) so no cost savings there.

Having lived with both the Isiris and the Arvo, my vote for HT use is the Arvo. Yes you could use a single RS390 instead of the 2-12"s. Chasw98 will be starting a thread pretty soon about building his Arvos, so you'll be able to get some info from it.

[chasw98]

You got my attention!
I went with dual 12's RSS315HF for the low end and am toying with a 15" sub for very lo freq. The dual mids everyone is telling you to use is good advice. The way mine is set up, I am running 2 8" drivers in series. A single driver would not be enough, but 2 of them gives you breathing room (easier to take away than to add). I have a very lossy room with carpet and heavy drapes so I am finding that the top end needs some extra boost. I am using the DCX2496 for active lo and to divide to the mid/hi passive crossover. I have just finished ordering a bunch of speed parts for the passive crossover. I am going try and reverse engineer the passive crossover to my needs in my room, kind of putting the 'Y' in DIY as JM would say. So far, though, these are very clear and open speakers with very good imaging and a wide soundstage.

[CraigJ]

Hi Mazeroth,

My input would be to spend the extra money and build the Isiris Junior with the DDS ENG-1 BMS 4540 combination. I have built at least five versions of the Isiris; RS28a, Peerless hds810921, with and without MCM waveguides, BMS 4540, single and dual Peerless 830883 , dual RS12s. Come to think of it, I've even made five different baffles To my ears, the design that Thomas is using integrates the drivers better than the prior designs, allowing the speakers to disappear. Yes, I think they sound the "best".

You are more than welcome to borrow my MCM waveguides if you'd like to go that route. Certainly the dcx makes experimenting easier- even I can do it.

Craig

[Deward Hastings]

sounds great . . .

The RS390 should work fine . . . it has about the same displacement as the two 10” in ORION. Larger flat baffle vs. smaller H-frame is entirely a matter of acceptable size in the room, they should sound the same. The 390 may offer a slight advantage, in that it can probably be crossed a bit higher (no H-frame resonance to avoid), reducing demands on the mid driver. That in turn allows a smaller mid than ORION’s W22 . . . the ER18 should do fine crossed maybe 150-175 Hz. It also (perhaps) allows moving the M/T crossover up a bit from ORION’s 1440 Hz, which increases the number of suitable tweeters. It is an entirely plausible and practical design approach for a full range dipole. Don’t forget to at least allow for the possibility of a rear-firing tweeter to make the radiation pattern more nearly uniform full range.

As you no doubt already know Linkwitz discusses all of this in detail at the linkwitzlab web site, including detailed discussion of M/T panels with passive crossovers (in the “Phoenix” section under “prototypes”). Doing the crossover design with a DCX is an excellent, and by far the easiest, way to get it all sorted out, especially since you already have a general idea of what the crossover response will need to be. If you are at all handy with electronics you might then consider buying ORION crossover boards (they’re quite inexpensive considering what you get) and populating them with components selected for the desired response as determined with the DCX. Or, of course, you could just use the DCX as a permanent part of the speaker.

Three particulars worth serious consideration are magnet mounting the mid driver, those rear-firing tweeters, and correcting for floor effects. It sounds like a great Winter project, and using the DCX in the design process makes experimenting easy . . .

[Paul Ebert]

Could you please elaborate on magnet mounting? I have a vague idea what's meant by this, but I'd like a better one. What, exactly, does this entail and what are the pros / cons?

Thanks.

[ThomasW]

Linkwitz decouples the driver's frame from the front baffle. To hold the driver in place he has a mounting system that attaches to the magnet.





There's info about this on his website

[Deward Hastings]

> please elaborate on magnet mounting?

What Thomas linked . . .

See also:

http://linkwitzlab.com/Driver%20Decoupling.doc

for results from another source.

[Mazeroth]

Since after the crossover losses are included the tweeter is much more efficient than the single ER18, why bother with the the hassles of using the waveguide?

The main draw with the waveguide is to improve the tweeter dispersion and minimize room reflections, with some emphasis on minimizing distortion in the lower frequency range. I'll also be using an active crossover so the efficiency thing isn't that important. I will need to do some testing to see if the output boost between 1-6 khz is going to make integrating with the mid more difficult. However, with the flexibility the DCX2496 offers I'm hoping it won't be too bad.

As far as going with dual mids, this is going to be used only for music duties with a seating distance of 7-8 feet. I'll be crossing the mid over around 150-200 hz. so I'm hoping the single mid will be able to keep up.

Now, don't get me wrong, I would love to build a pair of the Isiris Jr. but the cost difference is going to be astronomical (I just closed on a house!). If my calculations are correct the drivers alone are going to be almost $1100, not to mention the cost of the crossovers. Maybe in a few years I could swing that by the wife but definitely not this early in the game.

So, with that said, do you guys still think the waveguide is a bad idea? Can you think of any other mids I should consider for this build?

Thanks.

[Paul Ebert]

If I were you I'd give the waveguide a try. Assuming you go with the MCM, you can always remove it and not be out much $.

As for other mids, I'd strongly consider the Usher 8945P which is, at the moment, only $15 more than the Seas at PE. Personally, I don't like the looks of the Usher, but it's hard to argue with the distortion levels.

[Deward Hastings]

> do you guys still think the waveguide is a bad idea?

Maybe not a bad idea, but I don’t see it “all worked out” yet, so it’s maybe not without risks. That, combined with the disadvantages of greater center-to-center spacing (to the mid) might make the negatives greater than the positives, especially for a “budget” project. There are a lot of decent sounding speakers out there *without* waveguides . . . not so many with . . .

[ThomasW]

The MCM waveguide brings little to the table.

The BMS 4540ND/DDS ENG-1 combo is a different story.

[JimS]

Having lived with both the Isiris and the Arvo, my vote for HT use is the Arvo. Yes you could use a single RS390 instead of the 2-12"s. Chasw98 will be starting a thread pretty soon about building his Arvos, so you'll be able to get some info from it.

Got my Arvos up and running last weekend - haven't had enough time with music to give a detailed impression yet (kids' rooms are upstairs so I've been limited to low-level listening at night), however, for HT they are awesome :T :T :T :T (my son "made me" watch a few of the Star Wars series last weekend to make sure he gets his costume right for Halloween!!)

Very detailed, dynamic, great reproduction of low level effects, dialog very clear and imaging good enough that it'll be a long time (if ever) before I feel the need to add a center for HT . . .