Seas ER18 design thread

[Z3Sooner]

Sorry to revive an old thread as a first post, but I've been looking at John's website for quite some time and also wondered how the ER18 would work in his waveguide TMM design. The question was asked here, but not fully answered.

I was looking at this project as a first build. I have the knowledge and construction abilities to build speakers, but I've just started studying speaker design and don't have anywhere near the knowledge to know what adjustments would need to be made. John did mention in the post that several other drivers would work well in the design and I thought this might be one considering the better distortion measurements. Any thoughts would be greatly appreciated.

[1Michael]

A different driver usually means a different Xover design is needed.

[Jed]

JonMarsh is working on a waveguide tweeter/ER18 woofer 2 way and you might wait for that.

[JonMarsh]

Yeah, if I make any mistakes, at least you'll know what not to do! :W

Seriously, I've had this idea kicking around since first seeing test results for the ER18RNX and testing it myself- seems like a natural. Now I have some ex-inlaws who could really use this design to go with their new flat panel TV and music system (replacing some JBL's), so it's easy to justify proceeding.

My version is a little bigger than some of the 6-1/2" two ways done- another design to look at if you're interested in this direction is MarkK's Seas DXT design, on his Audio Heuristics site, using the Seas DXT tweeter, which incorporates a smaller basic waveguide, and uses a smaller 0.5 cu ft enclosure, if you're space constrained. The one I'm working on uses a PE 0.75 cu ft enclosure for those not into doing there own finishing.

I expect to have this one I'm working on posted in a few days, at least the prelim results, if not all the gory details all at once! Most of the woodworking is done, and waveguides and drivers ready to go together.

[jkrutke]

A different driver usually means a different Xover design is needed.

Yup, the ER18 doesn't have quite the same top end response as the CA18. It would need a little more massaging into LR2 slopes. As is, it wouldn't really "drop in" to the waveguide design. If you've really got to have a different woofer, the P18RNX/P might have a better chance of working without changes.

[Z3Sooner]

Wow! Thanks for all the responses, and on Christmas no less. I really appreciate it.

I've designed and built a couple of subwoofers with great results, but this will be my first, full range project. I'm trying to decide between two options. My first thought was to just select a proven design and build it. On the other hand, this kind of substitution might be a great learning experiment to "get my feet wet" in speaker design.

I've taken a look at MarkK's design and also the new Idunn using the DXT and the U18RNX/P.

Maybe the better idea would be to build one of these 2ways for a starter and then play around with designing a 2.5 with the same driver? That would give me some time to get a measuring rig set up and learn how to use it.

By the way, thanks also for the patient responses to a complete newb question.

[jkrutke]

I expect to have this one I'm working on posted in a few days, at least the prelim results, if not all the gory details all at once! Most of the woodworking is done, and waveguides and drivers ready to go together.

Sounds like a cool combo, keeping an eye out for your design.

Maybe the better idea would be to build one of these 2ways for a starter and then play around with designing a 2.5 with the same driver? That would give me some time to get a measuring rig set up and learn how to use it.

By the way, thanks also for the patient responses to a complete newb question.

I bet Mark K could easily come up with a 2.5 way of his DXT design using his current measured data set.

[Mark K]

I think there is a lot to be said for a slightly larger cabinet although the more than usual stuffing via the Martin King mathcad worksheets works well. Also, a slightly bigger waveguide might even work better.

I'm not reinclined to try to do a 2.5 way. If you really want that, build Zaph's waveguide. If you really want to pursue a TMM, well, start with SE... :

At this point there will be a number of really nice designs for the ER18-Jon's, Zaphs, and mine. Not counting the Idunn.

I mentioned it in a post on the PE board, but I'm much more inclined to morph this into a DXT/ER18 or other mid all passively crossed on a rather wide baffle unit, and having that actively high passed to a woofer module.

Think a passive MT speaker like this



Actively crossed to a woofer module.

See SL's comments on the Grimm here

http://www.linkwitzlab.com/links.htm#A sensibly designed loudspeaker

PS merry xmas!!!!

[Jonasz]

Another interesting thing about the Grimm is the low crossoverpoint at 1.2kHz (4th order acoustic) and is still capable of a max SPL of 109dB/1m.

[Z3Sooner]

[QUOTE=Mark K]

I'm not reinclined to try to do a 2.5 way. If you really want that, build Zaph's waveguide. If you really want to pursue a TMM, well, start with SE... :

[QUOTE]

No problem, If I do something like that It would be for a learning experience anyway.

Looking at the Grimm, what's the reasoning for inverting the mid and tweeter?

[JonMarsh]

I don't know the specific designer's reason, but one reason is that with the conventional arrangement, going above the tweeter axis produces a null in the crossover frequency range, whereas going below the tweeter axis produces a broad but gentle lift -- this is far more subjectively benign, so if you sit and stand listening to your speakers, having the woofer above will often be preferable. Has implications for a TMM 2.5 way system, too, as it's better to configure it more like an MTM in many cases.

[Paul Ebert]

As in a MTM 2.5 with the .5 woofer above the tweeter?

[Carl V]

Can't speak for Jon...but no.

I did it with the .5 below. The midwoofer is above the tweeter, the tuncated
FR woofer was below. This also allows for a less compromised horizontal
Center speaker if you have to go horizontal Sonically...romance is another matter :T

[JonMarsh]

0.5 below, midwoofer above the tweeter would be the way to go for the reasons I describe.

ALSO, if one is using a 2nd order crossover without any physical offset correction or all pass time delay network, then the crossover lobing will be directed upwards if the midwoofer is above, which is generally the superior solution, as we listen more often standing at a distance than lying on the floor (excepting ThomasW).

[AlanH]

ALSO, if one is using a 2nd order crossover without any physical offset correction or all pass time delay network, then the crossover lobing will be directed upwards if the midwoofer is above, which is generally the superior solution, as we listen more often standing at a distance than lying on the floor (excepting ThomasW).

Interesting. Does this apply to TMs or just to the MTM 2.5-way you were describing? If it applies to the former, does that mean for a speaker like Mark K's design a builder would be better off placing the speaker 'upside down' if the primary listening positions on the tweeter axis or above? And how significantly does it change as the crossover slopes increase?

[JonMarsh]

Interesting. Does this apply to TMs or just to the MTM 2.5-way you were describing? If it applies to the former, does that mean for a speaker like Mark K's design a builder would be better off placing the speaker 'upside down' if the primary listening positions on the tweeter axis or above? And how significantly does it change as the crossover slopes increase?

The degree of this effect varies with the tweeter and baffle design.

In the case of Mark K's ER18DXT, the waveguide mounting of the tweeter offsets it towards the rear and helps line up the acoustic centers. Same thing for Zaph's Waveguide TMM.

Now, for a design like the SR71, using asymmetric slopes and phase characteristics for what is basically a 4th order, the crossover region overlap is only about half as wide, and you can get away with phase tricks in the crossover, and get a horizontal tweeter/woofer lobe. For a second order crossover, the wider overlap complicates things, and it's difficult to compensate in the crossover except by an all pass delay network. (like Zaph's ZD5).

OTOH, when you're using a waveguide, or a tweeter with waveguide front end, you may get more matching of the acoustic centers and an easier crossover design.

This same concept would apply to a 2.5 way TMM.

It's funny this topic has come up this way, because the Modula MT MkII design I'm working on will have a few build options, modular if you will, as the original Modula's I built in the 90's were. The configurations in development are a TM with waveguide, a TM without waveguide (simpler build), and a TMM with waveguide, using two 0.75 cu ft PE cabs per side, for which the configuration could be inverted or not, depending on the preferred placement height of the overall system.

For the TM without waveguide, the recommended configuration is midwoofer above tweeter, mounted a little higher than a standard stand speaker, as usual, tweeter preferred at seated ear level. This way we don't have to jigger the crossover so much or add a delay network. I wasn't really planning on doing the more basic TM, but playing around a bit today, it was easy to come up with the crossover, and it is a simpler build.

[Bear]

I think there is a lot to be said for a slightly larger cabinet although the more than usual stuffing via the Martin King mathcad worksheets works well. Also, a slightly bigger waveguide might even work better.

I'm not reinclined to try to do a 2.5 way. If you really want that, build Zaph's waveguide. If you really want to pursue a TMM, well, start with SE... :

At this point there will be a number of really nice designs for the ER18-Jon's, Zaphs, and mine. Not counting the Idunn.

I mentioned it in a post on the PE board, but I'm much more inclined to morph this into a DXT/ER18 or other mid all passively crossed on a rather wide baffle unit, and having that actively high passed to a woofer module.

Think a passive MT speaker like this



Actively crossed to a woofer module.

See SL's comments on the Grimm here

http://www.linkwitzlab.com/links.htm#A sensibly designed loudspeaker

PS merry xmas!!!!

Pretty. :drool: Do you have to worry about reflections from the backwave in a cabinet that thin?

[benchtester]

In the case of Mark K's ER18DXT, the waveguide mounting of the tweeter offsets it towards the rear and helps line up the acoustic centers. Same thing for Zaph's Waveguide TMM.

Yes, but... I was disappointed to find that the DXT waveguide only provides about 8 mm of recess. (I am out of town, and can't verify this dimension right now.) In comparison, Zaph's TMM waveguide provides 29 mm of recess. For my speakers I was looking for 20 mm, hence the disappointment.

However, I like the Seas DXT in general, its sound quality is clear, accurate, and thanks to the DXT waveguide it has good dispersion. Mark K's ER18DXT looks like a great implementation.

I have a modest little blog entry on a DXT mini-project here:

http://techtalk.parts-express.com/blog.php?b=68

and a more successful ScanSpeak waveguide project blog entry here:

http://techtalk.parts-express.com/blog.php?u=4397&blogcategoryid=8

(feel free to leave comments)

[Mark K]

Pretty. :drool: Do you have to worry about reflections from the backwave in a cabinet that thin?

Yes, I think you're stuck with at least one dip where 2xDepth is equal to 1/2wavelength. Being shallow makes the dip probably deeper and wider. The flip side is that it also forces the first dip at 1/2lambda high enough that the subsequent dips at 3/2, 5/2 are above the mid passband.

Say the internal depth of the Grimm is 12cm. The first reflection induced dip would be around 700 hz. There would be a peak, correctable, at 1400 and the next dip at 2100, well outside the passband.

In a typical box, the internal depth might be double that, 24cm. The magnitude of the dip will be down, expecially if stuffed. But, the first reflection will now be at 350 hz, and there will be a peak at 700, another dip at 1100, and a peak at 1450, etc, etc.

So, which is worse?

[Mark K]

Yes, but... I was disappointed to find that the DXT waveguide only provides about 8 mm of recess. (I am out of town, and can't verify this dimension right now.) In comparison, Zaph's TMM waveguide provides 29 mm of recess. For my speakers I was looking for 20 mm, hence the disappointment.

However, I like the Seas DXT in general, its sound quality is clear, accurate, and thanks to the DXT waveguide it has good dispersion. Mark K's ER18DXT looks like a great implementation.

I have a modest little blog entry on a DXT mini-project here:

http://techtalk.parts-express.com/blog.php?b=68

and a more successful ScanSpeak waveguide project blog entry here:

http://techtalk.parts-express.com/blog.php?u=4397&blogcategoryid=8

(feel free to leave comments)

The depth from the DXT flange(i.e. baffle face) to the base of the grill is around 12mm. The dome is behind the grill a mm or two, probably. So there is probably 14-15mm of rear offset of the dome (as a quick rough estimate-don't hold me to it...). While it would be nice to have, say, 25mm or so, this is pretty good all things considered.

I too, would like to have a somewhat bigger waveguide. IIRC, wasn't Seas prototyping a larger DXT lens? That would be cool. Something like the RS28 motor on a 6-8 flange...oh, where are the cheap asian DXT knockoffs?

[dlr]

Something like the RS28 motor on a 6-8 flange...oh, where are the cheap asian DXT knockoffs?

Waiting on the patents to expire probably. I would guess that it will be a while.

DXT Home Page

Dave

[Mark K]

Waiting on the patents to expire probably. I would guess that it will be a while.

DXT Home Page

Dave

Oh, but then why is it they sell all though new release DVD's for only a buck in the far east... 8O

edit-I may try to contact DXT and see if it's possible to generate some sort of a group buy on a larger version. Not at all sure they would be interested, but it's worth an email.

[dlr]

Oh, but then why is it they sell all though new release DVD's for only a buck in the far east... 8O

edit-I may try to contact DXT and see if it's possible to generate some sort of a group buy on a larger version. Not at all sure they would be interested, but it's worth an email.

I'm sure that they could copy the DXT and sell them in the far east, it just wouldn't be sold here. There's likely not much of a market for the DXT there. Now if the market for it there was good enough...

I'd go in on a group buy for one that could go lower into the midrange, maybe a 6" diameter version. Maybe even use the same voice coil diameter. It might require bumping up the diameter a couple of mm to get enough volume displacement without excursion increase to benefit from it. The DXT is 26mm, but SS has used 28mm voice coils effectively for a long time. The trouble with that is re-tooling for an entirely new motor, no parts sharing as is now the case.

Dave

[Jonasz]

Would be interesting with a larger one 6-8" that fit the RS28F/A, Peerless HDS/SS D2608 or similar tweeter with gutsy low end. I would be interested in a group buy for sure.

Btw, did anyone test this new(?) SS tweeter yet? Smooth response and a very low resonanse at 475 Hz.

http://www.madisound.com/catalog/product_info.php?products_id=8724

[JonMarsh]

Well, it's described as a Danish built version of the DX25T, so I'd expect it to test similarly to the original Vifa version- the HDS tweeters in the Discovery series are essentially identical. If that pattern holds true, then I would expect the performance to be similar, but then, it does look at bit different physically from the original. If it's similar to the original DX25, then the D2608-9130 would be preferred, for the $15 or so cost difference.