SEAS 4-way Dipole

Return to the Archives page.

Update 2-27-07:

A number of people have requested detailed plans for the Seas 4-way but unfortunately this was one of those projects that was drawn for the most part only in my head and then just built. Also, while I believe the addition of the 4.5" Seas driver put this speaker a step above the Orion, the system does have a number of problems...

The magnet mounting system for the 8" drivers works pretty well but the rims floating inside rubber gaskets in the front panel sometimes rub due to excessive energy transfer from the 12" Dayton Reference drivers in the bottom. It's not excessive, but occasionally you will hear it especially on any material with sustained low frequency energy.

So, for my next system, I will do two things to solve the panel vibration problem. One, I simply will not put the heavy high excursion drivers in the main panel. These will be in independent baffles thus completely isolating them from the rest of the speaker. Two, the mid-bass drive units that will be in the bottom of the system will be using an opposed mounting system to cancel any energy these might transfer to the baffle. These two fixes should solve the problem.

After having the Seas 4-way for more than a year, I have concluded that a 3-way system simply is not adequate to cover the full bandwidth. I had for some time believed the addition of a small driver between the 8" and the tweeter would offer some improvement and this system has been my opportunity to explore this. I now believe that the improvement has been greater than my original expectations. I further believe that this concept should be taken farther. I am currently working on a six-way system that will include a 5-way panel plus separate dipole subwoofers.

So, sorry to those who might have an interest in building the 4-way Seas system, but I really never drew anything more than a few sketches to get me started. After that I pretty much made it up as I went along. It has served it's purpose and proven to me that 3-way is not adequate so from here I plan to move forward in more ways than three...

Last Update 1/14/06:

Continuing to listen... I have the spikes on the bottoms now and the crossovers have the covers on and are installed in the rack. I don't want to be so bold as to say I have made a speaker better than Siegfried's Orion but these will be replacing the Orion for my critical listening system. I'm not sure what I will do with the Orion, I might sell them or I might set them up in listening room #3.

I'm still looking into other amplification options for the new speakers. The Ushers are great but with four or five of them running it just pulls the AC line voltage down too much. I'm going to try an Anthem Statement P2 for the woofers because its idle current is only 140 watts compared with the R-2.5's 400 watts. The 400 watts by itself wouldn't be a big deal, it's when you add in the 175 watts each of four R-1.5's that I have a problem. That's a total of 1,100 watts being sucked off the AC line at idle!!! It's pulling the voltage down!

Update 1/13/06:

I spent some time listening last night. First impressions are very good! Is it better than the Orion? A very definite...maybe. I'm going to listen more when I'm not quite as tired as I was last night. The Orion uses the Seas 8" to 1400 Hz which is close to the 1600 Hz where stored energy problems begin. Also, the Orion pushes the tweeter to 1400 Hz. This system uses the 4.5" Seas driver to alleviate both the 8" and the tweeter from potential problem areas. I ended up going with 700 Hz to 2,100 Hz for the 4.5" driver and from listening thus far I believe this is an improvement over the Orion. The sound stage is very tall making it something like the performers on a somewhat elevated stage. The effect is not as much as you would expect however from the location of the tweeter. The polar pattern had been directed downward slightly using the all pass delays in the crossover.

Another advantage over the Orion is the high pass filter at the low end. Especially anyone using them with vinyl knows about the bouncing cone syndrome of the XLS drivers in the Orion. I'm filtering the woofers 4th order LR slightly below 20 Hz in the stand alone mode and selectable 25 Hz or 40 Hz in subwoofer mode. I don't think there will be any reason to ever use the 40 Hz mode.

There are a few cleanup items remaining like putting the tops on the crossover cases and mounting the spikes on the bottoms of the speakers.

More photos...

The additional unused output you see on the crossover is for the optional subwoofer. I have started cutting parts for twin subwoofers each using the same Reference Series 12" Hi-Fi woofers as in the main panel. They will however be in large sealed boxes driven by plate amplifiers and will cover the 16 Hz to 25 Hz range.

Update 1/9/06:

I'm working on wiring up the drivers today. The thing kind of dwarfs the Orion next to it doesn't it. These will be up and playing shortly and then I can start working on the matching subwoofers. I'm setting the crossover at 25 Hz between these and the subs. The subs will then cover the 16 Hz to 25 Hz range.

Update 1/7/06:

Almost finished! Should be loading drivers tonight or tomorrow. Hopefully they will be playing in the next few days. What? Yes, those holes do look too big for the 8" Seas driver don't they! That is because they are 1/2" diameter larger than the outside diameter of the Seas drivers. They won't even touch it! They will instead be mounted to the large vertical post in the rear only by the magnets. This will completely decouple them from the panel.

As you can see in the photo, it's a bit larger than the Orion. The tweeter and 4.5" driver will be well above the ear level while seated. This is corrected for in the crossover by using shorter than otherwise expected delay on the tweeter and 4.5". This will direct the polar pattern downward slightly toward a point just above ear level while seated.

The 12" RS HI-FI Dayton Audio drivers will have just a little more output in the low end than the 10" XLS drivers in the Orion. Not much more however because keep in mind that in the case of a dipole the excursion requirement increased as a cube function as you move down in frequency. So, the low end capability will be only slightly more than the Orion. These however will be used in conjunction with twin subwoofers which will take over at 25 Hz and extend down to about -6db at 16 Hz. The subs will use the same Dayton RS Hi-Fi drivers (two/side) but in sealed enclosures.

Below is a photo from earlier during the measurement phase...

The earlier stages of this project as they progressed are below...

(date...don't remember) I have worked up a 4-way +1 active crossover and I'm working on laying out the pcb. Below is a block diagram...

The crossover will be intended for use on up to a 4-way di-pole main panel with an optional sealed box sub-woofer output. The sub out will have no sub-woofer EQ, this will be expected to be included with what ever sub-woofer is used. An external (or internal jumper) switch can select the, "stand alone" mode which will place the HP rumble filter at perhaps 10 Hz. The rumble filter will be optionally up to 4-th order. In the sub-woofer mode the sub-woofer crossover will be selectable between two choices using jumpers. The frequency choices are determined by fixed resistors to be chosen prior to construction of the board. You could for example insert resistors for a 30 Hz and a 40 Hz choice that can later be selected by internal jumpers.

I did not include di-pole correction to the upper midrange circuit due to the expected frequency range. I expect this output to typically be used for about 600 Hz or higher possibly to a dome midrange or possibly a small cone driver. Although it might still be dipole it is likely the dipole correction requirement would be sufficiently below the pass band that it would not be required. It is also not likely to need Q correction on this driver since it will likely be operating well above its resonance frequency.

Recently I have been hearing some argument that a slight low Q boost in the lower midrange may produce a favorable effect. Apparently it gives vocals more body or something. Anyway, this crossover will have the ability to provide that boost with the, "Mid EQ" section.

As of today, 9/26/05, I am in the process of laying out the pcb. With 23 op-amps plus well over 100 resistors, and over 100 caps for a single channel, it's going to take awhile! I'm not sure when this pcb will be completed but I may make it available once it's done to offset the cost of making the boards. It cost a lot to make boards in small quantities so I will probably do several sets on the first run.

Update 10/16/05: The board layout is finished and the boards have been ordered. They should arrive by mid week.

Update 11-8-05: The boards are here and the project is underway. I'll be doing measurements today for entering into LspCAD. The plan is to get the crossover designed before I leave on the next trip to Taiwan around 11/15/05 so I can build up the boards while there. I don't expect to completely finish the cabinets on this trip, too many other things going on. Hopefully on my next trip home...probably in Jan 06' I can get this system done. It btw is using the same Seas 8" driver as the Orion and same tweeter. It is using two of the 8" drivers instead of one however and it is using the 4.5" version of the same driver to cover the range from about 700 Hz to 2100 Hz. This keeps the 8" farther from its problem area that starts about 1600 Hz. (It has stored energy issues above this frequency) The 4.5" driver also will take a load off the tweeter allowing it to crossover at 2100 Hz instead of 1400 Hz thus cutting the excursion of the dome about in half.

I already know of one minor mistake on the boards...holes too small in the 3-pin jumper package. Once a set is built up and working and verified to be ok, I will be willing to sell a set to anyone wanting one. My cost was just over $50/board so $100 for a set. (It's a mono board). On this first run I will only have at most 3 sets available for sale. After that if more people are interested I will correct all known errors at the time and make more but might up the price to $75/board for the effort.

Update: 12/8/05 Hey what do you know, the board actually works! I have a full blown crossover up and running on it and I didn't even have to cut any traces and put parts on the back of the board! I would make minor changes in any future builds but the biggest would just be to correct the hole size in the jumper pins. I found pin strips here in Taiwan that fit the existing holes however so it's not that big a deal. I would also adjust cap packages somewhat in some of the notch filter sections and possibly change the topography of the room mode notch but none of these are big issues.

1/4/06: I'm getting closer to completion. I should have photos of the speakers posted in a couple days. My ideas on powering these has changed. Although I bought one more Usher R-1.5 and an R-2.5 while in Taiwan I will not be using them. I'm going to opt instead for smaller packaging in the form of multi-channel amplifiers. I haven't decided exactly what yet but just something more compact and that uses less power at idle. So, the Ushers are for sale, the R-2.5 is already on Ebay, Item number: 5851140061 ... bid high and bid often!!! The thing is huge! About 110 lbs of pure class A power...250 w/ch into 8-ohms, and even 1,100 watts into 1-ohm single channel driven. If you don't need it for your speakers put it in the garage and use it for an arc welder. It really is a nice amp but for this speaker I would need it plus four R-1.5's which totals about 500 lbs, and sucks 1000+ watts of the AC line. It's also massive overkill so I'm thinking maybe just use two 6-ch amplifiers or more likely something like three 4-ch. That way I could put one 4-ch driving the woofers and let the other two drive the rest of the speaker. This way the woofer load would be off the power supplies of the other amplifiers. I don't know, I'll figure it out.