Well, I have taken some steps and now I need some help...please :)

For the last speakers you will ever build then you need to make the most of it as it is the midrange that really matters for voice That said I have a couple of suggestions for you that are near to your budget:

Zaph Audio mid/woofer

and the Scan Speak Discovery 10F/4424G and the T/S parameters and Freq plot

I see that you are aiming for about 8 Ohm impedance overall (based on the ribbon tweeters) so a pair of the SS 10F in series will get you there. 1 Litre internal volume sealed for each. For the low XO try around 300 (or the 250Hz you suggest) and for the high try around 3000 or better. John K raves about this driver and he takes them quite high. Of course, yours is a different application. You need to think about off axis dispersion and power response, especially matching up to a ribbon tweeter. I'm sure the Solen guys can optimise this for you and provide details of the volume for your bass drivers in a sealed box. I did a sim in Unibox and it needed 256L for 4, Qtc of 1 (and it wouldn't go much lower than this). It's flat to about 50Hz. Suggesting dual opposed top and bottom for the bass drivers (WW) to minimize vibration. And I think that WMTMW would be enough for the center channel.

Some are of the opinion you choose your midrange first and work from there. So many ways to approach this

note really possible with 4 4 ohm woofers. This will be a 4 ohm speaker, not that that has any bearing on what mid is chosen. Impedence does not need to be matched; sensitivity does unless you're going active (which you should based on the insanity of this project.... )

Out of the OPs choices the only ones that might work are the first TBs. The others arent sensitive enough, even taking into consideration a parallel configuration. The SS 10Fs might work as well, but over budget.

Well done, WMTMW's (or a 3.5 WWMTMWW) can produce excellent results...and I agree with the choice of a ribbon tweeter for that configuration.

With that much money in tweeters, I agree you should not cheap out on the midrange. Since this will be your "last" set of speakers, you will always regret not stepping up to better midrange drivers. The Scan 10F is also the first driver that came to my mind as a good match with the ribbon tweeter.

A potential problem with the 10F is they may not play loud enough, low enough, to mate well with inexpensive bass drivers. I have no knowledge of the particular bass driver you linked but, typically, the Peerless SLS series is good significantly higher than 250Hz. (The 12" SLS in my surrounds are crossed around 300Hz.) As a bit of insurance or "future proofing", you could size the enclosures and cutouts to facilitate both the bass drivers you have as well as something better later on.

Since the center is arguably the most important of the LCRs, I would make it at least as capable as the L&R...possibly doubling 10F's in vertical pairs on either side of a vertical ribbon to lower distortion and achieve a similar vertical pattern. Maintain a woofer complement at least as capable as the other two front speakers.

Seems like you are going to need a professional to take measurements and design proper crossovers for your system. Without optimal XO's, the drivers are a waste of money.

Paul, nice call on doubling up the mids. I only suggest less bass drivers as this is going to be a big speaker if my T/S values are spot on. Have to second the need for measurements, including T/S for the bass; this is not a simple XO to design.

Noxdowne, another way to look at the budget for a driver is how well behaved it is. The 10Fs look relatively easy to manage and wont need a lot of components in the XO. The TBs, as an example, will need to be tamed above 2K and the cost of components can add up.

You got a great deal on those tweets...thought you paid more for them!

Anyhow, I would not double the mids in the L&R speakers...might mess up what could otherwise be tidy polar response. Instead, I would return those 8" buy-out woofers and exchange them for a simple pair of 10" with better HF performance...something like the regular SLS or RS270 with aluminum or copper in the motor.

Both the 5" SS and ZA14 are too large for a good horizontal polar match with the narrow ribbon and both probably run out of displacement at 250Hz...but you don't have much choice with the "iffy" 8" woofers in a 3-way. The 5" SS has no advantage in displacement or directivity and runs out of "smooth" up high, so the 14 may be the better bet.

The reason I initially recommended doubling mids was the extremely small displacement, and diameter, of the 10F...and I didn't realize you had height flexibility with the center. Height opens the possibility of duplicating the vertical MTM in the center channel so that horizontal and vertical polar patterns match the L&R. (If you're using the center as a TV stand, don't forget to tilt the baffle up toward the LP...ribbon demands it.)

Woofers without a top end, and delicate tweeters with wide dispersion screams "4-way". However, to keep you out of the really deep end of the crossover pool, I'd opt for pairs of decent 10's, cross to Zaph mids at 400-500, tweeter at 3-3.5k. WMTMW for the L&R, W-MTM-W for the center.

Because of the very different baffle size and shape, make sure your crossover designer uses measurements from your drivers in both cabinet styles and you should be good to go.

Don't understand what this means???


Matching the mids to the tweeter:
It is better to match directivity whenever practical...especially the horizontal because the ear is more sensitive in the horizontal. A narrow ribbon tweeter exhibits very broad horizontal directivity but, because it is tall with respect to wavelength, a very narrow vertical pattern. A ribbon works well in an MTM because the M/M pattern, like the ribbon, is wide but compact vertically (fewer floor & ceiling reflections). However, if the ribbon and cone mids are substantially different widths, the horizontal pattern will narrow substantially, just below crossover. A little 2" mid would help minimize horizontal pattern mismatch with your ribbon...but it obviously wouldn't have enough LF capability...so no 2" mid. The ZA14 is marginally smaller than the SS and seems slightly more extended on the top end. Very minor edge to the ZA14.

Matching mids to the bass drivers:
For a given SPL, cone excursion increases with decreasing frequency. (Watch the cones in your subs for an illustration.) So the maximum LF SPL a driver can deliver is normally determined by how much air it can move while providing reasonably linear output. The SS mid has 80 sq cm of cone area and +/-1.5mm xmax (~linear movement). The ZA14 has only 68 sq cm of cone area but +/-3.25mm xmax. Using SL's SPLmax spreadsheet, the ZA14 can move enough air to achieve an extra 5db LF output over the SS! You need all the LF output/headroom you can out of the mids and that's the primary reason I'd pick the ZA14 over the Scan.

Didn't know you were using a projector screen...so much for assumptions. Have you considered an AT screen? Only $200 or so from Seymour AV and it would provide a far more convincing soundstage...you'd be surprised.

The boxes look good..be sure to design/angle the baffle planes so the center of each MTM is on-axis with your ears. They are going to keep your crossover designer busy...three different baffles and three slightly different crossovers!