3-way reference project??

Just another Moderator
Joined 2003
Paid Member
I think that sounds like the way to go moondog. I don't have the monetary constraint, but I already have too many projects, that are not getting done so adding another is not an option (though I would love to do it).

Moving to a 12" might start to blow out the cost somewhat. SB have some nice drivers out now (and they are widely available) and I'd consider them for the woofer and mid, definitely and quite probably the tweeter as well.

Tony.
 
If you want something like this to work I suspect you need to agree on the objectives first before considering suitable drivers.

Most of the projects discussed here are usually skewed by enthusiasm in one or two directions away from what an impassive balancing of design compromises would produce. Perhaps a good role for a reference design is to set a few bounds on size and cost, set a target of reproducing stereo music at, say, reference SPL in a medium sized room and then to document the design considerations that lead to a 1" tweeter + 4"/5" midrange + 10"/12" woofer being the conventional reference design. I would also suggest picking, say, 2 or 3 alternative drivers from the standard range of established driver manufactures rather than a single "winner".
 
diyAudio Moderator
Joined 2008
Paid Member
This is reasonable, I think defining the objectives makes compromises much easier. DIYers produce good speakers that can become larger, more expensive, or more over-specced than something that might be capable of producing similar results.

It's only been a few years since this began but waveguides have become more of interest.
 
Well if defining the objective is one way of looking at the problem and the 2-way was an effective answer perhaps all we need to add is the missing bottom octaves.
Getting real bass but not subsonics? 30Hz to 20kHz 3dB down at each end??
It is not so much looking at the problem as defining the problem in quantitative terms. In qualitative terms a reference monitor may mean low distortion, loud, full audible spectrum, etc... but this is insufficient to make decisions about what hardware qualifies and what does not.

So how loud? This should follow from picking a representative room size and the required SPL level. The minimum to qualify for reference would have to be at least the standard playback level of around 85dB average + 20dB peaks. There is a case for more to match, say, concert levels.

Frequency range? Flat from 15Hz-25kHz is ideal but needs qualifying to be useful. 30-20kHz could be argued. A +/-3dB tolerance would not qualify for a reference label for many. A +/-1dB tolerance is getting more like it even though a 2dB difference in level is going to be audible.

Reference distortion levels? Tricky. Needs to be met at the maximum required SPL. Perhaps something like less than X% THD at maximum SPL. Possibly a larger value for less than 100Hz. Tying it more closely to perceived distortion is going to have practical difficulties.

Directivity? The ideal directivity does not exist and so a target cannot be defined. Nonetheless we know that sudden changes with frequency are to be avoided in a reference designs.
 
diyAudio Member
Joined 2007
Or... TIW250XS or TIW300...(think i got the models right)

I then found out that the Visaton drivers are not as available in the USA as the Peerless. Getting commonality was always part of the problem in deciding on a "Reference" I think.
In my own build of the Reference 2-Way I stopped early and added a .5 woofer turning the project into a 2.5 Way speaker for SWMBO [almost finished after 4 years ] that could also be termed part of this exercise.

The cheap Visaton woofer models in a similar fashion to the cheap Peerless except it needs a bigger box. I think the Peerless is a perhaps little smoother in the XO region
All my plots are on the computer that died unfortunately
 
And how are they?
I was surprised to see this still active. Sadly I don't think it will ever be realized.
The Jazzman are awesome.. They have great extension for the narrow baffle. They image well and cast a big soundstage. Tonally they are pretty neutral.
Subsequently I have seriously upgraded the rest of my system and they continue to reveal more and more. I power them with Rogue 180 monos and they sound lovely.
Here's a link.
http://www.stereo.net.au/forums/index.php?/topic/21047-new-speaker-build/?fromsearch=1
 
Nice combination navin. I think there are advantages using drivers from the same supplier as they tend to be tonally similar and save crossover time and cost, not to mention they are well behaved without nasties to knock out in the crossover. It puts up a strong case for SEAS due to worldwide availabilty and they are generally available for long term.

You have mentioned it previously and I agree the 27TDFC would be a good option as is a good tweeter that is the darling of the DIY crowd. I've had good success with it as well as the 27TFF and 27TFFC and depends on the application to which is better.

A single 10" vented option would be a good idea as has great bang for buck and an economical alternative that's smaller and a bit more WAF.

At this stage this seems to be heading towards:
Tweeter 27TFFC with option of 27TDFC
Mid MCA15RCY
Woofer CA26RE4X with 2x sealed (vented?) alt 1x vented (sealed?).

Ted said the mid is the most important driver and I agree as it carries the bulk of the character of a 3-way. When you think about it the woofer and tweeter are there for support even though they have a large roll to play. I haven't used the MCA15RCY but from all accounts from what I've read would be an excellent choice and hard to beat for the money. You could use the CA15RLY as a choice for those who want to crossover lower but I think distracts from the classic 3-way speaker ideal. If crossing over very low then a 6" option would be better IMO.

This is the combination which was agreed :)
 
A single 10" in a TL is fine.
a) I figured the cost and difficulty of making a TL box might push the project outside the abilities of some of the DIYaudioers. But if most of us are ahppy with making a TL then 1 x10" in a TL box will work well.
b) As Andre has pointed out below, 2 x 10" will have a sensitivity of 99db (as against my assumed sensitivity of 91-92 db after baffle step compensation)



Is this after baffle step compensation? The spec sheet gives a nominal sensitivity of 90.5db, if you look at the FR curve (SEAS claims their curve is taken in a 28 liter sealed box) the sensitivity is below 90db (under 300Hz). Add 2 woofers pushes that to 96db. factor in baffle step compesation and it should have been 91-92db at most. Have I missed something?



So far the drivers most have agreed on are SEAS 26RFX or RE4X, MCA15RCY and 27TDFC or TFFC.

The object of this project is as a reference for diyaudioers. Using better or more exotic drivers (Seas-Excel, Illuminator-Revelator, RAAL, PHL, etc..) might push this project outside the skills, abilities and finances of most diyaudioers.




Besides this Ultimate systems demand better XO and cabinetry skills than most of us have.



This system is just a fun project. If any diyaudioer wants to build a 3way or 2 way there are dozens of projects, kits and designs available on the net. One result of this project should be something we diyaudioers can call our own. Ofcourse help from the pros (Lynn, Planet10, etc..) is more than welcome but the idea is that the ownership of this project will be diyaudio and not any one person - kinda like open source software.

Also there was consensus for CA26RFX as bass module single or double in BR..