Speaker rebuild/redesign help!

[djsixbillion]

I’m new on this forum, but have built a few pairs of speakers (ghetto though they may be) to date. I’m now looking to put together something “better,” and am in need of experienced advice. Here’s what I’ve got: LCR MTM’s with HiVi F6 mids and RT2II ribbon tweeters, using basically textbook 3rd order Butterworth crossovers @ 2800 hz (with a zobel on the mids and L-pad on the tweeters). I also run them with a DIY sub using two Peerless CSX 10” woofers in a vented, isobaric setup.

I’m pretty happy with the sound, but I think that it could be much better with a more “optimized” crossover. I’m also not all that pleased with the enclosures, and want to rebuild them (I made the L/R towers too tall, for one thing, so that the tweeters are a few inches above ear level when seated).

So here are the questions: how do I go about designing a “perfect” crossover for these drivers, if such a thing exists? And is it worth it? How about other options, like throwing some 4” or so mids in the mix, and making an MTMWW using those and my existing drivers? I’ve been wanting to build a bigger sub anyways, so how about putting one of the 10” Peerless woofers in each L/R tower, crossed over at, say, 300 or so? I guess that would be an MTMwwW...

As you can probably tell, I’d like to reuse as much of the “stuff” that I’ve already bought as possible. Any suggestions or ideas welcome! My goals are, basically, everything: great midrange and imaging, tight, punchy bass down to at least 40 hz or so....oh and good efficiency so I can keep using my 35wpc tube amp!

Thanks,
Desmond

[JonMarsh]

Are you prepared to get some measuring capability? At least an inexpensive mic like the Behringer ECM 8000, and free ware like Speaker Workshop, and also build a measurement jig for impedance? There's a good manual for Speaker Workshop written by a third party at Audio DIY Central. That manual also walks you through the measurement and design process.

It sounds like you didn't do baffle step compensation in your existing speakers- a text book 3rd order crossover at 2800 won't include that. That improves the in room frequency response, but lowers the sensitivity. The HiVi F6 starts to get kind of rough above 2 kHz, and considering CTC issues for crossover integration, a much lower crossover frequency will probably work better and clean up the presence range. However, the HiVi RT2II won't handle that, due to it's limited frequency range and travel. One other point to consider, some time ago I was interested in this tweeter, to, but came across measurements for it which were particularly disappointing- poor distortion performance, and a lot of high order distortion. Like other planar ribbons, such as the Yago, they tend to have poor energy storage characteristics from about 3 K down.

If you want to keep all of your drivers, you'd going to have some limitations in what you can do. One possibility for adding a midrange is a high quality dome unit, like the Dayton RS52, which is very clean from 800 Hz to about 10 kHz. Then a lower frequency crossover is possible for the midwoofer- maybe something like the MTMW we've been working on for a center channel and mains.

Modula NeoD R&D thread


For relevant comments about low cost ribbons vs well engineered low cost domes, this recent thread may be helpful.

Dayton PT2C-8


Some of us do a fair amount of driver testing, not just frequency response, but energy storage, harmonic distortion, and intermodulation distortion. All of these parameters affect the reproduced sound, and when you have have a 20 dB difference in performance between two drivers, it's not subtle in the audible results, in fact it's usually quite audible while running the tests. This is especially true if the high harmonics (4th, 5th) aren't at a low level, as the separation from the fundamental is quite obvious. This is what makes sound hashy and harsh- like what folks complain about in digital, frequently.

Some of the ribbon drivers have pretty good performance if you stay above 4 kHz with them, but many don't. And considering cost, if you can get a good dome tweeter with low distortion, a clean CSD profile (cumulative spectral decay), and extended response to 30 kHz for $30, then one has to work a bit to justify a more expensive approach.

My recollection of the results for the HiVi tweeter was that with muli-tone IM tests it produced a lot of intermod harmonics, over a wider frequency range, and at higher levels.


Just some things to think about.

[djsixbillion]

Thanks for the reply Jon. Your comments kind of reinforce what I had already been thinking, just based on listening impressions: namely that the F6 and RT2II both sound a bit rough at their respective frequency extremes. I don't really like the idea of ditching these drivers, simply because they will have no other home...but that may be what happens.

I have downloaded speaker workshop and started playing with a 3-way crossover, incorporating a midrange like the one you suggested. There's a bit of a learning curve, though!

[JonMarsh]

I have downloaded speaker workshop and started playing with a 3-way crossover, incorporating a midrange like the one you suggested.


There's a bit of a learning curve, though!

No kidding.... remember that some of us slow learners like me have been at this for 35 years+... still learning new things, better ways to do stuff, and sometimes, just plain having to dig myself out of a pickle or a corner I painted myself into... :W

Hang in there, and welcome to HT Guide Mission Possible.


~Jon

[djsixbillion]

Thanks for the encouragement. I'll keep plugging away with the software and see what i come up with. With reference to your earlier suggestion, what do you think of a TMWW with my current drivers plus an RS25 mid-dome (would be a funny-looking speaker, but might sound ok)? What would be the pros and cons of doing that versus an MTMWW with some 4" mid-cones? I can't seem to locate the part number right now, but there was a Vifa 4" mid I was looking at that seemed to be pretty flat from 300-8000 or so.

[djsixbillion]

Well, after more than a little fiddling with Speaker Workshop, I gave it up as a bad job and moved on to fiddling with Jeff Bagby's Passive Crossover Designer. This, combined with the FRD Consortium tools, proved to be a much easier method for simulating driver and crossover responses.

Here's what I've come up with so far: Put the two Peerless 10-inchers into their own 1 cu. ft. boxes (which I already happen to have), and set them directly next to the new enclosures I'm designing, which will each house two of the HiVi F6's in about a 1 cu. ft. volume, topped with an open baffle section containing a Dayton RS52 mid-dome and the RT2II ribbon tweeter (imagine something like a small Arvo Part, about 38-40" overall height, but with the bottom section being a sealed enclosure). I'll post a sketchup drawing when I get it into a more presentable state.

The mid-tweeter crossover will be passive, and I've currently been modelling a pretty simple 4-th order L-R at 5khz, with a zobel on the RS52 and small pad on the tweeter. The rest of the crossover will then be active, by way of a Rane SAC-23 (which I already own), bringing in the F6's from ~250-1250Hz, and the Peerless 10" below that. Bass below 50Hz or so will be augmented by a future sub build (Rythmik 15", perhaps?).

I've attached a graph of the simulated response below. Black is the overall response and green is the power response. Its not perfect, and doesn't include any BSC, for instance, but I plan to take care of that with some sort of eq (either a behringer deq2496 or a rane parametric, for the F6 and Peerless woofers, this will also help to smooth the lower midrange response a bit). The system will be tri-amped, with 100 watts to the 10-inch woofers, 50 to the 6-inchers (in parallel), and a 35-watt tube amp (modified Dyna ST-70) for the mid/high section.

Does anyone have any input/suggestions? Is this all a bad idea?!

Thanks