DIY Build tweeter question

[Mediocre Polymath]

Hello folks. I've had to call in sick to band practice today, so I've been sitting around at home coughing, drinking tea and going through some old project ideas.

 

Last year I started pondering the idea of making myself a small, lightweight 1x12 cab – something that I've discovered is an idea lots of people have had over the years. I made some fairly detailed design drawings, but didn't go ahead and make anything.

 

At the moment I use a cab that more or less fits the "small" criteria, but not really the "lightweight" one. It's a weird 1x10 cab that I designed back in 2015, using ineptly-operated winISD and random bits of information I could glean from the internet.

 

 

It's made from 18-mm birch-ply and it weighs somewhere in the region of 15–20 kg, mostly due to the hench Eminence Delta speaker (a funny discontinued model with a whizzer cone and a power handling of about 200 watts). It sounds fantastic, but it's generally a bit of a pain in the arse to move around.

 

My design criteria for the new design was that it should be 1. capable of handling around 250 watts at 8 ohms, 2. be no more than 350 mm wide (I'd like it to fit in a narrow alcove in my house), and 3. (most important) be light enough for one person to lug around on public transport.

 

 

This is what I came up with last summer, based around the Faital 12PR320 and a B&C speaker for the high-frequencies. I recently discovered that this is essentially a parrallel-evolution version of the Basschat 112, just slightly smaller and with, notably, a different tweeter.

 

Which gets to my question. Why do speaker designers favor compression drivers over small-diameter cone speakers?

 

This isn't an accusatory thing, nor am I trying to start a fight, I genuinely want to know what the reasons are.

 

I had a compression tweeter in my 110 cab originally, but I found it sounded extremely harsh and excessively loud. It burned out after a few gigs, and I replaced it with a small-diameter cone speaker (the same as the B&C one that's in this cab design).

 

Recently, Bill Fitzmaurice's detailed comments on this forum have led me to think this might have been due to the cheap off-the-shelf crossover I used, rather than the tweeter itself, but I'm still not sure if I like the full 40–20,000 Hz sound (but then again, I also don't like 40–4,500 Hz sound of tweeterless cabs). 

[Lfalex v1.1]

Might a tweeter level control be a worthy addition to your design?

 

They're quite common in commercial designs to help tame excessive/harsh trenle without having to mess around with your tone.

[Bill Fitzmaurice]

39 minutes ago, Mediocre Polymath said:

Which gets to my question. Why do speaker designers favor compression drivers over small-diameter cone speakers?

Because they're cheap, or at least the ones you see most often are. They cross over typically between 3.5 and 4kHz, because that's as low as cheap tweeters will go. They use barely adequate crossovers again because they're cheap. IMO the best high frequency option is a four to six inch cone midrange driver, crossed over between 1.2 and 2kHz. They tend to be good to 8kHz or so, which is all electric bass needs. You seldom see them because they're more expensive, as are well designed and spec'd crossovers. 

BTW, you can't lay the blame for this on speaker designers. Given free rein and a blank check most of us would do it the right way. This one falls on the bean counters.

[Mediocre Polymath]

Interesting, thanks. That tallies with my own subjective experience – I think the 4" speaker I've been using as a tweeter is a mid-range driver designed to go up to something like 8,000–10,000 Hz, and I've found I much prefer that sound. (I made the switch because I'd convinced myself that I needed a tweeter with a power handling greater than any easily-found compression tweeter, so I seem to have stumbled into the right answer by accident). I was worried that there some some arcane detail to do with impedances or phase problems or something that I was missing.

[Phil Starr]

First of all I'm going to say there are a lot of details to do with phase issues and a whole lot of other thngs involved in designing a really good crossover. Something that 'works' is simple enough and you can design a crossover using theory and software but speakers are far from straight resistive loads and neither are they flat response so combining two without a lot of distortion at the crossover is tough without a lot of experience and good measuring gear. That's not to discourage you or anyone from using a generic crossover but just be realistic about the possibilities. Having said that you can have a great sounding instrument speaker that falls short of perfection. However the BC 112T Mk3 has a really well developed crossover so don't dismiss it as a fully worked out design. You can change teh shape of the box to fit your alcove so long as you keep the volume of the box the same

 

On the other hand it's quite right that the top frequencies above 8kHz aren't really important for bass and won't actually be there art audible levels in the output of a bass pickup. Though even a 4" driver will be beaming at 2kHz so a well designed horn will give you better off axis response. I've been experimenting at home with a 6" driver and a 12" bass driver using DSP for the crossover and found that a 200Hz crossover worked well and gave a remarkably smooth sound. Using a small cone driver to cover all of the mid-range means that these frequencies won't be beaming and a smaller lighter cone is more able to track the mids accurately. Our ears are most sensitive in the 1-3kHz range so moving the crossover out of that region is really helpful. Having nothing above the pass band of a bass pickup is just not having something you will ever use.

 

The tests I've been doing have just been proof of concept, there aren't many reasonably priced midrange drivers available off the shelf. 6" is too big really and I've bought 4x3" drivers for the next stage of development. These little drivers don't have the power handling or sensitivity to achieve the right volumes on their own.

 

There's nothin new in all of this, it's been understood since the 1930's and probably earlier. Ideally you'd split up the entire audio range into 4 or 5 bands but passive crossovers are difficult to design and always bring in some distortion. Fortunately for bass the pickup itself acts as an inductor and cuts all the high frequencies depending mainly upon how many windings there are in the coils. The strings of course don't vibrate well at high frequencies so there is much top to start with.

 

So it's relatively easy to reduce the crossover point for a compression driver and horn to around the 2kHz point without crazy expense and minimise crossover distortion to  get great radiation control of the upper frequencies, which is what @stevie has done with the BC112T and his LFSys designs. It's also possible to crossover at much lower frequencies and get better dispersal at sub 2kHz frequencies and let a single driver or array of drivers cover the whole mid range but with significant beaming above 4kHz which is what Genzler have done in their bass arrays. You'll have to design your own crossover for that and probably need to wind your own coils as high inductance/high power coils aren't widely available off the shelf.

 

 

[Mediocre Polymath]

Hi Phil, thanks for the detailed response. I hope you didn't think I was casting any shade on the basschat 112 design – my current plans for this speaker can probably be most accurately described as "the BC 112, but a little bit smaller and a little bit worse".

 

I'm willing to make some fairly significant compromises on bass response and general fidelity if it means I can carry it down the rickety Jacob's ladder that goes up to my attic without worrying halfway down that I'm going to fall and die. My modelling in WinISD suggests it will be louder and bassier than my current 110, and that's worked fine in every situation I've used it in so far.

 

What's the crossover frequency on the basschat crossover, because I'd be tempted to just copy that design, again, imperfect though it would be in my set up.

[Bill Fitzmaurice]

Designing crossovers is no more difficult than designing boxes, thanks to software like XSim. https://xsim.software.informer.com/

The only issue is the need for FRD and ZMA files. They're available for most hi-fi drivers, with pro-sound you probably have to create them yourself. But it's no more difficult than any other part of the speaker design process.

Coils are readily available, I haven't had to wind my own since the 1970s. They are expensive and quite large when used in a passive crossover lower than 200Hz, but that's also something I haven't done since the 1970s. Below 500Hz active crossovers and bi-amping is the way to go.

The easiest method to get better mids and highs is with a sealed back midrange, like the PRV 6MR 300 SEAL. Crossed over at 800 to 1200Hz it will keep up with any woofer.

As for using higher order high pass filters to run tweeters to 2kHz, I've been doing that for 25 years. The bass cab industry as a whole has yet to catch up... or should I say wake up? 😉