audio-talk

Nice day, bright sunlight, the hound and I have the house to ourselves, and I've got bird flu. So since I don't feel like doing much, I thought I'd have a search through the archives to see what else I could stick up that might be of interest. This one's something I did a while back for a chap in Canada who asked if I could give him a hand, and since it turned out well and is a bit different, I thought I'd dust it off. Like most of my one-offs, it never had a name, but I discovered my old copy of Dungeon Keeper* at the back of a cupboard a few days ago, and it seemed a good handle. So, Dungeon Keeper it is.

The requirement was for a floorstanding 2-way, 2-driver loudspeaker that looked like a classic ProAc design: simple rectangular box, no chamfers or roundovers permitted. He already knew what drivers he wanted to use (next posts), and that while he wanted the ProAc look, he didn't want the typical ProAc balance. Fair enough. That's probably where this speaker is a little different. Kevin (chap who asked me to design it) is a bit of a 'perfectionist' on some things. So for all the simple box and driver configuration, there's a bit of over-engineering going on in the filter. Where the Saiga design I posted a little while ago uses a fairly simple crossover with a low component count, this is a little more ambitious. I chose it because it's different, and some of the features might give some ideas.

High value, high performance 8in bass driver required. Kevin's a martyr to pistonic operation, which has merits if done properly. Since he's in North America, the Dayton RS225-8 was the logical choice. However, the 225 is a pain to work with; it's closer to a bass than a midbass driver, has mass bell-modes at the top of its range which must be addressed unless you enjoy listening to huge uncontrolled resonances and the resulting IM / harmonic distortion in the rest of the range too.

On the up side, it kicks like a mule in the bass, has good extension in a vented box, plenty of excursion for dynamic peaks, low distortion if used properly, and looks like a stealth fighter.

And for something unusual...

The tweeter is the Usher 9980. This is not available to DIYers unless you knew someone, or could find a pair from OEM overstock, a pair from a broken finished speakers etc. However, it's a close cousin to the Dayton RS28a-4, which is built by Usher for Dayton. The only real difference is the dome material; black anodised aluminium for the RS28, beryllium for the 9980. Or at least, that's what they call it. In reality it's probably not pure beryllium. We don't know for certain; this is speculation only. The appearance (purple-grey) is right, but general behaviour is almost identical to the contemporary RS28a, which would be very unlikely if the dome was pure beryllium. However, titanium can perform quite closely to aluminium, so it's possibly either a titanium low-alloy with a small percentage of beryllium, or titanium with a beryllium layer added by some means.

Whatever the actual dome material, it's got the same performance as the RS28a-4 -nice motor, low distortion, low Fs, reasonably well controlled impedance, decent excursion and good build quality, so you can cross it pretty much how you like. Not really much to criticise. Does it sound the same? Pretty much, as you'd expect given the almost identical FR and distortion performance. If I was pushing it, I might say the 9980 had a touch more 'air' at the top end, but I don't trust impressions like that; I didn't have a pair of contemporary RS28as to hand when I heard them, and you always get differences between or even within batches depending on what production tolerances are. I haven't got a decent picture of it, but it looks (slightly different dome colour aside) like the RS28a-4, so the latter is standing in here.

Nothing unusual here, a straightforward reflex box using a moderately damped alignment. It's down-firing through the plinth, so needs raising on 1in cone spikes or similar. The room it's used in is about 25ft x 17ft as I recall, wooden floor, speakers about 2ft from the front wall on either side of a stone fireplace. The walls are also (insulated) wood, which leaches some LF output & was a consideration in the general voicing. The vent is shortened relative to the nominal alignment due to the resistive loading it encounters between floor and plinth. A basic sim of the target load is attached. Nominally with these dimensions you get some longitudinal standing waves interfering with the alignment above nominal tuning, but filling the box up to the upper end of the vent with Dacron, along with the down-firing configuration clobbers this sufficiently not to cause any issues; they're sufficiently low to be swamped by room effects; a degree of rationality is needed here.

Cabinet dimensions also attached; 18mm build material was assumed; take your pick of whatever floats your boat. I'd personally prefer to double the front baffle (at least). Kevin used BB plywood, lagged with 1in Ultratouch recycled denim. There are a couple of window-braces; the exact position isn't critical, but asymmetric spacing is generally preferred. The boxes are mirror images; the RS225 is on the centreline, the 9980 tweeter laterally offset by golden ratio to dissipate some of the energy from standing waves on the front baffle; useful when you've got sharp 90 degree edges. It does make a difference. The listening axis is on the tweeter, and the tweeters are intended to be toward the inner, rather than the outer edge.

Dungeon Keeper logo shamelessly borrowed from the original game. Since these are not a commercial speaker, I trust they don't mind me borrowing it for a bit of fun.

The fun part for me. Since I'm lousy at building things, I enjoy the acoustic and electrical design parts. Minimalist this is not. This is what happens when you have drivers that offer a lot of performance potential, but also have significant characteristics that have to be overcome. The 9980's a doddle, you can do pretty much what you want with it, which is good because the same can't be said of the RS225-8. I'll upload a single-driver speaker next; I haven't put one up here for a while. But sticking with this for the moment...

Attached is the response of the drivers, on the speaker baffle. Characteristic hump 1KHz - 2KHz in the tweeter response, as you always get from a baffle about this size. Otherwise, quite smooth, no nasties. Small hole in the HF response about 16KHz, but nothing tragic. RS225? What you'd expect. Typical baffle step losses, but otherwise flat as Kate Mara's stomach out to about 1.5KHz. Above which it goes to pieces. Mass & profile are taking over, big null centred on 3KHz and a very tasty 10dB resonant mode centred on about 6.5KHz that melts your ears if it's not bludgeoned into submission (partly the resonance itself, partly the distortion it causes & amplifies). Sadly it's not practical to fix this mechanically; you can get some way, but not enough if you want an accurate speaker: it's too severe to realistically kill that way, even with carefully targeted additional mass. So by the time you've finished mucking about with it, you still need fairly heavyweight filtering, and you've damaged some of its better properties. On the whole, slapping it into submission via the filter is the more practical option. Perhaps I should have called this the Mr Grey speaker? Still, in Dungeon Keeper, you can give your minions a slap to keep them motivated, so it's appropriate enough.

Yes, it's that time.

Do you iron creases in your jeans?
Do you like flasks of weak lemon drink?
Do you really, really hate crossovers?

You're not going to like this at all.

Good.



I need to make something spicy to kill the bird flu. Will go through the filter later though. Suffice it to say temporarily, it's a sledgehammer. LR6 (36dB/octave) at 1.25KHz. Don't try this with most tweeters; they'll be begging for mercy. The Usher is one of only a handful that can hack it, and this really is the limit of what even it can handle, even with a lot of protection from the crossover.

System response is also attached. I haven't spliced in the cabinet output < 200Hz because I've got bird flu and can't be bothered. Target response trend was an overall slight decline toward higher frequencies, with that slightly reduced at the very top end to compensate for the inherent dip in the HF response of the 9980, and prevent it sounding too dull / lacking 'air'.

Here we go again, do you really expect people to waste their time reading this - do you have ears

Yes I am being rude, but really.

Dr Bunsen Honeydew wrote: ↑Sat Apr 01, 2017 6:21 pm Here we go again, do you really expect people to waste their time reading this - do you have ears

Yes I am being rude, but really.

Well I'll be rude Richard.

Fuck off and leave this thread alone, some of us (like me) want to read it without you shitting all over it.

once again you are first.

I ask you a question - do you understand what all this gobble-de-gook is about - really, in other words what the feck it has to do with a speaker playing music.

Dr Bunsen Honeydew wrote: ↑Sat Apr 01, 2017 6:49 pm once again you are first.

I ask you a question - do you understand what all this gobble-de-gook is about - really, in other words what the feck it has to do with a speaker playing music.

Please STOP polluting this thread, you really are making yourself look like a dumb arsehole.

I wish we could do a survey / poll here to see how many people agree with you. As far as I am concerned the OP is taking the piss so deserves the piss being taken back to him. So I ask again, in plain English, explain to me what he is trying to tell us, or is the point to make it so full of obtuse illegible and illogical tech bollocks that all it does is make you think he knows what he is talking about

OK no point in pursuing the point, I have had my say, now all you pretend ego driven experts like Noddy just get on with nodding your heads as though you know what this gobble de gook is all about.

I don't understand a lot of what Scott has written, but like yourself Richard, he produces loudspeaker designs that other people enjoy listening to. He does so by a method that is the polar opposite of yours, but the fact that he produces results that others enjoy would indicate that his method is equally valid. There is room on this forum for both of you to propound your theories, and each should be able to have his say without the other belittling him. Of course you can disagree with him, but you do yourself a disservice by doing it the way you do.


And it's bad for Phil's blood pressure.



"There are two ways of doing things. My way and the wrong way". Janet Street-Porter.

Dr Bunsen Honeydew wrote: ↑Sat Apr 01, 2017 7:02 pm I wish we could do a survey / poll here to see how many people agree with you.

You do seem to be pursuing Scotts posts Dr B in the same way you suggest Phil does to yours.

I can understand Dr.B's reaction in connection with that crossover. Wouldn't have put it that way myself , but ....

Nick wrote: ↑Sat Apr 01, 2017 7:35 pm

Dr Bunsen Honeydew wrote: ↑Sat Apr 01, 2017 7:02 pm I wish we could do a survey / poll here to see how many people agree with you.

You do seem to be pursuing Scotts posts Dr B in the same way you suggest Phil does to yours.

Different, Noddy pursues the individual it doesn't matter what I say. I can assure if this thread had any sense in it I would agree with it and be happy, but quite simply it is taking the piss, so it deserves it in return. Is he trying for the job of head designer at B&W, he would confuse and bullshit the owners so much (as they haven't a bloody clue what make a good loudspeaker) they would all nod their heads knowingly and employ him, and they would continue producing non musical shit!