advertisement


following on from the 'glass ceiling' thread

Rockmeister

pfm Member
I often muse on designing a 'perfect' bit of kit, and mostly, speakers.
Gilbert Briggs is to blame.
Bloody upward firing 12" bass drivers in concrete corner cabs.
Me and a pal did this at Art college in 1970.
He paid so he kept them (not that they were going anywhere) :)

Look at all the expertise here!
Surely we can design something amazing
Perfect phase coherence
no colouration
easily driven
......
I'm semi serious

It'd be great for the PFM collective to invent something weird and then some bored nutter (Rocky puts his hand up) to actually make them in the garage (domestic acceptability NOT essential ) and show them off (sound them off I mean) at the 2023 wam show

Anyone fancy this idea. Amps and front ends follow
....
 
I'd be interested, but I've just spent £11k on a car, so funds are tight for me. At the very least I'd be more than happy to take the measurements and design the crossover. Obviously once you've got measurements, anyone can have a go at designing a crossover.

If anyone wants to bring some of their favourite music to my place in Cheltenham, I'd be more than happy to prove I know what I'm doing.
 
A few questions -

What's the budget?
Sealed, ported or?
Maximum size?

Ask any speaker designer to suggest their perfect speaker, and they'd all give a different answer. For me, with every design, I always learn something new and when thinking about the next design, I try to improve on the worst aspects but keep the good bits.

Perfect phase coherence
no colouration
easily driven

Perfect phase coherence - The problem here is, without a hell of a lot of attention to detail when it comes to measurements, and processing those measurements correctly for the crossover design software. You won't have perfect phase coherence, even though the software is telling you, you have. I've lost count of the number of times that someone uploads a graph showing the phase tracking beautifully throughout the crossover, and then we find they forgot to include offsets in their simulation! With an even order crossover, you can reverse the polarity of one of the drivers, and take a measurement to make sure you have a good 'null', but that doesn't mean the phase is as good as it could be.

No colouration - Obviously that's impossible, but your average speaker is such an imperfect thing, that there's plenty that can be done to lower colouration. Most speakers are rectangular boxes, and this is bad because as the sound diffracts off the sharp edges, it causes peaks and dips in the frequency response. It's argued that this is what causes a narrowing of the soundstage.
The ideal speaker enclosure shouldn't sing along with the music, IMO. The stiffer the cabinet, the higher in frequency the cabinet resonates, and this often ends up where your ears are most sensitive. Thin wall cabinets with no bracing, don't sound as tight in the bass. I think I've found a good solution with stiff, reinforced foam, but although it's easy to work with, it's not easy to cut accurately or achieve a nice finish.

Easily driven - Ported speakers are usually easier to drive than sealed, but I know there are a lot of fans of sealed on pinkfish. Ported done properly beats sealed IMO, but it's not easy to get right, and that's why sealed betters ported the large majority of the time.

I find higher order crossovers sound too forward, but low order are difficult to get right when crossing over at higher frequencies. The drivers need their relative acoustic centres to be somewhere near each to achieve good phase alignment, either by having a step in the baffle (causes diffraction problems), using a waveguide and small midrange or an angled baffle. The other option is a dome mid because the acoustic centre will be almost, if not completely identical, to a typical dome tweeter. You also need drivers with a smooth, extended frequency response, and because of the shallow slope you don't want a breakup peak that you get from your typical hard cone.

For a dome mid I quite fancy the Scanspeak D7608/920010 as used in the Alchris Audio AR-66 or the new Bliesma M74S

There are plenty of nice tweeters out there. I personally try to avoid ferrofluid, because removing it in experiments made a surprising improvement - This is a nice tweeter for the money (Scan D2604), and looks like it could be the same one used in the Proac DB1

There are lots of nice high efficiency 10" to 12" pro audio drivers out there. It just depends on what size driver you'd want, or if you'd rather go sealed, Peerless and Scanspeak make suitable drivers.
 
Hi Mat.
I didn't understand any of that:)
Which sums up my knowledge very well i think ;)
Well i did know some but once we get all technical??
My thoughts were that we would start by defining the very basic needs of a speaker. It's essential purpose. That would give us a list of design priorities.
After that we might try to define what type of drivers each section of the frequency range would deliver our needs best, and then how to construct, to engineer a structure that gave each driver the best chance.
The we need to tie the sound down to actually sound coherent and real at a listening position, and then design a crossover to suit.

I had no idea of cabinet structure except that, do tweeters need a cabinet?
Should bass and mid be separate?

My imagination had drifted onto something like big Wilsons, with a series of separate boxes (or no box) linked by scaffolding and clamps that were adjustable for position. By loosening a clamp, we slide the tweeter around until it sounds right for this particular room.

Domestic arrangements are irrelevant here, and as to budget, I have no idea, but as an exercise, let's say a million quid and then get real when there is some design to think about.

This is fun. A knowledge pool and input from anyone who knows anything what we need first. (Massive shout off followed by 3 flounces followed by some great ideas being knocked around and.....):)

It may never work, but it hope it will be enlightening and fun to chat about anyway. Who knows?
 
Just to be more useful.
The best bass I ever heard was from something very BIG, enclosed in a concrete box. So, large diameter driver, in cabinet with zero vibration.
I know nothing whatsoever about mid units. Might electrostatic panels be best? A horn?
The top end? I love Scanspeak soft dome tweeters. I never heard better, but ribbons might be a shout. I don't like metal dome tweeters. just, NO.

Bass is less directional than top and mid, so the bass cabinet is most flexible and easiest to place. Being big and heavy it makes a good support for what goes above.
If anyone made a point source mid and top unit, I think the phase business would be sorted. Like Tannoys but without the bass, and sweeter. One might need a supertweeter sort of thing, but like bass, I think I know that they are more flexible where they go. Hello scaffold pole, about which I'm serious btw. Filled with sand or expanding foam, they are fairly dead, strong and rigid and inexpensive. Also interesting from a design standpoint and easily painted.

I thought maybe decorated, but lets not go there yet?

And ther is the total of my ignorance.
 
Hi Mat.
I didn't understand any of that:)
Which sums up my knowledge very well i think ;)
Well i did know some but once we get all technical??
My thoughts were that we would start by defining the very basic needs of a speaker. It's essential purpose. That would give us a list of design priorities.
After that we might try to define what type of drivers each section of the frequency range would deliver our needs best, and then how to construct, to engineer a structure that gave each driver the best chance.
The we need to tie the sound down to actually sound coherent and real at a listening position, and then design a crossover to suit.

I had no idea of cabinet structure except that, do tweeters need a cabinet?
Should bass and mid be separate?

My imagination had drifted onto something like big Wilsons, with a series of separate boxes (or no box) linked by scaffolding and clamps that were adjustable for position. By loosening a clamp, we slide the tweeter around until it sounds right for this particular room.

Domestic arrangements are irrelevant here, and as to budget, I have no idea, but as an exercise, let's say a million quid and then get real when there is some design to think about.

This is fun. A knowledge pool and input from anyone who knows anything what we need first. (Massive shout off followed by 3 flounces followed by some great ideas being knocked around and.....):)

It may never work, but it hope it will be enlightening and fun to chat about anyway. Who knows?

Tweeters are mostly sealed units, so no need for a separate enclosure.
Bass and mid are best kept separate.

The tweeter (and the rest of the drivers) would need to be in a fixed position. If you change the position of the tweeter on any axis, the crossover would have to be adjusted.

I highly recommend you read this - https://www.diyaudio.com/community/...-design-your-own-speaker-from-scratch.332688/
 
I find higher order crossovers sound too forward, but low order are difficult to get right when crossing over at higher frequencies. The drivers need their relative acoustic centres to be somewhere near each to achieve good phase alignment, either by having a step in the baffle (causes diffraction problems), using a waveguide and small midrange or an angled baffle.

My objection to the angled baffle solution is that the sound in the room will be brighter than on the listening axis.
 
My objection to the angled baffle solution is that the sound in the room will be brighter than on the listening axis.

I've always avoided angled baffles because I was worried the top octave or two would suffer where the tweeter would be beaming, and have a lower output at your ears. Thing is, I listen off-axis anyway, so I'm not sure if it would matter.
 
I've always avoided angled baffles because I was worried the top octave or two would suffer where the tweeter would be beaming, and have a lower output at your ears. Thing is, I listen off-axis anyway, so I'm not sure if it would matter.

I mentioned that because of my memory of the Spica TC50 speakers. They were great in the sweet spot, but failed the "sounds good in the next room" test. They seemed to spew 1K-5K all over the ceiling .

Still, I would love to see a dynamic speaker with a good-looking step function, or perhaps even pass a recognizable square wave.
 
There are some measurements of the TC-60 here. He says it's a 1st order crossover and then goes on to say they are not in phase and should be down 6db at the crossover. He's wrong, if it's an odd order crossover, it should be 90 degrees out of phase, and down 3db at the crossover point.

 
Remember the old WAD KLS3 DiY design from HiFi World?
It had an open backed mid range section. That speaker failed because it had the HD-3P tweeter that deflated.
 
I often muse on designing a 'perfect' bit of kit, and mostly, speakers.
Gilbert Briggs is to blame.
Bloody upward firing 12" bass drivers in concrete corner cabs.
Me and a pal did this at Art college in 1970.
He paid so he kept them (not that they were going anywhere) :)

Look at all the expertise here!
Surely we can design something amazing
Perfect phase coherence
no colouration
easily driven
......
I'm semi serious

It'd be great for the PFM collective to invent something weird and then some bored nutter (Rocky puts his hand up) to actually make them in the garage (domestic acceptability NOT essential ) and show them off (sound them off I mean) at the 2023 wam show

Anyone fancy this idea. Amps and front ends follow
....
I too must admit to bring a fan of Briggs, I have all his books from the 50s and early 60s.
He writes with an entertaining logic that I like and understood easily when I started out.

I started on a similar quest to what you describe, in the 90s, building the best I could and taken my creations to most wam shows since,
I quickly found out that to get a really good sound, good large efficient drivers are required.

Hence I probably have close to a hundred drivers of all types stashed away that have either been stepping stones or I've got plans for. It's a slippery slope.
 
...
Surely we can design something amazing
Perfect phase coherence
no colouration
easily driven
......
Perhaps there's an element to add that it "goes loud enough before level compression"? Or maybe that needs qualification as in "goes loud enough for the desired bass extension and the room size"? And for some definition of "loud enough".

Not going loud enough for my room and my musical taste was one of the defects I think was present in my previous loudspeakers. I decided to address it the last time I upgraded them. From ones with a 170mm bass unit of unknown SPL capability to ones with a 235mm bass unit and a 112 dB SPL (continuous) rating. Should I want to go there, close-miked piano now sounds as superbly dynamic as if it's not that far away.

AFAICS with loudspeakers it's easy to list the potential defects to address but it seems you always have to compromise in some way in practice. So in a "no compromise" design it looks to me more like a case of listing the key compromises and designing for them to be "too small to matter".
 
Remember the old WAD KLS3 DiY design from HiFi World?
It had an open backed mid range section. That speaker failed because it had the HD-3P tweeter that deflated.

Brian of this forum has or had a pair. I can't remember seeing an open backed midrange section in his, but I do remember his tweeters had deflated.
They are excellent speakers and usually an absolute bargain if you can find a pair.
 
I often muse on designing a 'perfect' bit of kit, and mostly, speakers.
Gilbert Briggs is to blame.
Bloody upward firing 12" bass drivers in concrete corner cabs.
Me and a pal did this at Art college in 1970.
He paid so he kept them (not that they were going anywhere) :)

Look at all the expertise here!
Surely we can design something amazing
Perfect phase coherence
no colouration
easily driven
......
I'm semi serious

It'd be great for the PFM collective to invent something weird and then some bored nutter (Rocky puts his hand up) to actually make them in the garage (domestic acceptability NOT essential ) and show them off (sound them off I mean) at the 2023 wam show

Anyone fancy this idea. Amps and front ends follow
....
I often muse on designing a 'perfect' bit of kit, and mostly, speakers.
Gilbert Briggs is to blame.
Bloody upward firing 12" bass drivers in concrete corner cabs.
Me and a pal did this at Art college in 1970.
He paid so he kept them (not that they were going anywhere) :)

Look at all the expertise here!
Surely we can design something amazing
Perfect phase coherence
no colouration
easily driven
......
I'm semi serious

It'd be great for the PFM collective to invent something weird and then some bored nutter (Rocky puts his hand up) to actually make them in the garage (domestic acceptability NOT essential ) and show them off (sound them off I mean) at the 2023 wam show

Anyone fancy this idea. Amps and front ends follow
....

Look forward to reading the outcome/s. 'Perfection' is a hard taskmaster, my aim was to be 'good enough'. Having farted about with DIY speakers for 60 years I ended up with those in the avatar, which were at the limits of my skill set and pocket. I remember the concrete boxes from way back, great idea, and at the time cabinets without colouration were a very attractive approach, concrete moulding was possible, but it is b.....dy heavy stuff, the corner cabinets in concrete are brilliant in concept, but movable they are not. As you say there are some bright and knowledgable contributors on pfm, how about defining and agreeing some parameters first, which would of course be restrictive. However without constraints, there are just too many variables, I think; especially for a project designed by committee.
One interesting outcome might be the individual pearls of wisdom that are already emerging. Would good to put them in priority order, or give each a score.
Forgive me, the wanderings of an aged mind, have fun
 
Brian of this forum has or had a pair. I can't remember seeing an open backed midrange section in his, but I do remember his tweeters had deflated.
They are excellent speakers and usually an absolute bargain if you can find a pair.
I have still got the HFW supplement with the design. The mid range section had an open back with a curtain, the idea was to stop cabinet reflection
 


advertisement


Back
Top