Several years ago I made quite an investment in a DIY pair of fairly large 3-way speakers, based around Scanspeak 6" and 10" revelators and a tweeter in a Pellegrene waveguide. Then turbulent times hit and I suddenly found myself in a tiny flat, in which they could barely be stored - let alone used sensibly. However, I kept the drivers and am beginning to come up with a new plan for them:
The aim is to make things much more flexible this time, to account for wherever I'm living over the next several years or more. So the 6" and tweeters will form a pair of small sealed 2-ways that I can use almost anywhere in their own right. They will then be joined by either a new subwoofer, or else mounted on a pair of separate bass cabinets using the 10" drivers (to make them effectively 3-ways again). Should space etc permit at any point, then it could all be set up at once, as 3-ways+sub. I'm thinking active crossovers, that can be easily changed or switched-between to suit the different configurations.
The aim is to make things much more flexible this time, to account for wherever I'm living over the next several years or more. So the 6" and tweeters will form a pair of small sealed 2-ways that I can use almost anywhere in their own right. They will then be joined by either a new subwoofer, or else mounted on a pair of separate bass cabinets using the 10" drivers (to make them effectively 3-ways again). Should space etc permit at any point, then it could all be set up at once, as 3-ways+sub. I'm thinking active crossovers, that can be easily changed or switched-between to suit the different configurations.
I'm just starting to model roughly what I could expect from this, and it looks like the small 2-ways used alone would peak at 95dB (@1m) before linear travel is exceeded, whilst reaching down to 45hz at -f6. The addition of a sub handling below 80Hz would raise the 2-way's peak SPL to 106db (@1m) and of course overall reach much lower frequencies. Or the 10" bass cabinets handling at least up to 150Hz (or likely higher, by crossover preference) would stop the 2-ways being limited by excursion at all - though in contrast to a good sub, the 10" cabinets would only reach down to about 35hz at -6dB (unless possibly teamed with a linkwitz transform).
So, the initial rough modelling makes it 'seem' reasonable to attempt, though there will be much more to do before I get to anything very specific. The 2-ways will be the priority to build, but less sure about the order of the sub or bass cabinets, or if I should just settle on one or the other rather than both. I'm tempted by the sub alone, but the 6" drivers (edit: 15W/8530K01) actually only have a 4.3" effective diaphragm diameter so it might be quite a leap without some bigger bass drivers.
I'd be interested to hear if anyone has thoughts on that, or any other general aspects of the concept at this early stage; things to consider, avoid and so on?
Cheers,
Kev
So, the initial rough modelling makes it 'seem' reasonable to attempt, though there will be much more to do before I get to anything very specific. The 2-ways will be the priority to build, but less sure about the order of the sub or bass cabinets, or if I should just settle on one or the other rather than both. I'm tempted by the sub alone, but the 6" drivers (edit: 15W/8530K01) actually only have a 4.3" effective diaphragm diameter so it might be quite a leap without some bigger bass drivers.
I'd be interested to hear if anyone has thoughts on that, or any other general aspects of the concept at this early stage; things to consider, avoid and so on?
Cheers,
Kev
Your plan is very feasible. The 6" revelator will really shine in the midrange if it is high-pass filtered at 150 to 200 Hz.
I have a similar modular system. I have a pair of woofer cabinets with 12" sealed drivers. I have an active 3-way amp system built into the speaker stands, and I can swap out the small 2-way speakers (which I call head units).
You may find something useful in my two threads: This is the first project, using SB drivers
https://www.diyaudio.com/community/threads/new-active-3-way-hypex-and-sb.352767/
The second project uses Satori textreme drivers for the mid and tweeter, but I kept the woofer cabinets and used the same stands with the built-in Hypex amps.
https://www.diyaudio.com/community/threads/new-active-satori-textreme.366347/
Personally, I think you should stick with the 10" ScanSpeak woofers for now. You may find that there is no need for a true "subwoofer".
j.
I have a similar modular system. I have a pair of woofer cabinets with 12" sealed drivers. I have an active 3-way amp system built into the speaker stands, and I can swap out the small 2-way speakers (which I call head units).
You may find something useful in my two threads: This is the first project, using SB drivers
https://www.diyaudio.com/community/threads/new-active-3-way-hypex-and-sb.352767/
The second project uses Satori textreme drivers for the mid and tweeter, but I kept the woofer cabinets and used the same stands with the built-in Hypex amps.
https://www.diyaudio.com/community/threads/new-active-satori-textreme.366347/
Personally, I think you should stick with the 10" ScanSpeak woofers for now. You may find that there is no need for a true "subwoofer".
j.
Thanks Jim, that is very encouraging! It is particularly helpful to see those threads with all the info and inspiration they contain, as well as the improvements that you yourself made between one and the next. Very smart looking results, too! I just skimmed them for now but will devote some quality reading and thinking time to them later in the week 
Thanks also for the thoughts on the 10" woofers; I guess it would make more sense, given the crossover points they'd work so well with the 6" drivers. I shall do some more simulations to look at what might be done with box volumes and linkwitz transforms etc. I'm perhaps more willing to compromise on SPL than when I designed the last 3-ways using them, so smaller boxes could be more viable and so make them usable in more situations.
Thanks again,
Kev
Thanks also for the thoughts on the 10" woofers; I guess it would make more sense, given the crossover points they'd work so well with the 6" drivers. I shall do some more simulations to look at what might be done with box volumes and linkwitz transforms etc. I'm perhaps more willing to compromise on SPL than when I designed the last 3-ways using them, so smaller boxes could be more viable and so make them usable in more situations.
Thanks again,
Kev
Well I spent the weekend reading threads and taking measurements, trying to decide on some of my basic design goals and doing some sanity checks on the drivers and enclosures. To help simplify things, I divided the possible future into three categories of increasing size and SPL:
So that seems to be working out; all very general so far but at least the plan looks to be on
For (current) tiny rooms: Taking some measurements in my small apartment, by an average of 70dB @1m things are getting loud enough to make it through the thin walls to the neighbors with some significance, so I shall call that adequate volume for the small 2-way when being used by itself. Though I shall want headroom for peaks - so lets call the minimum requirements as being 90dB @1m clean peaks.
WinISD suggests a single 15w8530k01 will reach 90dB within its linear range down to an -f6 of 31Hz; to me, that seems low enough response for such a modestly sized speaker. Natively it would want about 16L (if sealed), but with a Linkwitz Transform taking that down to 10L (just as an example) it would still only be using 5watts at peak, so power is not significant. In fact it could easily be made smaller if wished, and likely will.
For somewhat bigger rooms where the 26W8861T01 could also be employed: the 15W would be high-passed to only cover the mid-range and could then reach more dB than I'd ever personally want. So the 26W would become the limiting factor instead; WinISD suggests that within its linear travel, an -f6 of 27Hz is possible at around 100dB. Allowing for 20dB peaks that would mean 80dB average, which seems adequate to me for what would then be a smallish 3-way in a modest (UK) sized living room. Natively the 26W would want about 75L (sealed), but a LT and a 15L enclosure would still only demand about 45watts even at peak dB, so again power is not great and probably the box could be made smaller in practice.
If I ever wanted THX SPLs, and/or got a bigger room, and/or wanted lower frequencies: then some new sub-woofer could be introduced. If the 26W was high-passed even pretty low then it would go louder than I'd ever need, so the subwoofer(s) would be the limiting feature.
So that seems to be working out; all very general so far but at least the plan looks to be on
Though.. it has been a while and I'm rusty with WinISD; for the wattages mentioned above I was looking at the 'system input power' (which I believe for some reason ignores the extra power required due to DSP/filters etc). The 'apparent amplifier load power' is therefore higher when using a Linkwitz Transform, so in the 26W's example above, the power requirement actually peaks at about 210W.
Which might seem a lot for a nominal 170w driver, except that 20dB down at the average SPL it isn't even a quarter of a watt. So I don't think that significantly changes anything about the speakers - just places some demands on transient peaks that the amp must be able to power.
Which might seem a lot for a nominal 170w driver, except that 20dB down at the average SPL it isn't even a quarter of a watt. So I don't think that significantly changes anything about the speakers - just places some demands on transient peaks that the amp must be able to power.
Hi Kev,
Yes, you're right about WinISD's treatment of input power levels - your 5Watts for the 15W driver would be calculated before the effects of the Linkwitz Transform, so if you had say 12dB of gain in that filter then you're up to 80W total.
As you recognise though, music is dynamic (and most music doesn't have all it's power in the very LF regions where the Linkwitz Transfer would be active), so in the real world you may still be OK.
Yes, you're right about WinISD's treatment of input power levels - your 5Watts for the 15W driver would be calculated before the effects of the Linkwitz Transform, so if you had say 12dB of gain in that filter then you're up to 80W total.
As you recognise though, music is dynamic (and most music doesn't have all it's power in the very LF regions where the Linkwitz Transfer would be active), so in the real world you may still be OK.
Thanks for that, it fits with my (delayed!) understanding so that's good to hear!
So.. a few one simulations woth the LT suggest that at 35L sealed enclosures for the bigger woofer, it's excursion is the limit. But down at 15L we're starting to require hundreds of watts before Xmax can even be reached. So if we imagine a more normal hifi amp that can deliver say 150w peaks, then something like 25L or so is needed to allow reaching xmax. So that's still not too bad at all, thankfully.
So.. a few one simulations woth the LT suggest that at 35L sealed enclosures for the bigger woofer, it's excursion is the limit. But down at 15L we're starting to require hundreds of watts before Xmax can even be reached. So if we imagine a more normal hifi amp that can deliver say 150w peaks, then something like 25L or so is needed to allow reaching xmax. So that's still not too bad at all, thankfully.
Starting to think about the wider project now..
So, still quite early thoughts at this stage, more about the options I'm heading towards rather than firm decisions as yet. As always, I'd be interested in any thoughts people might have on these kinds of things.
Cheers,
Kev
I might end up getting into computer-based DSP for room correction, seeing as how the future rooms are unknown at this design stage. However, whilst DSP could handle crossovers etc too, my current inclination is to build actual circuits for the active crossovers, linkwitz transforms and baffle step compensation - the idea being that (for future flexibility/future-proofing) the speakers could stand alone if wished, without any external gizmos for their normal/basic function. Possibly that is less efficient though, to have both DSP and hard-wired active circuitry, so it isn't yet a firm decision.
Secondly, the system will use a power amp per driver. I may build a case comprising four amps for stereo tweeters and mids for when these are used alone as a 2-way speaker, and a second case comprising two amps for when the bass driver cabinets are also set up (to make them 3-ways). Somewhere I have several official ESP boards for Rod Elliott's P101 mosfet amps and also his P3a amps, and most of the components needed to build them. Though I could decide to order a couple of his P68 subwoofer amp boards for the bass drivers (and maybe also a couple more for any future subwoofer that i may make).
Or... I suppose alternatively just getting a 7.1 amplifier would be simpler. Though being a stereo man, I don't know much about them except that they look fairly costly, and I'm far from rich. Maybe if they included functionality like DSP it might be a more attractive bundle though, providing all the amps, crossovers, transforms etc etc needed in one package and making the two previous paragraphs redundant.
So, still quite early thoughts at this stage, more about the options I'm heading towards rather than firm decisions as yet. As always, I'd be interested in any thoughts people might have on these kinds of things.
Cheers,
Kev
Well a bit more modelling has helped make some decisions - picture of max/peak spl hopefully attached.
Which seems great to me, for small rooms, larger rooms and then bigger rooms or THX levels. In fact the 3-way even without a subwoofer looks more than fine to me for my music requirements, especially as this'll be a tri-amped active crossover setup, which has an easier job than a single amp. So, whilst I'll allow for a subwoofer in the design, I probably won't make one for quite some time, if ever.
Further good news is that on getting my old project stuff out of storage, there is much more of it than I remembered. Not just the P101 boards but lots of peripheral ESP ones like soft start and speaker protection and LR-crossovers, plus the main/costly semiconductors, 2x500va transformers and some big heatsinks. So I will at least build the P101 power amps for this project, even if upstream stuff like dsp/crossovers is still open to question.
Kev
# At a modest scale.. the 15w as the only bass driver (lower yellow line) seems good in 6L combined with a linkwitz transform. It is limited by its linear excursion. This would be the expectations for a small 2-way setup in small rooms.
# Then.. if the 26w bass driver were also employed in an additional 20L cabinet combined with a linkwitz transform (the lower blue line) to make it a 3-way, I get a little bit more bass extension but a lot greater spl available; again limited by the linear excursion. [The 15w driver crossed over at 150hz (or more) has no problem matching the 26w's spl (upper yellow line)]. This would be the expectation for a 3-way setup.
# Then.. if some future subwoofer were employed and the 26w crossed over at 60hz (upper blue line) lots more spl becomes available - but this time limited mainly by my amplifier's peak power. However the 15w is less sensitive, so that will dictate the maximum SPL of the system as a whole (upper yellow line). In theory my amp (esp P101 with a 2x 40Vac transformer) will peak at 150w so the results of that are shown. Though such is the law of diminishing returns that even if the amp only reached 100w the spl would only lose a couple of dB.
Which seems great to me, for small rooms, larger rooms and then bigger rooms or THX levels. In fact the 3-way even without a subwoofer looks more than fine to me for my music requirements, especially as this'll be a tri-amped active crossover setup, which has an easier job than a single amp. So, whilst I'll allow for a subwoofer in the design, I probably won't make one for quite some time, if ever.
Further good news is that on getting my old project stuff out of storage, there is much more of it than I remembered. Not just the P101 boards but lots of peripheral ESP ones like soft start and speaker protection and LR-crossovers, plus the main/costly semiconductors, 2x500va transformers and some big heatsinks. So I will at least build the P101 power amps for this project, even if upstream stuff like dsp/crossovers is still open to question.
Kev
Thinks Jim, thats reassuring. I have been through this kind of process before but it was quite a while ago and the goals are a bit different this time, so it has been helpful to work through some general things with winISD and get the basic intentions sorted out. I'll still need to consider many other aspects for the actual design of course, but this looks like the general plan, now. Exciting times
Just working on various shapes and sizes that could be employd to give the cabinet volumes needed. I'm currently heading towards this kind of thing:
So, it does seem that with a mixture of electronics and cabinet design the drivers can be wrestled into speakers of fairly modest size and reasonable aesthetics. Thats great news; I have never been very excited by the rather boring (if practical) boxes that I've made before, and so wanted something a little different this time.
Basically, a small 2-way that can be sat on top the 26W's bass cabinets if/when there is room for 3-way floor standers, which would then be about 1m tall in total. The sides are in lighter-coloured timber to act as a slight optical illusion, making them look narrower (the 2-ways look somewhat more squat and boxy otherwise).
I've seen the pointy-bottom idea somewhere on this forum before and really liked it; it seems to add a bit of elegance and lightness to an otherwise quite solid looking pyramidal shape. Wish I could find the thread to give credit, but if that was you then thanks!
I intend to have fabric covers on permanently, once finished, and might well use felt behind them to alter edge diffraction (since large round-overs would need extra baffle width). In fact I might design the front baffles accordingly, to avoid needing an extra frame for the fabric and the diffraction that these tend to cause.
The tweeter and midwoofer could be inverted, since the waveguide is quite wide. However I like the midwoofer at the bottom as it is then less equidistant from the vertical and horizontal edges, so will probably keep the tweeter on top - and maybe re-form the waveguide slightly to make it narrower.
The whole thing may only need to be about 8" deep which will suit small rooms and (if necessary) modestly sized shelves and window-sills etc. for the 2-way.
So, it does seem that with a mixture of electronics and cabinet design the drivers can be wrestled into speakers of fairly modest size and reasonable aesthetics. Thats great news; I have never been very excited by the rather boring (if practical) boxes that I've made before, and so wanted something a little different this time.