Page 1 of 6

#1 Lii Audio Speakers

Posted: Tue Dec 31, 2019 12:29 pm
by Ray P
I meant to post this previously and was reminded this morning that I hadn't. I thought there might some interest in Lii Audio's 'Flagship' ready to run loudspeaker design - it uses their 10in full range drive units like the Crystal and Fast 10s. They feature on the Lii Audio website but there are some nice pictures here;

https://www.decware.com/newsite/Liiaudio.html

There are drawings on the Lii Audio website. They provide most of the information you would need to DIY build them except for a decent drawing of the 'pear' shaped internal parts.

#2 Re: Lii Audio Speakers

Posted: Tue Dec 31, 2019 4:14 pm
by Wolfgang
I was playing with the idea to build "Hypercubes" but wasn't sure how I would be able to cut those angles precisely enough so that gluing the pieces together wouldn't end up with a complete mess. Building BLH looks like a walk at the beach compared to building the Hypercubes. But post #23 sounds very intriguing:

https://www.diyaudio.com/forums/full-ra ... ers-3.html

The Lii Audio Crystal 10 design seems to go in the same direction considering the part about cancelling of sound waves inside the box and not using damping material: "...he designed the shape to reflect everything to the exact center of the box where the waves become scalar which is to say cancel each other out." They are most likely not easier to build than the Hypercubes but are available as finished products. And they don't need a x-over and a bass unit like the Hypercubes.

Being addicted to SE OTL sound the disadvantage of the Lii drivers (only 8ohm available ) could be overcome by using a 8ohm (or even higher) resistor in series with the driver and get some nice current-drive side effect. It works and sounds a bit smoother in the mids (starting above 1kHz). With the InvOTL there wouldn't be any problem driving such a set up. Even with the 300B SE OTL I could get enough loudness with 8ohm DX4 Lowther drivers, 8 ohm resistors in series and reduced NFB.

Is there anybody who has enough background knowledge in order to compare the two "speaker principles" ?

#3 Re: Lii Audio Speakers

Posted: Tue Dec 31, 2019 5:26 pm
by Ray P
Wolfgang wrote: Tue Dec 31, 2019 4:14 pm They are most likely not easier to build than the Hypercubes...
I don't think the Lii speakers woud be that difficult to build if you were to use flexy-ply and similar materials. The shape looks challenging until you compare it with, say, a boat shape when, because the curve is only in one plane, it starts to look more simple. Still time-consming though.

Staying with the boat theme, an alternative approach would be to construct the curved sides as a cored laminate using something like end-grain balsa as the core. A single male plug would enable repeatable laminations to be made using vacuum bagging techniques.

If you can make the side panels the rest of the design looks fairly simple especially if the internal braces etc. were to be CNC routed.

#4 Re: Lii Audio Speakers

Posted: Tue Dec 31, 2019 5:51 pm
by Ray P
I'm pretty sure Colin (Toppsy) has build speakers with curved sides before so maybe he'll have some experiences to share?

#5 Re: Lii Audio Speakers

Posted: Tue Dec 31, 2019 5:57 pm
by Ray P
Other build techniques the curved shape would lend itself to include;
  • Build some shuttering and mould the shell in concrete or plaster
  • With a suitable autoclave you could heat form acrylic over a plug (actually you would probably need three plugs as you would need to shape it in stages

#6 Re: Lii Audio Speakers

Posted: Wed Jan 01, 2020 11:04 am
by Toppsy
Ray P wrote: Tue Dec 31, 2019 5:51 pm I'm pretty sure Colin (Toppsy) has build speakers with curved sides before so maybe he'll have some experiences to share?
Indeed Ray I tried two different construction techniques, bending 5mm bendy ply around a skeletal frame and translam built up CNC frames.
The build thread of both with positives and negatives on both methods is here: http://www.audio-talk.co.uk/phpBB3/view ... f=4&t=5729

#7 Re: Lii Audio Speakers

Posted: Wed Jan 01, 2020 1:31 pm
by ed
Ray P wrote: Tue Dec 31, 2019 12:29 pm
There are drawings on the Lii Audio website. They provide most of the information you would need to DIY build them except for a decent drawing of the 'pear' shaped internal parts.
I found ply pressing to be the best method for internals. Lay up layers of carcass veneer(very cheap) and clamp the layers in a press with glue.
plypress.jpg
and
plycurve.jpg
if doing the cabinet sides then strap layers of carcass over a skeleton frame like aircraft wing ribs. I can dig out some photos if interested.

#8 Re: Lii Audio Speakers

Posted: Wed Jan 01, 2020 3:29 pm
by Scottmoose
Which 'two speaker principles' are meant above?

Without accurate dimensions, it's a bit difficult assessing the box, but curved sides apart it appears to be akin to a classic bass reflex alignment (closest to, say, an Ultraflex or Onken). The curve to the rear will provide some of the desired length for the ducting; the curved sides will prevent a lateral standing wave in the c. 550Hz - 700Hz region depending on how wide the box would be if it were rectilinear and of the same height, depth & internal volume. If there are no internal structures to prevent it, there will still be a vertical standing wave between the top / base panels: given the comments on the Decware site about scalar cancelling it seems improbable that there is anything to disrupt it, since that would then be in the way of the lateral reflections. Generally most large vented boxes are not stuffed full of fibreglass either (as Steve Deckert himsef knows perfectly well, having designed enough of them) and resonances in the aforementioned region have a rather short wavelength and are quite easy to absorb with very modest amounts of lagging material. The less you need the more acoustically efficient your design is of course, and I doubt anyone would argue about that, but there is a limit before rationality kicks in... ;) A rough stab at the alignment suggests it should be reasonable though, so not a bad product for anyone in the market for something like this if shipping costs are reasonable. Good luck to them, I say. :)

I note parts of the published electromechanical T/S data on the Crystal 10 contradicts itself, so presumably they've averaged each particular value from a batch, without trying to make a consistent data-set. The most obvious / significant is Qts: if Qe = 0.48 and Qm = 7.78, Qts cannot possibly be the stated 0.51, but 0.452. Vas / B*L / Mms also appear inconsistent, presumably for the same reason. No criticism, but a point to keep in mind when working with the information.

Re adding an 8ohm series resistor, it will also raise effective Qts from 0.452 to almost 1.0 (0.9965 if you want to be really pedantic) which has a concomitant effect on box alignment: it'll be peaking significantly above tuning followed by a drop in output due to it wanting more box volume than is available. May be liked, but it's not ideal either. Can be useful in other loads e.g. a baffle though to compensate for the cancellation if it's suitably sized for one to balance off against the other. YMMV as always.

#9 Re: Lii Audio Speakers

Posted: Wed Jan 01, 2020 4:21 pm
by Ray P
Scottmoose wrote: Wed Jan 01, 2020 3:29 pm Which 'two speaker principles' are meant above?
Thanks for the useful input Scott, though I'm not sure who the question is for?

TBH, I considered it to be a variation on the usual bass reflex box and I didn't buy any of the stuff about cancelling standing waves as my basic O level physics tells me that a fixed shape isn't going to do that across the audio spectrum.

My rationale for starting the thread was simply because, to me, the speaker is attractive, a little different and looks like a potentially good implementation of the Lii drive units. There are other designs for the units on the Lii website, IIRC (I didn't pay that much attention to the other rawings), there's a folded Voight pipe and a bass reflex box.

#10 Re: Lii Audio Speakers

Posted: Wed Jan 01, 2020 4:23 pm
by Ray P
ed wrote: Wed Jan 01, 2020 1:31 pm I can dig out some photos if interested.
It's alway interesting to learn how people problem solve Ed so, yes do share...

#11 Re: Lii Audio Speakers

Posted: Wed Jan 01, 2020 4:46 pm
by Scottmoose
It was in reference to Wolfgang's comment in his post above.

Right, cancellation of all eigenmodes isn't going to happen, lovely notion though it is. ;) Steve Deckert's prose can get a bit OTT on occasion, though he does sometimes have some interesting ideas. Either way, it looks to me like it has the potential to be a nice classical reflex variation for these new take on Corals, so all (probably!) good from my perspective, within the context of what it is. :)

#12 Re: Lii Audio Speakers

Posted: Wed Jan 01, 2020 6:12 pm
by Wolfgang
I considered the Hypercubes a "sealed" box and the Lii speakers "vented", referring to this as "two principles", but I cannot understand from the drawing how the Lii speaker would work exactly.
official1.jpg
It doesn't look like a typical bass reflex design.
Re adding an 8ohm series resistor, it will also raise effective Qts from 0.452 to almost 1.0 (0.9965 if you want to be really pedantic) which has a concomitant effect on box alignment: it'll be peaking significantly above tuning followed by a drop in output due to it wanting more box volume than is available. May be liked, but it's not ideal either.
I was refferring to E.Merilaeinen, Current Driving Of Loudspeakers, p. 182ff, where he describes how to use a voltage drive amp partly in current drive operation. I tried it with BLH/Lowther DX4/PM6A and it works and it sounds actually really nice. It obviously helps to weaken the driver's EMF currents (and does a couple of other things too like changing Qts) which is quite noticeable.

#13 Re: Lii Audio Speakers

Posted: Wed Jan 01, 2020 6:29 pm
by Ray P
Scottmoose wrote: Wed Jan 01, 2020 4:46 pm It was in reference to Wolfgang's comment in his post above.
Ah! I thought I had skimmed the previous content but obviously not very well!

#14 Re: Lii Audio Speakers

Posted: Wed Jan 01, 2020 9:06 pm
by Ray P
Toppsy wrote: Wed Jan 01, 2020 11:04 am Indeed Ray I tried two different construction techniques, bending 5mm bendy ply around a skeletal frame and translam built up CNC frames.
The build thread of both with positives and negatives on both methods is here: http://www.audio-talk.co.uk/phpBB3/view ... f=4&t=5729
Thanks Colin, I enjoyed revisiting that thread. You seem to have concluded that the stacked layer approach worked best for you.

Just out of interest, do you have any idea what the tightest radius 5mm bendy/flexy ply will conform to?

#15 Re: Lii Audio Speakers

Posted: Wed Jan 01, 2020 10:04 pm
by Scottmoose
Wolfgang wrote: Wed Jan 01, 2020 6:12 pm I considered the Hypercubes a "sealed" box and the Lii speakers "vented", referring to this as "two principles",
Ah, got you. Yes, that's precisely what they both are.
but I cannot understand from the drawing how the Lii speaker would work exactly.
It doesn't look like a typical bass reflex design.
I hadn't seen that drawing, many thanks for the heads-up. :) Well, it does (look like a typical [traditional] bass reflex) to me, but perhaps I'm more used to working with them. You've got the main volume, with the structural ribs drawn in lilac, and the rectangular [relatively large cross section] vents located on the rear panel. By squeezing it out like a bottle-neck at the back, they've got the duct-length they need. That's pretty much all there is to it in operating terms; based on that drawing & the others, it's a fairly simple box as far as that goes. Note the lack of anything to break up the vertical eigenmode, so I'm afraid the Decware comments as far as vertical modes go are 'mistaken'. I'll reserve judgement about the horizonal as I don't have time to look at it in more detail; I suspect Steve (Deckert) is waxing lyrical a bit in his enthusiasm for the Lii product; fair enough, it's nice that he's helping give them a bit of publicity. No doubt it helps generates a bit of traffic for him too, so good for everyone, including consumers after a product of this type.
I was refferring to E.Merilaeinen, Current Driving Of Loudspeakers, p. 182ff, where he describes how to use a voltage drive amp partly in current drive operation. I tried it with BLH/Lowther DX4/PM6A and it works and it sounds actually really nice. It obviously helps to weaken the driver's EMF currents (and does a couple of other things too like changing Qts) which is quite noticeable.
I can well believe it, but whether it will be an improvement varies on a case by case basis. Most of my Fostex & similar designs for e.g. assume they are to be used with an amplifier of xyz output impedance because that's what the drivers were intended to be used with (or equivalent compensation via a series power resistor / high resistance speaker wire). But ideally you factor the altered effective Q into the enclosure design from the get-go as the alignment can otherwise change very rapidly away from what may have been intended. In the absence of that, it's a case of luck-of-the-draw; you may get results you find positive, you may not. I've just done a box series for an upcoming Markaudio driver with variations for various amplifier output impedances (or series resistance) from very low output impedance voltage-source up to about 6ohms. Ignoring other factors, box volume requirements rockets from about 35 litres out to 120 litres, as you'd expect from a Q that had roughly doubled.