PERSPEX QUASAR ver 2.0 prototype build

[Greg]

- it doesn't work quite so well as a conventional two way in the Quasar as the gap between the two drivers becomes noticeable and distracting...

Now that is interesting. That was exactly what I thought when I listened to your original Quasar’s. I could readily hear a break between the main driver and the bass helper. I concluded that it might have been down to the unsuitability of the Eggfest venue or that I might have been too close to the drivers, but without doubt, I detected separation which I didn’t like.

[Paul Barker]

I didn’t notice that effect at the time Greg, but I never got on with my attempts at home Greg hence why I made the stupid decision to sell the Supeavox’s to Mark, who also sold them on. Wish I had them back!

But to me the Quasars were totally magical or poor. I never got totally magical at home. Might be room effect.

[chris661]

Forgot to say that the orange curve looks familiar Thank you for showing it here. What do you think caused the two-way curve to roll off earlier? Some cancellation effect in the room from driver interaction or room mode?

The GMF285 is a lovely sounding driver in its own right and makes a great conventional two way with the Supravox tweeter that can be used from 2KHz - it doesn't work quite so well as a conventional two way in the Quasar as the gap between the two drivers becomes noticeable and distracting... Having said that I must try the small sonido with the GMFs in my Quasar once the baffles are out of store...

Glad the orange curve looks familiar - it suggests my measurements aren't too far off-base

The next avenue of exploration does involve running the Supravox higher. I have some large-format compression drivers that'll get well below 1kHz, so we're going to bring those into play to see what's what. In terms of a high-efficiency way of producing 800Hz-20kHz, a good compression driver should work well.
Given adequate power (300w per driver), they did this (one per side, covering 1.2kHz and upwards) -
At home HiFi levels, distortion is crazy-low. Something like 0.05% THD.

Greg does make a good point, though, that the vertical separation will likely be a problem. I'm hoping that crossing them over lower than 2kHz will rectify the audible problem (and I'm not surprised that it is audible, although the sensitivity of individual listeners will likely vary), but if not then it's likely back to trying to find a full-range driver to do the job.

To answer the question of near-field vs in-room LF response, I think the low-frequency drop is too broad-band to simply be a cancellation. I've gone back over the graphs, and there's a close vs ~1m vs ~2m comparison which can be made. The close-mic'd one gives a gentle rolloff below 70Hz, and as the mic moves away, there's more of a 70Hz peak, and less below 70Hz. Now, I suspect the 70Hz peak is a room interaction, but the declining output below that feels to me like dipole losses.

FWIW, the same mic, USB soundcard and software reported this:


At my listening position at home (similar-sized room), with the closed-box loudspeakers that Colin heard the other day.
So, I don't think the measuring gear is at fault. There's a slim chance that the particular mic locations I measured at showed wide-band cancellation below 70Hz, but IMO it's the dipole losses that are most likely to be the cause of the lack of low bass.


I'll be going back to Colin's for further experiments in the coming weeks, and will try taking LF response measurements in 10+ locations just to make sure the room's not doing something weird. At this point, I can't say with certainty whether it's the room or the speaker, but I can say the (limited) evidence suggests it's the speaker.

Chris

[IslandPink]

Forgot to say that the orange curve looks familiar Thank you for showing it here. What do you think caused the two-way curve to roll off earlier? Some cancellation effect in the room from driver interaction or room mode?

Just thinking about this, the 'solo' Lii response curve extends a fair bit lower than an AER would - I think because the Qts is higher ? It's quite bassy without a helper.
What was the crossover ? - it's likely going to need a smaller cap than the crossover Nick was using with the FE167E. Otherwise you might get a cancellation of the bass between the two drivers.

[Toppsy]

What was the crossover ? - it's likely going to need a smaller cap than the crossover Nick was using with the FE167E. Otherwise you might get a cancellation of the bass between the two drivers.

Mark,
this is the XO used:


When Scott designed the filter he had very limited published data on both the drivers (more on the Supravox than the Fast-10) and what was published by Lii Audio was suspect at the best. Hence in my early posts on this thread I did quantify this fact that the XO used is likely not to be optimised for the driver combination. Though Scott did design a standard parallel XO to mimic the series one we have not used this since James designed the Quasars for a series filter where both driver effective 'see' each other via the series filter. That said changing the cap to a smaller value will also necessitate change in value to the inductor so a completely new XO based on Chris's actual measures of the drivers would be required. This however would still not fix the issues with the Fast-10 driver. Hence our search for a better performing replacement driver to mate to the Supravox.

[Ray P]

How many hours have the Lii Audio units clocked up?

[IslandPink]

That crossover looks a long way off the mark. The inductor provides the low pass on the Supravox, usually it's in the range 8.2 to 12mH ( it was 12 or even 15mH on James's original ) . This is to get the crossover down around 150 to 200Hz. If you have only 2.2 , the LP is going to be over 600Hz.
Then the cap provides the high-pass on the midrange driver. With the Lii's natural roll-off being fairly low, a 100uF is not going to put the HP at much more than -3dB at 120Hz . There will still be some bass coming from the Lii, out of phase with the Supravox.

[Toppsy]

Around 5-6 hours whilst in my possession Ray. That would be around 2-hours ran solo without any filtering I used my P6 amps with a speaker work-out CD that Chris (stratmangler) gave me that is recorded specific for running in new drivers. The cones certainly moved with that CD. After that coupled to the Supravox through the above posted filter. Can't say for how long they were ran before that.

I know Nick has emailed the manufacturer about the measured performance we got but not sure he has had a response back from them yet. However Lii Audio do not specify anywhere on their web site about the need for a lengthy run-in time for any their drivers. Perhaps they should?

[Scottmoose]

At the risk of shoving my unwelcome nose very briefly into proceedings, if I may, series crossovers do not work like parallel filters, Mark. The drivers and filter are inherently locked together, and the low pass filter for the Supravox is in fact formed by the capacitor, not the inductor, while the high pass is formed by the inductor, not the capacitor. That's how a basic series filter works. If anybody is interested, 1st order series electrically speaking (please note that major caveat), also have the peculiar characteristic of summing flat, or being able to sum flat on a consistent basis. Assuming Butterworth (Zeta = 1.0) then this will be in either polarity. Changes in the Zeta function will alter both the individual rolloff characteristic and filter frequency, but it will theoretically continue to sum flat, albeit, since you also alter phase by doing so, only in one, rather than both, electrical polarities. Whether any of this applies acoustically is a different matter entirely as the inherent response & rolloff characteristic of the driver, its VC inductance & the physical spacing between the drive units in the X, Y and Z [i.e. front-rear] axis all come into play.

The aforementioned filter was always anticipated by me to be more or less junk acoustically speaking, because we had essentially nothing to go on. It was a lash-up to get things going, no more. Supravox's published data isn't exactly distinguished when you look at it (lots of things missing), excellent though the drivers can be, while Lii Audio's is fiction, as was obvious from the get-go, with that ridiculous FR plot they print. You can get a very loose idea of what it might be sans so much smoothing & on a smaller amplitude scale (Chris's is on about 40dB) but inference is ultimately inference; some wideband units are better than others & can be closer to the rather linear claims, but the only way you can know is to measure them and actually find out, as Chris has been kind enough to do. Unfortunately, I am not in a position to be able to do that, greatly to Colin's understandable annoyance, but I'm obliged to consider the physical and mental health of my family, and not to put too fine a point on it, also keep a roof over my head.

Trying to find a relatively linear (and relatively sensitive) wideband is the holy grail of wideband design; realistically speaking the WE 755 is still one of the best, and since the pioneers were happy with about 12KHz for understandable reasons, you need a supertweeter if you want that HF 'air'. The Fostex FE208ESigma is better than some; the silver coil Lowther PM6A is reasonably reasonable from that perspective also. The question is how much linear and non-linear distortion you're willing to accept in order to obtain whatever you want elsewhere e.g. sensitivity, lack of filter > 200Hz or whatever. As always, name your poison. We all want different things. From Chris's measures of the units in question, the Supravox is clearly a solid unit -no surprises there, even if the manufacturer's data is a bit dodgy. The Lii driver, even accounting for the unflattering graph scale, clearly 'has issues' -sufficient that I would want a different unit. As noted, that's just me.

[Toppsy]

Unfortunately, I am not in a position to be able to do that, greatly to Colin's understandable annoyance, but I'm obliged to consider the physical and mental health of my family, and not to put too fine a point on it, also keep a roof over my head.

Scott, many thanks for your input, which are wise words as always. No I am not annoyed with you as you have put it. I do understand your situation though do not fully agree with you. Notwithstanding our disagreement I'll give you a call tomorrow and we can have a good old chat and hopefully put the misunderstanding all behind us.

[Ray P]

Around 5-6 hours whilst in my possession Ray. That would be around 2-hours ran solo without any filtering I used my P6 amps with a speaker work-out CD that Chris (stratmangler) gave me that is recorded specific for running in new drivers. The cones certainly moved with that CD. After that coupled to the Supravox through the above posted filter. Can't say for how long they were ran before that.

I know Nick has emailed the manufacturer about the measured performance we got but not sure he has had a response back from them yet. However Lii Audio do not specify anywhere on their web site about the need for a lengthy run-in time for any their drivers. Perhaps they should?

Thanks Colin. IIRC the guys who posted about them over on the lenco heaven site mentioned that they sounded better after a period of running. I also seem to recall someone measuring them and commenting about smoothness. I'm following this with curiousity as they've been subjectively well received elsewhere.

Perhaps I should try and find the time to but a quick and dirty speaker together and give mine an outing. I'm pretty sure there's a drawing for an easy TQW design over on LH. Perhaps I'll head over to LH later.

[Greg]

Hopefully not off topic, but just a general observation. My WAD KLS3 (improved) speakers are more than 25 years old. Although they have been subject to upgrade periodically, essentially, the bass and mid drivers are as original, so also 25 years old. With the regular use they get, one, could say they are well run in. When I sit down to listen, they sound fine. However, occasionally, I have them on for a much longer time than usual, like today when I was listening at considerable volume for beyond six hours. Without doubt, the last two hours of the session had the best speaker performance, particularly noticeable on bass punch (Hit you in the chest) on drums. It does beg the question, when is a speaker run in, and when it is, how long does it take in a session for it to sound at it’s best.

[ed]

However, occasionally, I have them on for a much longer time than usual, like today when I was listening at considerable volume for beyond six hours. Without doubt, the last two hours of the session had the best speaker performance, particularly noticeable on bass punch (Hit you in the chest) on drums. It does beg the question, when is a speaker run in, and when it is, how long does it take in a session for it to sound at it’s best.

it can also 'beg the question' has our perception changed after listening for so long and maybe the speaker hasn't changed at all?

[Scottmoose]

Certainly can happen. But there are physical mechanisms too. The extent depends on the driver design, but voice coil heating causes temporary changes in the electromechanical damping; this can also affect the internal box temperatures and that + dynamic load conditions will also affect suspension compliance. Drivers with vented coil formers, venting to the basket under the spider etc. tend to be less affected, but still see changes. A set of fully broken in drivers I recently measured shifted by about 5% in Fs, Qt & Vas after about an hour of running on ordinary programme material, which would definitely cause an audible difference to the cabinet alignment. Not a vast difference, but it would be there.

[ed]

now I'm having one of those 'Brain, I wish you'd stop following all these mad paths' days.

So, what do we do with the knowledge that:

A pair of speakers could improve after some hours of listening.
or
We know that if we listen for a length of time our appreciation of the music may change, either because the equipment has changed or our brain/perception has changed.

How do we measure such variables and ultimately decide which piece of equipment is best for us. Or, dare I say it, better objectively.

Is it possible to compare different equipment with such variables in place?

I feel a song coming on ....'hifi is a nightmare...yeah'