'LGT' Construction Diary

[wigginjs]

Also, just thought about this. How is this thing not going to tip over?

[ShinOBIWAN]

Originally posted by HBarske
Great design.
Build it. Put a 50000$ price tag on it. I bet you'll be successful.

Originally posted by speaker


That's not marketing, it is great industrial design.

Love the first set of images.

Wow!

Originally posted by edjosh23
Very nice!

I really like Tarantism, great word. I'm sure it will fit well with those speakers.

Josh

Thanks all, appreciate the sentiments

[ShinOBIWAN]

Originally posted by m0tion
Awesome plan. Interesting driver choice. Two questions:

1) Do you have a plan to deal with the very poor vertical off-axis response of the tweeter or do you not view this as a problem?

Ta and hi Motion.

I do still view it as problem but the RAAL despite being a fairly long line source still offers OK vertical dispertion in comparison to ribbons of a similar sizes thanks to it concentrating the higher frequencies into a narrowing radiating area. I don't have the drivers yet but looking at the manufacturers data appears to show a smooth off axis response in both the horizontal and vertical which is definitely important. The RAAL does have good horizontal performance too.

The MTM will allow me to control the dispertion of the mids around the 1.5Khz XO point and provide a better match than could be afforded just from a simply mid-tweeter arrangement.

Both the 4" Scan mid and the RAAL both have an extended and flat response so crossing will be made that bit tidier and produce good acoustic slopes without the need to start shoring things up with the digital route.

2) Why passive xover? You seemed to be one of the largest proponents of digital active crossovers. Does this mean you will also not be implementing room correction with these speakers? I was under the impression you were also a big proponent of room correction.

I brushed on it above but the Perceive 2 really forced the need for an active solution because I was running the mid right to the limit of its operating range and also had one of the nastiest FR you'd ever likely to see. So to get the best from that you needed 4 notch filters and 48db+ slopes. After messing around with both passive and active it was clear things benefitted from the digital trickery.

From all accounts this next project looks to imminently suitable for a passive network and a fairly simple one at that. I'd also like to have a bit of change this time. Passive filters don't crash, suffer from latency issues or require any maintanence once setup Things might change if it turns out the active route once again is very definitely superior.

I wouldn't be without the room correction though and this can still be implemented when the PC is acting as the source regardless of passive speakers.

3) Since the choice of 8" woofers limits the lower frequency response of these speakers do you plan to use them with a subwoofer? Any ideas for that design and what you would use as a xover?

Yes these will be good to 30-35hz so a sub will still be essential for HT, but not for music. I simply couldn't get away with a design that included hefty and capable subs yet still remain fairly compact. I did look at that idea though.

[ShinOBIWAN]

Originally posted by m0tion
Also, just thought about this. How is this thing not going to tip over?

Not sure what you mean... well I do but I don't think it will be problem - remember the concern about the leg that held up the Perceive's?

In the end it turned out that the only danger of cabinets falling over came from them being placed on flimsy tables

I think sheer mass just makes things tough to move unless you really do something daft. It worked last time so fingers crossed anyway.

[wigginjs]

RAAL despite being a fairly long line source still offers OK vertical dispertion in comparison to ribbons of a similar sizes

This might be true, I did a little math though just to get an idea for what the response might look like. At 10 degrees off vertical axis the response is down ~8dB@20KHz. Now, maybe it's cause my mommy didn't breast feed me long enough, but I'm 24 years old and I can't hear 20KHz tones. So lets look at a frequency I can definitely hear, 10KHz. At 10KHz the response is down ~4dB 10 degrees off vertical axis. Thats definitely noticeable, but maybe your listening position isn't a full 10 degrees off axis so you won't get all of that effect.

I sit about 11 feet away from my stereo speakers, so for me 10 degrees off vertical axis is +/- ~2ft assuming my ear height and the height of the tweeter are equal. So, at 11 feet away 5 degrees off axis, probably a very "safe" bet as far as frequency response goes is +/- ~1ft. Not too bad at all, although I think most people listen a little closer to their speakers than I do. So, lets look at an 8 foot listening distance. At 8 feet you'd get +/- 1.4ft for 10 degrees and +/- 0.7ft for 5 degrees off axis.

It appears that these probably won't be speakers you'll be able to walk around the room and still enjoy similar quality to your listening position, but maybe that isn't a goal for you. Also, one other thing I noticed is that the tweeter height on these speakers is significantly lower than the tweeter height on the Perceive V2's, any reason for that?

[Tenson]

Nice design Ant!

I think you could get away with the 70-10 ribbon and Xover at 3KHz. That should, I assume, have a better vertical dispersion being smaller.

[ShinOBIWAN]

Originally posted by m0tion


This might be true, I did a little math though just to get an idea for what the response might look like. At 10 degrees off vertical axis the response is down ~8dB@20KHz. Now, maybe it's cause my mommy didn't breast feed me long enough, but I'm 24 years old and I can't hear 20KHz tones. So lets look at a frequency I can definitely hear, 10KHz. At 10KHz the response is down ~4dB 10 degrees off vertical axis. Thats definitely noticeable, but maybe your listening position isn't a full 10 degrees off axis so you won't get all of that effect.

I sit about 11 feet away from my stereo speakers, so for me 10 degrees off vertical axis is +/- ~2ft assuming my ear height and the height of the tweeter are equal. So, at 11 feet away 5 degrees off axis, probably a very "safe" bet as far as frequency response goes is +/- ~1ft. Not too bad at all, although I think most people listen a little closer to their speakers than I do. So, lets look at an 8 foot listening distance. At 8 feet you'd get +/- 1.4ft for 10 degrees and +/- 0.7ft for 5 degrees off axis.

It appears that these probably won't be speakers you'll be able to walk around the room and still enjoy similar quality to your listening position, but maybe that isn't a goal for you. Also, one other thing I noticed is that the tweeter height on these speakers is significantly lower than the tweeter height on the Perceive V2's, any reason for that?

In terms of listening position and will this be a problem, the answer is no. I always listen in virtually on axis vertically but do like to have a wide horizontal dispertion so there's not just one person enjoying the treble.

For me the issue could be power response, I'll certainly have a very nice one in the horizontal but the vertical will be compromised - how important is this? Well less important than other attributes that I've placed above and over this consideration, in other words I believe its better live with this compromise and reap the greater benefits.

Power response is important IMO though. It dictates more about the overall sound than is apparent at first. Offaxis sounds eventually reach your ear at some point, albeit reduced in amplitude and time delayed. But what this contributes is a reverberant sound field much like you'd experience in any venue. How smooth this offaxis frequency response created by the speaker dictates similarly how smooth the reverberant field is. Its fine having a speaker thats either wide or narrow dispertion but make sure that the offaxis performance is at least smooth. Personally I prefer a wide dispertion characteristic with smooth offaxis performance, the Perceives gave me that with the use of 3" mid domes. This wide dispertion tends to make the speakers seem to disappear more effectively, you can't definitely hear particular sounds coming from them but rather its from a larger soundstage. Narrow dispertion can tend to sound like its constantly reminding of you that sound is coming from the front of the room from two specific spots but they tend to image very well because the time delayed reflections are reduced.

Its really about picking a compromise because the two on one extreme you have omni and on the other you have an 18" driver used for a mid

So overall, the RAAL and Scan crossed at 1.5Khz will offer a smooth response on and off axis to +/- 60degrees horizontal and vertically its about +/-20 degrees. This is going on datasheets which could well hang me So its wide horizontal dispertion and fairly narrow vertical the important thing is that the overall power response is smooth.

BTW: The tweeter is lower on this one because that's where it needs to be when sitting.

[wigginjs]

It seems that RAAL provides foam triangles that limit the vertical dispersion of their tweeter so that you won't have to worry so much about getting really poor quality reflections from your ceiling and floor, but one thing that confuses me is that the foam triangles appear to cover up much of the actual ribbon! Why would that be? Maybe I'm misunderstanding the function of the foam.


Side note:

Would you mind editing your drawing to show metrics for size and perhaps a side view? Sorry for the bother, but I'm very interested in this project.

[LineSource]

Originally posted by m0tion
It seems that RAAL provides foam triangles that limit the vertical dispersion of their tweeter, but one thing that confuses me is that the foam triangles appear to cover up much of the actual ribbon! Why would that be? Maybe I'm misunderstanding the function of the foam.

True Ribbon tweeters have a rigid clamp at the top and bottom of the ribbon foil which creates non-uniform vibrations plus reflections for the first few mm's or so of the ribbon. The RAAL foam may be positioned to absorb these vibrations so the more uniform center portion of the ribbon is the main contributor to the sound dispersion. I have run 3D finite element ribbon simulations that show these effects.

I have only done 2 WMTMW designs. I have not been able to get a ribbon tweeter to generate the correct dispersion in an WMTMW topology. I would use a dome tweeter, crossed to 5" ScanSpeak Revelators, crossed to 12" treated paper woofers, probably with 3rd order Xovers.

[Tenson]

Actually, I think they use the foam to change the phase of the sound and improve vertical dispersion limiting destructive interference.