Aussie specific newb questions

[fb]

So I built myself a Peerless XLS12 sub a few months ago, and I'm very happy with it. So I'm going to upgrade my bookshelves.

First q: The tools I used for my sub are no longer available to me. What are the basic requirements for high quality work with MDF (ignoring finish for now). I figure plunge router, table or circular saw and clamps? I already have an adequate drill. I'm hoping someone can recommend specific products available in Australia.

Second q: Which design? I'm thinking of either Zaph's Seas All-metal or XG18 MTM. I'm not worried about the low end, it seems quite well covered now. Considerations include sensitivity and the type of music I listen to.

Every now and again I like to crank it. My current speakers are rated at 88dB and are quite acceptable, however I'd like to leave my amp a little more headroom at the high end if possible.

The music I listen to is often very 'busy', from psytrance (yes it's very different to 'dance' music ) to metal, I often find everything sounding cramped together.

Will either of these designs perform better? I've also considered open baffle, but figure it would fail for sensitivity, size and room placement reasons. Any better designs with parts available in Aus?

Thanks

[mike.e]

Ive done the same thing-started with subwoofers and then craved better everything else.

http://www.advanceae.com.au/catalog/ Is local to you.

The 'cramped' sound output could well be a combination of high power transistor amplifiers with alot of NFB,combined with loudspeakers acting nonlinearly(Ie you should get a system with higher sensitivity)

Youl notice that many 2/3way cone driver systems seem stuck at 83-90db/1watt. This is because people dont enjoy massive enclosures with no bass output,but in your case you could use a more sensitive mid unit,which happens to roll off right where your subwoofer operates.

Are you actually looking for an existing DIY design or are you looking to fiddle about with xovers yourself?

Heres what one Oz guy did
http://www.geocities.com/adrian_mack/homepage.html

[fb]

My amp is the Panasonic SA-XR50 which seems to be fairly highly regarded, so I doubt that is the issue. Maybe the midranges on my bookshelves are struggling when there is too much low bass signal? I might try implementing a high pass filter one day.

I would prefer an existing design at this stage, as measuring equipment is another added cost and complexity. Are there any designs you could suggest that have this more sensitive midrange?

Thanks for your help

[rcw]

I wrote an article that appears on Rod Elliots website that shows that the typical small satellite is driven into quite severe non linearity by the standard 90dbm. average with 20db. headroom specification regarded as the minimum for hi fi.
The best bang for buck solution is to use a filter assisted QB5 alignment as this can typically increase the level at which distortion becomes objectionable by around 12db.
If you have a subwoofer that has a typical second order output at around 80Hz. for the satellites, then this is easily implimented by just a suitable box tuning.

[Bill Hobba]

Originally posted by rcw
I wrote an article that appears on Rod Elliots website that shows that the typical small satellite is driven into quite severe non linearity by the standard 90dbm. average with 20db. headroom specification regarded as the minimum for hi fi.
The best bang for buck solution is to use a filter assisted QB5 alignment as this can typically increase the level at which distortion becomes objectionable by around 12db.
If you have a subwoofer that has a typical second order output at around 80Hz. for the satellites, then this is easily implimented by just a suitable box tuning.

Hi RCW

Read your article. Very interesting and informative. As a proponent of closed box systems there are a few points I would like to make. From experience, and from other articles I have read, I find neighbors tend to complain if listening levels rise above 88db. And for a stereo pair this means each speaker is delivering 85db. Thus I think a much more meaningful level is 85 db not 90db. I tend to do most of my listening at 80 to 85 db - at least that is when I actually put a sound level meter on it. You also mention an excursion of 16mm at 78hz. I have had a look at the maximum excursion of some modern drivers like the Scan Speak revelator and the Seas magnesium drivers - they have excursions well in that - eg for the seas it was 19mm.

Personally I believe the future is closed box actively compensated systems using devices like the DEQX once its price comes down.

Thanks
Bill

[rcw]

I would be very surprised if that driver has a linear peak excursion of that sort, the article mentions peak excursion not peak to peak.
I would say that anything that can reduce the overall amount of distortion that occurs in the midrange must be a good thing, and the scheme outlined also substantially reduces the various forms of modulation distortion you get when a higher frequency signal rides on a large amplitude lower one.
At some future time dsp will no doubt be able to remove much of the artefacts mentioned, but to my mind it is something like the story of the man who built a car with square wheels, and then designed a very complex and sofisticated suspension system to give it a smooth ride, a great technical achievment, but why bother, when you can get the same effect by just making the wheels round in the first place.

[Bill Hobba]

Originally posted by rcw
I would be very surprised if that driver has a linear peak excursion of that sort, the article mentions peak excursion not peak to peak.

Agreed - its linear excursion is 10mm. I was thinking of the much worse distortion when it bumps against its mechanical stops - which probably is not too good for the driver either.

Originally posted by rcw
At some future time dsp will no doubt be able to remove much of the artefacts mentioned, but to my mind it is something like the story of the man who built a car with square wheels, and then designed a very complex and sofisticated suspension system to give it a smooth ride, a great technical achievment, but why bother, when you can get the same effect by just making the wheels round in the first place.

I was not actually thinking of that. I was thinking of correcting for frequency and phase errors. The transient performance of a vented system can never be corrected (resonant systems must ring) - frequency and phase errors can. Also they have rather steep slopes so it is quite feasible to remove the driveing signal below say 100hz and have that handled by the subwoffer.

Thanks
Bill

[Bill Hobba]

Originally posted by rcw
I would be very surprised if that driver has a linear peak excursion of that sort, the article mentions peak excursion not peak to peak.

I just wanted to add that I double checked - it is PP not peak. The revelator is a bit better - it has a 6.5mm linear peak excursion. So for 110db peak at 189hz you would indeed be getting close to the maximum and over it and into possible mechanical damage at 78 hz. And indeed the maximum recommended power for the revelator is 60W which gives a maximum output of about 105 db for its 87 db sensitivity - or allowing 20db headroom a maximum listening level of 85db. That is for the 7inch revelator - the 5 inch one has a sensitivity of 84.5db or a maximum listening level of 82.5 db. If you wish to stick to a closed box the only remedy I see is listening at lower volume levels or using multiple drivers. The 85db max I listen at would have a maximum excursion of about 1/4 of that at 78hz ie 4mm. Although I prefer a closed box; driving your speakers to possible destruction is much worse IMHO. And if you are using a device like the DEQX that actively corrects for bass roll off you are in an even worse situation.

Thanks
Bill

[fb]

Can the high pass for the QB5 alignment be used in conjunction with the already designed crossover? I'd keep the baffle size the same and modify the box depth and port tuning to modify an existing design?

Thanks

[rcw]

Yes FB, you just need to modify the box and port to give the required peak.
Some plate amps have a fixed high pass output with only the bass high pass variable, in this case you can compliance scale the qb5 alignment to suite the fixed high pass filter, details of this are in the compliance scaling article.