Hi Guys,

Well rather then add to my bloated RTA thread, I thought I'd post some data here that I recorded tonight from the Aura NS3's.

My first impressions of the speakers are wow, they are so cute. Like a really little midbass/sub.
Click to view attachment
Click to view attachment

The driver is classified as a 3inch extended/fullrange driver, specs can be found here...

http://www.madisound.com/catalog/PDF/auras.../NS3-193-8A.pdf

Aluminum cone
Underhung voice coil
NRT neodymium magnet system
Inherently shielded
Rubber surround
Specs :-
Fs 80 Hz
Nominal Imp. 8 ohms
Power (RMS) 20 W
Re 7.6 ohms
Sens. 1W/1m 80 dB
X-max peak 3.1 mm
Qms 8.0
Qes 0.73
Qts 0.67
Vas 1.25 ltrs
BL 4.7 TM
Mms 4.3 gr
Cms 920 mM/N
Sd 31 cm2
Net weight 216 g
Max Exc. p2p 20 mm
VC diameter 19.3 mm

I ran some sweeps and tested in ARTA via the soundcard output and through a small Coustic AMP 207 amplifier.

The response curves don't look too bad at all Almost flat from 100hz up to 2khz where it jumps up to about 5/6dB increase in level. This is different to the provided response from Aura in the above PDF hoewever, which shows fairly smooth to 6khz. I've included a plot of each driver measured seperately but superimposed. The Mic was placed 5mm proud of the rubber surround of the driver in the centre of the cone in order to reduce environment impact on the results. Drivers were free air sitting on the table.

Click to view attachment

I've also included a plot of the amplifier output measured with ARTA also, to see if for some reason the sounds card was adding to the response curve, but it seems fairly flat also, bar the peak in the low freq and some mess after 20khz.

Click to view attachment

I'm potentially looking to run the drivers as the midrange sitting on the dash, either on axis to the listener, or firing into the screen as my tweeters are currently. I have access to a AC EQL and a 2-way active crossover that can crossover anywhere up from 640hz to 10khz so have a bit of flexibility in respect to crossover points. Unfortunately they are fixed at 12dB slopes however.

Some rough testing moving the mic around the driver showed the off axis reponse to vary in the lower end more so then the upper end, but this won't be a true indicatino as the drivers were tested in 'free air'.

To obtain good measurements, you'll need to move the microphone further away from the cone. 6" is too close. Move it further away (around 0.5m) and test on a single wide baffle ('IB'). So long as its more then 1m away from surrounding objects, it should be fine. Just to make sure, fire an impulse through ARTA, and select (If I can remember, its an option) to ignore the reflections (which will be clearly visible).

And if I can remember correctly, its commonly called "The Whisper". BTW, I love Aurasound!!

Cheers ~thematt~. I'm limited with where I can do my testing and the ambient noise (fish tank filters etc) so I didn't set the mic too far from the driver.

I've since dropped the driver into a small cardboard box (220mmx160mm face) and given it a quick test also, I can clearly see the 2khz peak still, so I think it's to do with the driver.

This one isn't the 'whisper'. The whisper is the smaller 1inch driver, with a plastic basket etc.

It seems inadvertantly that my mic input on the laptop has decided to stop working, so I can't do anymore testing until
I resolve the issue.

Anyone have any idea on how the internal sound card on a Dell Latitude D620 is confirgured/located? For some reason the card does not seem to accept any input now I'm hoping that it hasn't given up the ghost

maybe an excuse to get a decent outboard one, can be had for a few hundred

I've been looking at external USB pre amp (phantom power)cards, as you say this might make it a necessity.

It seems only the left channel has stopped working, which was the channel I used to measure as the right channel is the loopback (built in to my cable) so I can still do testing with the right channel only, but will not be able to take accurate impulse response readings. I'll talk to the IT dept tomorrow to see if they can fix it for me

Further testing has shown as ~thematt~ suggested there might not be a peak as shown in initial testing. I get varying response depending on where I locate the mic and the levels I test at, Most of my latest test do not show the same 2khz peak.

If you're laptop has an expansion slot available or firewire port, I'd HIGHLY recommend a firewire 2 (8 pin) preamp/input over a usb one. Faster and more reliable in my limited experience with them. Also generally the better stuff is only available in firewire too.

Is there a 2" / 2.5" version of those? I LIKE the qts on them! Good response and power handling too, and small mounting ring too. These are all very good things for my plans as well.

I think it has an expansion/PCMCIA port? I'm assuming they are one and the same. My only issue with getting something like that is that they will not have the 48V phantom power supply built in to them to run the external Mics.

The benifit of the USB pre-amp is I can drive two Behringer ECM8000's completely from the USB pre-amp, using the laptops battery power. IE I won't need to runa power lead out to feed my UB802 Mixer/Preamp.

Aurasound make the 'Whisper' also, which comes in 1 and 2inch configurations.

http://www.madisound.com/catalog/product_i...?products_id=85

These will not play low as the NS3 though, but appear to be more sensitive.

I ran the NS3 from my little Coustic amp(2 x 30WRMS 4ohms), that might have been giving 5-10WRMS into the speaker(8ohm) and they were playing pretty loud. With 40WRMS behind them, I suspect with a slightly higher crossover, ie 250hz they will be able to keep up with the rest of my front stage.

You can however get a 4-ohm version of the NS3 which would bet better suited to car audio to get more power from your amp.

http://www.madisound.com/catalog/PDF/auras.../ns3-193-4a.pdf

If I can fit the NS-3 I will get 4 of them. 8ohm doesn't bother me as i'll be running pairs in parallel anyway for a 4ohm load. I wonder i buying a whole ton of these will yield me a bulk discount... Maybe it's time to make some MTM speakers (just for fun) for my laptop as well

you dips and peaks in the high freq's can be cause from baffle step deflection, like ~thematt~ said go as big a baffle as possible

When I get a chance I'll look at making a reasonable baffle and do some better testing. I'm almost dead certain it will be practically smooth from 150-6k.

There are plans on the net for satelites using the NS3's.

Click to view attachment

To be honest with you, I dont think your orginal problem was baffle step related (especially since it changed when you moved the mic).

Higher frequencies are generated by the driver differently to the lower frequencies, and when your mic is so close, its smack in the middle of the nearfield, and potentially being influenced by large degree's of constructive and destructive inteference. Basically, that close to the driver, and your unit isnt behaving like a piston.

Easy fix is to simply move the microphone away.

I have done a little more testing with these drivers tonight.

The drivers have been fitted to some tiny enclosures, namely 700-800mL (I am only guessing) sealed enclosures.

I tested the each driver in it's own enclosure, with the mic roughly 600mm away from the driver directly on-axis.

Results have been interesting. I've set the measurements at a higher resolutin, ie 1/12th Octave smoothing.
Click to view attachment

And the same result with 1/3rd Octave smoothing employed.
Click to view attachment

I'm unsure as to why I have a difference at 1.5k and 2.5k between the two drivers, and I'm also not quiet sure how to deal with the large dip in response at 15k. This suggests to me that I will requirea tweeter to smooth out the upper octaves, but it 'might' be possible to get away with the single driver from 250hz up...

lookin good!!

i think the next step will be to overlap the in cabin response onto the graphs you've generated and you will have a lot of valuable information that will keep you busy correcting for a long time!!

the graphs appear to be pretty linear to me? that dip at 15kHz is going to be pretty audible if your listening ability extends that far.

what kind of music do you listen to anyway?

Floyd Toole found a correlation between a peaks Q, the type of music listened to and how audible the peak is.

for example a jazz track, a peak with a very wide Q is audible, but not the same as in a rock track.

Dont know how conclusive the study was, but interesting all the same.

For drivers response i usually only use a filter that has a very wide Q and only very little amplitude boost/cut (ie, only 2 or 3dB) unless of course you are certain that any change in the time domain is going to be improved.

That would be the way to go SCorpion. It's is a bit tricky to implement with the trial versino that I can't save in however. I'll jsut get an incar response graph and work from that.

I seem to find it easier to real time tune in the car then trying to work back. A 'sweep' takes 5-seconds to take and allows you to see exactly what the last EQ change achieved.

I have been using Q's not much less then 3, but usually 4/5 and I have a real issue with EQing more then about 3dB so try and stray away from anything more aggressive then that.

Tieing this thread in to my use of the driver...
http://www.mobileelectronics.com.au/forums...mp;#entry995625

Musical choice is all over the place but I want the system to play 'anything' well, not just specifically one Genre.