, dus <copy past> 
Link: Distortion measurements
...to compare a few woofers I had in mind for a project. I 'slightly' nudged the drivers into a flat SPL over a more or less sensible frequency range for a fair comparison. Since it might be of interest to others I've put up a quick&dirty pic page on my website and added a few well known drivers too.
There is a glitch in the SLS12 plot at 150Hz, which is not from the driver (obviously, but I thought I'd better mention it). I had a few other drivers waiting to be tested as well, but after a a couple of hours of measuring, eq'ing, remeasuring, calibrating the levels, bursts, sweeps... my cat and I got fed up with the whole thingSo I put on a movie and my cat returned to her usual purring and sleeping. More drivers in the future after I set up a proper website.
Noisefloor is at -90dB above 1KHz (driver level at +4dB) and rises to slightly under -80dB at 100Hz. The graph of the HDS205 Exclusive marks the noise level very well with its F4 plot and the F5 above 1KHz. Level is at 2.8V into 8ohm for the SLS10 and all other drivers were calibrated to this SPL. I'll probably do higher SPL in the future as well to see how soon the drivers give up and start to compress/distort more. The woofers smaller than 10" were not amused by 20-40Hz (remember, flat spl...), so I left that out.
Cheers,
Hans.
Re: measurement conditions
Posted By: Hans L <DIY_Freak@hotmail.com> (s5592553d.adsl.wanadoo.nl)
Date: 5/10 2:23p.m.
In Response To: Re: Did some distortion measurements... (jkrutke)
Hi John,
Yes, those SLS woofers show exeptional value for money! End of May or so I should receive a couple the other HDS midbasses, the HDS tweeter and the XXLS12DVC as well so it will be interesting to see the difference, as far as harm distortion goes anyway.
Measurement setup:
Mic IBF-EMM8 with calibration file (relevant above 10KHz)
Mic preamp IBF-MP21
Soundcard Delta410 using spdif out and analog in at 48KHz
EQ with DCX2496 (modified, spdif in)
HK receiver for volume control and amplification.
The noise of the chain is hugely dominated by the mic and the mic preamp. There is also a pesky peak at 50Hz and a few multiples which I can't seem to get rid of. Luckily these are normally swamped by the spl of both the fundamental and the harm products if the spl is high enough to begin with. I suppose in-line transformers would help, but that has its disadvantages as well, apart from begin a rather large investment considering all the groundloops I would like to break. I might consider breaking the ground altogether temporarily when needed. With the AC noise at 50Hz, low freq&spl measurements are worthless, though the pattern is quite easily recognized.
Ever since I installed an expensive, ultra quiet fan PSU from Nexus (especially for audio use of PC...), I'm also experiencing a pesky peak at almost 30KHz. Had a look at the psu guts not too long ago, but it wasn't easy to reach the components that probably require better filtering. At this point I'm assuming it's the PSU, but I have to wait for my scope to come back from repairs to verify that and start modifications.
What type PSU are you using btw? Any noise problems like above?
I bought an external soundcard (usb, m-audio Sonic Theater 96KHz) for clean measurements concerning the 50Hz peak, but it turns out the distortions figures of that card are absolutely appalling... too bad. It does make a very good portable freq measuring setup though. I'm also eyeballing the Audiophile 192KHz (ADC(!), one of very few afordable), also by m-audio for even higher freq, high order harm dist measurements, including elextronics, not just speakers. I hope Bohdan will expand the non-linear options a little like I suggested a few days ago on the Soundeasy list.
As you can see, I used Soundesy for these measurements, but prefer the GUI of Lspcad 5/6 Pro for design work and quick spl measurements. So I used Lspcad for determining the DCX EQ/filter settings for flat spl, mostly +/- less than 1dB.
Measurements where done close range, less than an inch from the plane of the basket, unbaffled. That close up I don't think dipole behaviour plays a role. The driver size itself (apart from being unbaffled/dipole) might influence spl a little at close proximity, I'll check that some other time. Freq corrections were done with help of an MLS measurement with a suitably long window.
Did I leave anything out?
Hans.
