Shredhead Posted February 7, 2014 Report Share Posted February 7, 2014 I have been using REW to figure out T.S. parameters using the impedance sweeps with added mass. The data seems trustworthy all except for the sensitivity in SPL figure for the driver. I don't know how REW is figuring this number out just based on doing impedance sweeps through a headphone amp. Does anyone have experience doing this and if so -are REW's sensitivity numbers accurate? Link to comment Share on other sites More sharing options...
lilmike Posted February 7, 2014 Report Share Posted February 7, 2014 How far off are your results? When I did my "tutorial" over at AVS, things matched pretty well. SPL might have been off by a dB or two, the rest of the data was generally within 10% of the published specs, which is entirely reasonable given the differences in measurement setting and technique. It's my understanding that with the Thiele Small parameters and the Sd of the driver, the 1W SPL can be approximated, because we "know" the motor force and cone area. This approach will be an approximation though. I find it a bit more accurate to apply an appropriate voltage and measure the SPL. Link to comment Share on other sites More sharing options...
Shredhead Posted February 8, 2014 Author Report Share Posted February 8, 2014 Oh so it is derived from the T.S. parameters and the cone area, thanks! Unfortunately, there are no manufacturer specs to refer to on the drivers I'm testing. I think the T.S. are pretty darn close it's just that the SPL number seems wrong. I would love to do 1W tests outside but good lord, that would be an enormous amount of stuff to carry outside and I couldn't really trust the test because there would still be reflections off of houses. Do you do your SPL tests outside? Maybe I'll try one inside and see what happens. Link to comment Share on other sites More sharing options...
lilmike Posted February 9, 2014 Report Share Posted February 9, 2014 1W at 1M with a simple sine wave source will get you close enough indoors. No, it's not perfect, but at ~60 Hz, you'll be close enough, and then you can use a typical AC voltmeter to set your level. When I am measuring something that I plan to share on the 'net, I drag it all outside, grab a beer or two, and make an afternoon of it. When it is just to see if I am in the right zip code with respect to how a design turned out, I measure low-level and close-mic indoors. Most of my crossover design measurements are indoors as well, but for that work, I use a dual channel MLS-based measurement system, so I can do a decent job of gating the reflections and noise. MLS signals are also a LOT less annoying than sine wave sweeps for the rest of the people that live here. With REW and a headphone amp (and all the low-level testers (DATS, WT2, and so on), larger drivers may not respond well to the low-level signals. I've found that testing them at volt or two of drive is usually sufficient to get decent data, there are headphone amps that will deliver than cleanly. There are also small power amps that can be used, or for that matter, most of us have larger ones sitting around. If you can see the cone moving, you're probably OK as far as level goes. When working with a real amp, mind your levels and be careful. I let the magic smoke out of my M-Audio Transit trying to measure a driver, that sucked. Link to comment Share on other sites More sharing options...
Shredhead Posted February 10, 2014 Author Report Share Posted February 10, 2014 I think you're one hundred percent right on the headphone amp thing. You've inspired me to re-do the T.S. test in REW with an amp. I think I will use the voltage divider idea that REW's help menu suggests to avoid the smelly smoke thing. It will take me a while to figure out what parts to use but when I'm done, I'll post the differences. Your measurement system sounds interesting. That is using impulse responses right? How does that response compare to frequency sweeps -does it graph different because of gating the reflections? Thanks for the info Mike. Link to comment Share on other sites More sharing options...
lilmike Posted February 11, 2014 Report Share Posted February 11, 2014 I'm learning how to use SoundEasy now. Not exactly "easy", but it offers a LOT of capability when it comes to crossover design and optimization. ARTA is every bit as good, if not even better when it comes to measurements, but SoundEasy does a LOT more than just measurements. I actually bought it for crossover simulation and optimization. The MLS signals are sort of a static noise, and only last a few seconds. A quick "pssssshhhhht" vs "woooooooooop, woooooooooop, woooooooooop, woooooooooop" is a lot easier to deal with. Noise rejection is pretty good too, but environmental noise certainly influences acoustic measurements, there is no way around that. Frequency response, impedance, and Thiele-Small measurements all seem to be on the money though. The impulse is calculated from the MLS, so I can look at that and gate out everything up to the first reflection. When I gate things while measuring inside, it rolls off the low end depending on the width of the gating window, but for crossover stuff, who cares? I'm not too concerned about what is happening below 200 Hz when I am crossing 3 or 4 octaves above that point. I tend to focus on the 2 to 3 octaves to either side of the crossover point, as the rest is typically outside the area that the crossover impacts. The SoundEasy learning curve for me is most definitely vertical at this point. I have a LOT to learn. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.