Measuring Frequency Response
Please give me your opinion about my idea for a way to measure frequency response of a loudspeaker system far-field (microphone about 1 meter away from the loudspeaker system) in a room.
My method is a modification of the method of Project 58 at Elliott Sound Products. Here is the link:
At input to the cosine wave generator of Project 58 is a sine wave generator running continuously. The output of the cosine wave generator is amplified and the amplified output signal transduced by the loudspeaker system under test. The voltage output of a pre-amplified measuring microphone (not Linkwitz's) is displayed on a cathode ray oscilloscope (CRO). At some reference frequency of about 200 HZ, the RMS output of the sine wave generator at input to the cosine burst generator is adjusted to produce a suitable peak voltage of the cosine wave displayed on the screen of the CRO. Measuring response of the loudspeaker system under test at other frequencies, the voltage output of the sine wave generator is adjusted to produce the same peak voltage on the CRO that occurred at 200 HZ. Thus the extent to which the output voltage of the sine wave generator is reduced or increased indicates sound intensity produced by the speaker under test relative to that occurring at 200 HZ.
-Does this seem like a legitimate method to measure FR? Comments and suggestion are invited.
I may be missing something, but it seems to me that going digital and obtaining the impulse response of the loudspeaker is a way superior method for evaluating the linear aspects of the system. Today this is not very expensive, software is affordable or even free and measurement equipment can be had for little money.
free software can be found on the net and you will need a decent mic anyway which bring us to the soundcard which can be had (if you don't already have one) for 50-100USD or so.
Steady state meaurements (sines) will give huge variations from reflections in the room so the proposed method with a scope is not very useful if you want to know what the speaker puts out.
With PC based software you can get semi-anechoic mesaurements from about 500Hz and up by using gating techniques.
Joseph D'Appolito wrote the book on it ( Testing Loudspeakers ).
If you are measuring indoors isn't gating necessary?
Anyway: I found this DIY approach to polar measurements interesting
PC based FR measurement
Please name the free to cheap software available for measuring FR.
There is the question of the quality of the sound card that is currently installed in my PC. If I had to install a new card, I'm not certain that I would have the skill to do that. In addition, I believe that whether or not the software will function according to plan depends on the type of sound card.
There is the question of adjusting the output level of the microphone to the level that a particular sound card will accept.
For the above reasons, and probably additional ones, and the fact that I am not very skillful at using a PC, I desired to avoid the digital route.
Project 58 is about a sound generator that produces an impulse of sound that is of lower distortion than gating a continuous sine wave signal. The generator was devised by Linkwitz some time ago, I assume. It is most likely true that Linkwitz currently measures FR digitally.
Thanks for the responses. How about comments about whether or not what I have in mind would give an accurate result even if methods using a PC might be considered to be easier?
Speaker Workshop is free, a little strange in teh user interface , but has a powerful crossover design function with optimizer.
If you can tell us what computer sound card you have, someone should be able to tell you if it will work. Speaker workshop also has a "test card" function, but it can sometimes report compatibility for cards that just barely pass and don't give good results.
ARTA is free in demo mode and not very expensive to buy.
SoundEasy is about $300, and very powerful.
What I did once was take the software cool edit, generate a pulse, play it through the speaker and record it and do an FFT to get a frequency response. The frequency response I got was not anything like the response I get with Speaker workshop, so trying to cheap out with a kluged together mic and method will not necessarily work.
If you want to measure impedance and T/S, you will need to build a jig, but that is a lot easier than the pulse generator. For measuring Frequency response, you can build a mic and preamp and get by for a few 10's of dollars , or buy a mic (Behringer ECM-8000 is often used) and preamp and get by for about $100. Either way, you will want to get the mic calibrated, which can be had for about $40-70 from either Kim Girardin or Herb Singleton.
what's wrong with rew???
Thanks for naming some of the software available.
The type of sound card that is part of my computer I think is identified somewhere on the hard drive, but I'm not certain where to search.
So, but anyway, I suppose that I will be exploring using my computer to make the measurements.
When I achieve something I'll make a report and/ or have questions for anyone willing to help or having an interest.
Best Regards to all,
My sound card is Crystal Beta WDM Audio Driver Version 4000. A "working feature" of the card is "DOS Box Sound Blaster Emulation (sound and music = SB Pro)".
From my previous looking into sound card requirements for measuring audio, I seem to recall that "Sound Blaster Emulation" fits in with many of the the measurement softwares that have been developed. -That's my vague recollection.
Is anyone able to say if this is a generally usable sound card?
Thanks in advance, where possible,
Sounds like an older computer. It will probably work fine, make sure you limit any signal you input to the card to less than 2 volts or so or you will blow your inputs. This is a risk with any measurement software. Some build a diode clipper into their jig circuits as a hard limit. IF you use Speaker Workshop, make sure to read the unofficial manual - there are a lot of tricks, and while you don't have to do everything exactly as it is in the manual, it is a good start.
Note that while it is tempting to try to use the mic input, don't. IT is noisy and you will give up in frustration trying to take this shortcut.
|All times are GMT. The time now is 10:49 PM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio