![]() |
Tweakers' Asylum Tweaks for systems, rooms and Do It Yourself (DIY) help. FAQ. |
|
I am a newbie to measuring the Thiele-Small parameters for drivers, and I am trying to set up a reasonable system to measure them for a vintage 15" Jensen Golden Voice driver that is clearly very efficient. I want this driver to be in a correct cabinet, as it goes with a Motorola system I am restoring. Forgive me if I don't use completely rigorous terminology...I am new with this.I am using a freq generator that I have on my computer fed into a SS JVC integrated amp. The speaker outs for one channel are going to either a 10-ohm 5-W resistor, to calibrate volts with impedance, or the driver itself without the resistor. This is based on a method published in a speaker book by David Weems. The speaker outs for the other channel on the amp have an 8-ohm dummy-load connected.
Here is the issue I am having in getting Fs and the correct voltages for Fl, Fh, etc. If I set the freq to 400 Hz, and adjust the amp output to about 200 mV when measured across the 10-ohm resistor, I find that voltage around 14 Hz is about 260 mV, rises to 290 mV around 45 Hz, then trails back down to 200 mV slowly by 400 Hz. Naturally, I would rather have a completely flat output spectrum before I measure the driver itself, but here is what I did: I measured the output spectrum with the driver at the same frequencies as without, then used Excel to subtract the original amp spectrum voltages from the amp+driver voltages. This gave me a picture perfect voltage spectrum for the driver, just like what I see in the books, with Fs at around 75 Hz and then another hump around 260 Hz. Everywhere else on this difference spectrum is flat.
My question is this: can I calibrate the voltage to say 100 mV where the amp's spectrum is flat (around 400 Hz) with only the 10-ohm resistor, and then the voltage at Fs is 100 mV + the difference between the voltage at Fs with the driver and voltage at Fs with the 10-ohm resistor? Long question, and I hope I'm asking it correctly. In other words, can I come up with a correction function to take into account the non-linearity of the amp itself around Fs?
If I can't do this, what inexpensive amp would people recommend that I use. Of course, maybe it is the sound card and not the amp that is not flat-spectrum.
Thanks in advance!
Kurt
This post is made possible by the generous support of people like you and our sponsors:
Topic - Need T/S parameter advice (a bit long) - kurt23 10:34:14 06/14/04 (0)