Speaker Impedance Testing

[jaekfre]

I'm in the research phase of my first speaker project. I'm wanting to test my woofers to find the exact impedance at the desired crossover frequency. I came across the article linked below but wasn't able to find any threads on diyaudio referring to this method so I wasn't sure if this is an accurate way of testing. Does anyone have any experience with this method or other recommendations?

http://www.installer.com/tech/sit.html

Thanks

 

[PeteMcK]

ARTA http://www.fesb.hr/~mateljan/arta/download.htm
ARTA Jig - http://zobsky.blogspot.com/2008/01/simple-loudspeaker-measurement-jig-for.html

 

[sparkless]

Why not just use a Zobel (Impedance equalizer) to flattern the impedance curve?

Just in case you don't know it's a resistor & a capacitor connected in parallel (across the speaker terminals).

C = Le / Rc squared

C = capacitor
Le = voice coil inductance in mH
Rc = 1.25 x Re

Or C = 0.16/ (Re x f)

Re = DC resistance of speaker
f = frequency at which the drivers impedance is equal to 2 x Re
And again the resistors value is 1.25 x Re

Try this site for a calculator.
http://www.diyaudioandvideo.com/Calculator/ImpedEqual/

I believe this method is more foolproof

Steve

 

[jaekfre]

Thanks for the replies.
I'm planning on using a zobel network. However, all the graphs I've seen of the impedance after a zobel network is installed still shows an impedance a little higher than the speaker nominal impedance. From what I've seen, aboout 1 or 2 ohms higher. Is this normal? Also, I'm going to be wiring two woofers in series. I assume (which I know what this usually causes) I should add the resistance of both drivers together. Should I also add the drivers inductance together when calculating the values for the zobel network?

 

[sparkless]

It is not suggested that you run speakers in series. No two speakers will be exactly identical, even if they are the same model from the same manufacturer. This means that they will act slightly different from each other when presented with the same input signal. When wired in series, these differences will cause distortion in the form of back EMF.
You can however run both coils of a DVC speaker in series. The cone movement for both coils will be identical, and will not cause distortion problems. This is the formula to calculate the impedance of speakers in series.

http://www.diyaudioandvideo.com/FAQ/Wiring/

When the woofers are connected in parallel & through one low pass filter you can use either one zobel using half Re & Le, or one zobel for each driver based on Re & Le.

I'm considering using two seperate identical low pass filters for each woofer & a zobel for each in my next project (two way, dual woofer). (think of it like a three way crossover but without a bandpass filter). This will also have the benifit of keeping the impedance at the speaker terminals unchanged (not altered by either series or parallel connection).

 

[jaekfre]

Well, dang it. Not is series?!? How have I not read this yet out of the 100s of hours of research I've doing?!? Makes sense though.

I've already bought 4 woofers (4 ohm nominal impedance each) for a MTM configuration for 8 ohm nominal impedance. Guess it's all a learning process anyway. That's what makes it so much more enjoyable than just buying them from Best Buy right?

Maybe I should do a 3 way with both woofers. Decisions, decisions.

Thanks for your help and advice!