LCR Impedance Compensation - Base Values

[DR ANALOGUE]

Actually I measure a Scan Speak 10F. However, the issue is the same for other speakers.
Re (or Rdc) is measured lower than at 20Hz and the lowest point above fs.
In this case Re amounts to 5,7 Ohms, while the impedance at 20 Hz is 6,0 Ohms and the minimum above fs around 6,4 Ohms.

In terms of calculating the compensation, I was wondering which value should be targeted. Which value results in more appropriate compensation?

Do you have any opinion on this?


Many thanks
Michael

 

[Lojzek]

Hallo Michael,

the proper compensation values (L,C,R) are the ones that let you achieve the desired acoustical/electrical goals. Under Additional circuit calculators in a spreadsheet "Passive Crossover Designer" you will find the one to calculate textbook RLC parallel notch filter. You can use "Response Modeler" to create an impedance profile to feed the PCD and test the rlc values, if you haven't got already a zma file that represents your drive unit. XSim is simpler to use for some.

 

[TBTL]

Note that fs sometimes depends on the drive level.

 

[Dave R]

The reason that the impedance is slightly higher (6.4 ohm on your case) above the resonance peak is mainly due to the influence of the voice coil inductance. The voice coil impedance has less effect at lower frequencies.

Re is the dc resistance (5.7 ohms in your case) that is typically used.

If you want to try some hand calculations for a notch filter, consider this link:
Crossove_Design

Note that the shape of the impedance peak is important. Qes and Qms are factors that affect the shape. You need Qms and Qes for the calculations, in addition to fs and Re. The results may only get you in the ballpark. Simulation software as previously noted can be very helpful, and may save you a lot of trouble.

 

[DR ANALOGUE]

Yes, there are nice calculation tools available. Still, I prefer to know how things are calculated and proceed the old fashioned way and my beloved excel sheet
I improved my measuring equipment and can now measure impedance to a precision of 0,05 Ohms .... believing in a trail and error free compensation circuit.
Also calculated in-box-Q corresponds to the measured Q within 1%.

However, following your comment, I would assume that
- firstly, applying first the voice coil inductance correction based on Re
- secondly, measuring impedance including the correction
- thirdly, applying the LRC peak compensation based on the measured impedance value
may give more appropriate results.

I will try and report...

 

[DR ANALOGUE]

I respect all your comments! Without simulation tool, you won't get a very flat impedance curve.

Approaching Rdc gives more flexibility in shaping. If the tweeter has a higher SPL tha the woofer, this is the way to go.

 

[eriksquires]

I kind of cheat. I use XSim, import the zma chart for the driver in a box, and then use the notch filter calculator.



However, doing this for a woofer can require quite a bit of power through the components, make sure you properly adjust power handling.