AES Journal Forum

Interesting article about capacitor for closed boxes. I feel the following references deserves to be mentioned in this context:

German Magazine "Klang & Ton" - 5/97.
Lautsprecher im geschlossenen Gehäuse mit Vorkondensator
(Loudspeaker in closed boxes with seriescapacitor)
by Dipl.-Ing. Ulrich Ewerlin

For second order passive circuit in connection with the loudspeaker, see patent by Dr. Clemens Kroll, EP 0 568 721 B1. As far as I can understand this patent is about a specific work method.

In my opinion, none of the above references are as elaborate as the present paper.

Best regards,
Claus Futtrup

If the output capacitor is simulated by the use of a frequency-dependant current-loop around the power amplifier a different kind of trade-off is possible.

The clear advantage of a "real" capacitor is the frequency-dependant transformation that it offers. The other advantage of the topology comes at a cost at a different place: The LF protection function does not keep the power amp from unnnecessary voltage swings at out-of-band frequencies.

With the simulated capacitor the transformation effect is lost but amplifier headroom is gained by keeping it from amplifying out of band content.

Which one is preferable is dependant on application of course.

Kind regards

Charles Lehmann

My sincere thanks to Claus Futtrup for this further information, that I was not aware of. The general procedure of integrating a series capacitor with a closed-box loudspeaker goes back, of course, to Ernest Benson. The purpose of this paper, and of my earlier listed publications that added second and third order passive filters, has been to present rational methods that a designer can use, and also to emphasise that the parameters of the driver, box and capacitor, are inter-related, constituting a unique alignment.

Charles Lehmann raises an interesting point. Without a series capacitor, the amplifier load is a high resistive impedance around the acoustical cut-off frequency. At lower frequencies, the load impedance falls until, at very low frequencies, it is the d.c. resistance of the driver.

When the capacitor is in series, the load is close to the driver resistance around the acoustical cut-off frequency. As the frequency falls, the load impedance rises, becoming more and more capacitive.

I have not investigated these impedances in detail, but would estimate that, with the capacitor in series, the load impedance may well be more favourable to the amplifier, though dependent on the magnitude of the reactive components.

Neville Thiele

Dr Thiele,

I remember a KEF loudspeaker design (Kef Constructor PL 301 designed around 1984) ) using a series capacitor to extend LF to a certain degree.
The value of this cap was about 600uF. The capacitor was used with a 12" woofer (B300B in a 80liter closed box.)
It was the time that KEF swore on 18 dB/oct. crossover designs. The crossover filter was commercially available as DN28.


It would be interesting to investigate a QB3-alignment versus this present design.


Best regards

Stefan Willems
Senior Acoustic Engineer
Premium Sound Solutions
Denon&Marantz Holding
Leuven, Belgium

the PL301 kit also appeared in the ELEKTOR magazine january 1986 .

My thanks too to Stefan Willems and Koen Weijand for their additonal information. I can only repeat my reply to Claus Futtrup that the purpose of the paper was to present a rational design method for this already-known procedure.