Zobels and output impedance

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I think you'll have some losses anyway, because the output signal is divided by the impedance of the network and the impedance of the speaker. I think the output impedance will raise because the feedback trace is connected before the zobel network and not between the speaker and the network, so the feedback network will not handle it this way. I just wondered if you can connect the feedback trace between the speaker and the zobel network, but probably it won't work that well because the network occurs a phase shift (or isn't it that big???).

For those who're asking what a zobel network is:

Zobel network or Zobel filter [Also called Boucherot cell after Paul Boucherot who worked extensively with electrical networks and power.] 1. A filter designed according to image parameter techniques. 2. Audio amplifiers. Zobel networks are used in audio amplifiers to dampen out high frequency oscillations that might occur in the absence of loads at high frequencies. It is the commonly seen series resistor-capacitor combination located directly at the output of the driver stage, just before the output inductor. Typical values are 5-10 ohms in series with 0.1 microfarads. The network limits the rising impedance of a loudspeaker due to the speaker coil inductance. The output inductor found in most power amplifiers used to disconnect the load at high frequencies further aggravates this phenomenon. See Douglas Self's book for a good discussion of audio amplifier Zobel networks. 3. Loudspeakers. Some loudspeaker crossover designs include Zobel networks wired across the tweeter (high frequency) driver to compensate for the rise in impedance at high frequencies due to the inductance of the voice coil. The goal here is to try to keep the load seen by the crossover circuitry as resistive as possible. [After Dr. Zobel of Bell Labs.]

so this isn't correct (see forward), excuse me :D :
NOTE: zobel networks can also (and are often) be build of a resistor and an inductor placed in parallel.

Best regards,

Hi, I'm back...

so the impedance of the zobel network:

Z= R + 1/jwC => Z = R + 1/wC

w = 2*pi*f
R = 10 ohm
C = 0.1 µF

so Z = 1601 ohm at 1kHz and 169 ohm at 10kHz

the output of the amp (with feedback) is very low (~ 1 ohm or less) (note: haven't calculated yet, just took an excisting example)

so the total resistance (Zamp // Zzobel) will change a very very little, let's say that it remains stable.

best regards,
Zobel Suff

Hi Marcus,
"If you add a Zobel network to the output of an amp, will the output impedance of the amp change or will the negative feedback take care of that?"

As I understand it, the series inductor will add amplifier output series resistance according to the DC series resistance of the inductor, thus reducing overall damping factor, hence stout wire is usually used.
The inductor effectively isolates the amplifier output and summing point (feedback node) from high frequency loads, therefore a shunt RC network is required at the summing node point to ensure amplifier HF loading.

After the series inductor, an additional shunt RC network provides a load to compensate the rising frequency/impedence of typical inductive cable/loudspeaker system loads.
This second RC network is not always included but is required (and needs to be optimised) according to 'Douglas Self'.
Commercial amplifier zobel outout networks are included to provide amplifier worst case load immunity, but are never optimum for your cables and loudspeakers.

RC networks across each of the drivers in the loudspeakers will further help to provide a zero phase angle (resistive) load to the amplifier, and much contribute to clean, correctly dynamic and extended system sound.
In my experience, with an optimised cable/speaker load, on most amplifiers the series inductor is not required, and this gives a better lifelike sound.
Naim amplifiers are an exception.

Regards, Eric.
I don't have access to Self's book (yet). But the second shunt RC network would have to be located at each speaker since it have to be optimized. This is how it is done if I'm not mistaken.

If the first RC network and the series inductor are not included at all, would it be enough with only the optimized RC shunt across the speaker to ensure proper loading of the amp? Or is it OK without any correction circuits at all?

Why do the amp need HF loading at all? The lighter the load is, the easier for the amp, or have I got it all backwards?

so if I understand it correctly:

If you use a RL network in series with the feeback node, you must use a RC (zobel) - network too.

So, this is only necessary in bad (or worst) load conditions and I don't need the RL-network for my normal speaker equipment at home, but if we're going to play music 'on the road' where we have to deal with long speaker wires and other (not known) speakers it's better to use the RL-circuit.

to Marcus:

Why do the amp need HF loading at all?

the Zobel network seems to be there to protect the amp against HF-oscillation when no load connected (like Douglas Self said); and believe me, this happens very often :)D been there). I'm going to try now to modify my first (toasted) amp-PCB. The devices blew when I disconnected the speaker, I hope they won't do that anymore.

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.