Biamping and Zobel Networks

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
It still depends a bit if you are just bi-amping (splitting the passive crossover) or you drive your speakers active (no passive crossovers).

1. Zobels can also help keeping the speaker an easier load to the amp if your amp has a low damping factor or works better this way for whatever reason (this refers to some tube amps mainly). But it must be a designated Zobel network (generally behind the actual crossover network).

2. In active mode Zobels are generally not used and also not necessary.

Some expensive speaker cables even have Zobel networks built in, but I personally would reject all this post-correcting stuff.
It is my understanding that Zobels correct for a rising impedence of a driver, with frequency, due to the inductive
component of the voice coil.
As Grey correctly points out it helps the correct operation of a passive crossover.
I have found that sometimes it helps to Zobel active systems if the amplifiers used are not quite up to the task so I would not say that they are not necessary.
I do not believe that Zobels improve the damping factor as Iohk suggests as they decrease the impedence at high frequencies and hence the damping factor is reduced at said frequencies.

Zobel networks in the loudspeaker crossover obviously do not correct the damping factor of the amp.
But they flatten the impedance of the speaker, making the network of crossover and speaker a more easy, non-changing load. This can change the sound significantly with certain amplifiers. That is one reason why some loudspeakers sound so totally different with certain amps.
I had once a lenghthy chat with Peter Quordrup (Audionote England) about that subject. He agreed that the influence interaction between dynamically changing impedance and damping factors are often highly underrated.
I am no expert in this task, but after having combined so many different amplifiers with so many different loudspeakers, there must be something about it. This also "can" be an explanation why some synergies of hifi chains work so splendid.

While the zobels benefits in passive networks are primarily in terminating the crossover resistively to help ease the design of the crossover, they are not always needed in order to design the network. Zobels are still however useful even in active systems. Power dissipation in the output stage of linear amplifiers is critically dependant upon the reactance of the load, as is peak current demand. The more resistive the load, the lower the power dissipation and current.
Additionally, the frequency response measured at the loudspeaker input terminals due to the interaction between amplifier output resistance/cable resistance and the loudspeaker impedance will be minimised using a zobel network.
I suspect that someone in this thread might be confusing the Zobel network attached to some amplifiers output to improve stability and the Zobel network used in passive crossovers to flatten the load impedance that the filter components see.
I have never used a speaker Zobel network on any directly amplified drivers, (read active crossover). I can't see that they would be of any advantage to any well designed amplifier and speaker system.
The idea that an amp may run cooler seems reasonable but they are designed to drive reactive loads. I suspect that they could degrade the dampening of the system by adding a positive reactive component to a circuit that already has a negative reactance.
Anyhow that’s my 2 cents worth.
Regards WALKER
I agree, Walker, that the different Zobels do not have to be mixed up.
Also I myself have no experience with loudspeaker Zobel networks in active systems, the idea always seemed quite contradictory for me. But many of the driving amps had their output Zobels nevertheless.

It is strange, but we all are talking about loudspeaker crossovers so often - but on the other hand we all know very well that active systems are so much better most of the time !
How many passive crossovers I already tried to adapt, modify or make it better with little or doubtable effort ?
But, alas, active systems are, besides being more expensive, even more tricky to set up, they can be much more of a go-no go affair. I mean proper active systems, not the biwiring or biamping "bastards".
In virtually no recording studio in the world the musicians and the engineer listen only to passive monitors, even smaller sized monitors are getting almost inevitable. Have you ever listened to, say, a Genelec 1029 ? These are very small loudspeakers in heavy non-resonant aluminium cases and can sound very un-hifi-ish clear, neutral and punchy. All this is in a cheap opamp powered active two way system package. It may not be the best loudspeaker of the world, but it is a slap into the face of so many much more expansive systems.
And there are, btw, no zobels needed anymore.

I am sorry to have left the path of this thread almost, but should we better not search better ways of getting our systems active ? I am asking myself this question quite often although I know that it is not that easy.
There must be better active crossover schematics or different ones for the critical cases. Then discussions about zobels in passive networks could be a thing of the past.
In a different thread maybe...

lohk said:
Zobel networks in the loudspeaker crossover obviously do not correct the damping factor of the amp.

I don't know about correction, but they can surely change it.

If Zobel's flatten (by lowering a portion of) the impedence curve they could alter an amp's daming factor at a given freq.

DF = speaker impedance / amplifier output impedance

Say an amp has a DF of 300 at 2khz where the driver's impedence is 13ohms 'before' Zobel. If a Zobel circuit is applied and the load 'seen' by the amp is droped to 8 ohms at the same 2khz, the amp's damping factor would now be 184.

(unless i'm mixing Zobels. And we know what can happen when one mixes Zobels- whew! don't even think about it)
It is easy to proclaim that Zobels do not affect the sound of drivers in an active speaker without trying it but let me assure that there is a difference most of the time. The difference may be small at times but it is there none the less, after all are we not trying to get every last bit of performance from our designs.
To prove what I say hook up an amplifier to a driver and feed it with a square wave (say 1k keep the amplitude low to avoid harming the driver) and observe the output with a scope. Repeat the test with the zobel attached and you will see what I mean.


Sorry to keep you hanging.

When an amplifier drives a reactive load you will notice a certain amount of ringing or overshoot on the tops of the waveform. A zobel will change the amount and character of this overshoot.

I suggest you try Zobels on an active system and judge for yourself. If it works use it, if not leave it out. I have found that Zobels work in most instances, even in subwoofers.

Jam, I believe I know where your coming from ,buuuuut. The ringing that you will see without reactive correction is worst with correction. You just need to know where to look for it.

If you were to monitor the current between the Zobel network and the driver you will find the ringing alive and well. If you look for voltage ringing between the amp and speaker, (or in this case the Zobel network) you should see less ringing.
With reactive correction the amp has very little control over the LC circuits oscillations without correction the amps dampening has more control. These conclusions I have reached from a purely theoretical position. I have not tested it, I could be wrong but I don't think so. I will setup the experiment tomorrow if time allows.

Look up tank circuits.

The other misconception is that a Zobel network flattens the drivers rising impedance. Well it does and it doesn't, the driver's rising impedance doesn't change but there is a new element with falling impedance, (with rising frequency).

Look I could be missing the point here, feel free to shoot this down, we'll all gain from the outcome.
Regards WALKER

I am glad you are about to put this to the test.
I should have clarified that I mount the speaker Zobel network right in the driver terminals themselves, which I believe is important.
I have found instances when the Zobel was determental but rarely. I will be waiting for the results of your experiments, please also listen to the results.


[Edited by jam on 12-05-2001 at 08:27 AM]
Jam, having the network at the driver is, as you said, the way to go but the oscillations will still increase over connection without the network I suspect.
I will listen but I won't have the time to measure it I suspect. I can take oscillograms but I still don't know how to post images, HELP?

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