I know that the primary use of a Zobel network on a speaker driver is to equalize the impedance so that the crossover network will work properly. I also know that the DIY users of full-range drivers often use a Zobel network to tame the rising response that such speakers commonly produce.
My question is this: is it possible to predict how a Zobel network will affect the frequency response of a full-range speaker? Can one use the Thiele/Small parameters of the driver to do so accurately, and thus, tailor the response of the speaker to one's taste? Or, is it strictly a trial-and-error proposition?
I understand that manufacturers' specifications can not always be trusted, so that if one uses the published T/S values, one can not always be sure of the desired result, but in theory, should it be possible to predict the effect of a Zobel on the FR of a speaker? If so, what are the equations to do so, and can the type of filter be defined as, for example, 6dB/octave, or shelving, or something else?
Thanks to anyone who can educate me on this matter.
My question is this: is it possible to predict how a Zobel network will affect the frequency response of a full-range speaker? Can one use the Thiele/Small parameters of the driver to do so accurately, and thus, tailor the response of the speaker to one's taste? Or, is it strictly a trial-and-error proposition?
I understand that manufacturers' specifications can not always be trusted, so that if one uses the published T/S values, one can not always be sure of the desired result, but in theory, should it be possible to predict the effect of a Zobel on the FR of a speaker? If so, what are the equations to do so, and can the type of filter be defined as, for example, 6dB/octave, or shelving, or something else?
Thanks to anyone who can educate me on this matter.
Hi,
The consequence depends on the drivers impedance curve, with and
without the zobel, and the output impedance of the driving amplifier.
Normally with most amplifiers the zobel will have no real effect.
With unusual amplifiers with significant source impedance it will.
You need to know what that is to predict the effect.
rgds, sreten.
The consequence depends on the drivers impedance curve, with and
without the zobel, and the output impedance of the driving amplifier.
Normally with most amplifiers the zobel will have no real effect.
With unusual amplifiers with significant source impedance it will.
You need to know what that is to predict the effect.
rgds, sreten.
A zobel compensates the rise in impedance of the driver at high frequencies due to voice coil inductance, as you say usually for the purpose of providing a flatter impedance for a crossover to drive.
If you just place a zobel across a full range driver that has no passive crossover though, and it's connected directly to a low output impedance amplifier, it will cause no difference in the response at all, just the load seen by the amplifier. (In other words its a waste of time unless you have some sort of passive compensation network like BSC in front of the driver)
If you just place a zobel across a full range driver that has no passive crossover though, and it's connected directly to a low output impedance amplifier, it will cause no difference in the response at all, just the load seen by the amplifier. (In other words its a waste of time unless you have some sort of passive compensation network like BSC in front of the driver)
Right. As long as there is significant resistance in the BSC filter, a zobel will help correct a rising response. With a SS amp or low Z tubes,no BSC = no effect.
Bob
2 X
Placing a Zobel in that way on a FR speaker, would actually be putting 2 X Zobels in parallel on the Amp output. As most Amps do have one in place already. Often there can be a wirewound inductor after or before the Zobel, which could make some difference ! How much though ?
Placing a Zobel in that way on a FR speaker, would actually be putting 2 X Zobels in parallel on the Amp output. As most Amps do have one in place already. Often there can be a wirewound inductor after or before the Zobel, which could make some difference ! How much though ?
My knowledge of filter theory is very basic, so I am having a bit of difficulty understanding. If the Zobel network is placed in parallel with the driver, does the Zobel not constitute a high-pass filter, especially if the resistor has a lower value impedance than the driver at its lowest value? It seems to me that there should be some effect, even without a BSC filter.
Can anyone elaborate?
Can anyone elaborate?
I want to clarify my previous post. What I meant to say was that it seems to me that the Zobel network in parallel with the driver would appear to pass some of the high frequencies to ground, thus bypassing the driver at those frequencies. Is that incorrect, and if so, can anyone explain why?
Fraid not.
A filter needs a pass element and a shunt element.
For a low pass filter the pass element is a resistor and/or inductor and the shunt element is a capacitor.
You can change the Filter into a shelf rather than having an extended roll-off by adding a resistor to the shunt capacitor. This is the Zobel you are referring to.
But going back, for the LP filter to be effective you need a pass element. That pass element can be a resistor or an inductor or a combination of these.
If the Power amplifier is the only pass element and it has the usual low output impedance then the pass element is so small that the filtering effect is miniscule.
All the speaker Zobel does is draw extra reactive current through the output stage. This also applies when a crossover is in place. Zobels lower the sensitivity of the crossedover speaker.
A filter needs a pass element and a shunt element.
For a low pass filter the pass element is a resistor and/or inductor and the shunt element is a capacitor.
You can change the Filter into a shelf rather than having an extended roll-off by adding a resistor to the shunt capacitor. This is the Zobel you are referring to.
But going back, for the LP filter to be effective you need a pass element. That pass element can be a resistor or an inductor or a combination of these.
If the Power amplifier is the only pass element and it has the usual low output impedance then the pass element is so small that the filtering effect is miniscule.
All the speaker Zobel does is draw extra reactive current through the output stage. This also applies when a crossover is in place. Zobels lower the sensitivity of the crossedover speaker.
I just finished reviewing the article by Nelson Pass entitled "Current Source Amplifiers and Sensitive/Full-Range Drivers", and this has helped to answer my questions and clear up my confusion.
I had read this article some time ago and had incorrectly recalled some of the ideas presented. I confused some of the techniques illustrated in the article with the use of a Zobel filter and assumed that the Zobel presented some equalization effect where it would not apply.
Thanks to all who responded.
I had read this article some time ago and had incorrectly recalled some of the ideas presented. I confused some of the techniques illustrated in the article with the use of a Zobel filter and assumed that the Zobel presented some equalization effect where it would not apply.
Thanks to all who responded.
I made zobels for my WR system. There is an rc zobel for the highs and an rlc zobel for resonance. This was very effective to cancel out the reactive impedance and lowered the response less than 1/2 dB.
It doesn't seem worth while unless your amplifier is sensitive to load impedance.
It doesn't seem worth while unless your amplifier is sensitive to load impedance.
I realize this is an old thread but thanks to the original poster for asking this question. I was just wondering the same thing myself.
I also came here with a wish to find a solution for a speaker with not so linear response. Thought a Zobel would help... but apparently not.
BG 20 - 8 Ohm - any ideas how to repair the resopnse of this driver without using any kind of DSP? Ideally, it would be done totally speakerside.
BG 20 - 8 Ohm - any ideas how to repair the resopnse of this driver without using any kind of DSP? Ideally, it would be done totally speakerside.
What do you want to repair?
The quality and bandwidth of that speaker is *impressive* for a single unit/full range driver.
Reading that frequency response, it is very smooth (we are talking Real World speakers here and I commend Visaton for that, instead of idealized smoothed brochure type) I can clearlysee 2 contributors whose output is summed:
a) most of the curve shows what the 8" cone puts out, from lowest frequencies to around 9 kHz, where, like cone drivers do, it falls like a brick (usually around 24dB/oct, just check any other speaker around)
b) above that I see an extension, which must come from the added whizzer cone ; it extends frequency response up to about 16kHz which is impressive and in due turn also falls.
A Zobel might correct for impedance although any decent amplifier must drive that speaker with ease, but won´t do an iota to change the acoustic output.
Do you want ruler flat response?
You have 3 main ways to achieve it:
a) (best) add a matching tweeter and crossover, it´s thye best passive way to achieve that.
If so, buy same speaker without the whizzer cone attached so as to avoid any interference with the added tweeter; I bet Visatone also offers same (or if you feel adventurous, cut it off with a sharp blade).
b) technically possible but it carries a penalty: design a "crossover" which attenuates response between 1000 and 9000 Hz, it will leave end to end response incredibly flat.
c) add some EQ to correct what we see.
DSP? Analog? pick your poison 😉
I´m an old school semi-retired Engineer who never updated and too old now to learn new tricks so would do it with a few Op Amps, but be my guest.
My experience is that an optimised zobel directly across the terminals of a FR speaker is very worthwhile, regardless of the amplifier.
Theory states that with a low output impedance/high quality amp there should be no difference, but in practice I find this is not true.
I was running low inductance Canare Star Quad speaker cable, very fast/high power/high quality amp and very high quality CD player source.
The optimising needs to be done with the driver mounted in its enclosure, and is an iterative select on test/measurement process.
The network needs to be mounted directly at the driver terminals and away from the magnet.
I used paralleled values of BP electro and PP capacitors, with one 0.15 uF Polystyrene....this cap was important and completed the improvement.
I also avoided WW resistors and used paralleled 1W MF resistors....the average power dissipated is not actually very high for typical music.
I was able to trim the driver/zobel combination to flat impedance past 40kHz.
The results were that the driver gained subjectively much/nicely cleaner response, much more extended/higher response, and much better depth information.
The amplifier also ran markedly cooler, and the useable acoustic power was markedly increased before objectionable distortions.
All you need is a signal generator, a low value series resistor and a wide frequency range meter of some sort.
The optimising takes some time to home in on the correct values, but is well worth the effort/expense.
Dan.
Last edited:
The 1000-9000hz range could be adjusted down to 90ish db/w with an band shelving filter. A resistor, capacitor and inductor in parallel. Something like 1mH, 6Ω and 1.5μF.
The ~13khz dip could then be adjusted up by bypassing that filter with a inductor and capacitor series. (Something like 0.5mH/0.25μF?)
The ~13khz dip could then be adjusted up by bypassing that filter with a inductor and capacitor series. (Something like 0.5mH/0.25μF?)
@JMFaahey
So you say just stick the driver in a big tqwt and enjoy its performance?
@Max Headroom
Once you say directly on terminals, and then away from the magnet. Do you mean on "box terminals"?
@M8e
Wouldnt that many components affect the sound too much?
So you say just stick the driver in a big tqwt and enjoy its performance?
@Max Headroom
Once you say directly on terminals, and then away from the magnet. Do you mean on "box terminals"?
@M8e
Wouldnt that many components affect the sound too much?
Yes, directly on the driver terminals but bent away from close proximity to the magnet assy...ie, don't tape/glue them to the magnet or basket, IOW away from metals.
Mount the driver with terminals facing down.
Dan.
A zobel can act as RC circuit where Capacitor act as a lowpass and Resistance can be a voltage divider.
It has more purpose than that.
amp zobel is only there to make amp see flat impedance
speaker zobel act as a smooth attenuator which makes the inherent sharp treble come down, correct phase and make overall smooth sound
speaker zobel are RC lowpass and impedance matcher and Amp zobel is only there for impedance mismatch
and if people are moving with time, nowadays solid state amp dont even need a zobel and are also zero ohm output impedance
So i think adding Zobel to speaker is pretty good as an RC circuit
i did it on knowles RAB32063 BA in IEM project as a single fullrange and it somewhat eliminated the need of dampers and resistor damping and fixed 2kHz and trebly peak.
It has more purpose than that.
amp zobel is only there to make amp see flat impedance
speaker zobel act as a smooth attenuator which makes the inherent sharp treble come down, correct phase and make overall smooth sound
speaker zobel are RC lowpass and impedance matcher and Amp zobel is only there for impedance mismatch
and if people are moving with time, nowadays solid state amp dont even need a zobel and are also zero ohm output impedance
So i think adding Zobel to speaker is pretty good as an RC circuit
i did it on knowles RAB32063 BA in IEM project as a single fullrange and it somewhat eliminated the need of dampers and resistor damping and fixed 2kHz and trebly peak.
- Status
- Not open for further replies.