Back EMF.....

[timkolody]

Trying to learn as much as I can about how this affects sound and why some amps seem to do a better job handling it than others.

I understand how it is generated and fed back into the amp via speaker wire. I understand that a higher damping factor helps to reduce the effect it has on the amp. What else makes one amplifier better at handling this than another? Is there a way you can control back EMF through speaker design (driver).

All information is greatly appreciated. I have the crown paper on higher DF to reduce back EMF.

Tim

[marlboro]

Read this, and other articles on the same site:

<A HREF="http://sound.westhost.com/impedanc.htm#_damping">http://sound.westhost.com/impedanc.htm#_damping</A>

<A HREF="http://sound.westhost.com">http://sound.westhost.com</A>

[timkolody]

> I am not talking about damping factor, I am talking about back EMF. A high damping factor (low output impedance) on an amplifer is just one way of reducing back EMF, however, there are other factors involved. These other factors are what I am looking for.

As far as the links you posted, I disagree with the gentleman's logic. He has some understanding, but gets tangled in his own pajama's:

"Indeed, my very own (bi-amped) hi-fi uses an amp for the bass-mid with a designed output impedance of about 4 Ohms. This provides a useful extension of the bottom end (I'm using sealed enclosures), without excessive peaking at resonance (there is a bit, and the correct impedance is probably 2 Ohms, but I just haven't gotten around to making the modification yet). To some, this is absolute nonsense - how can you have tight bass with no damping? Easy, damp the enclosure properly, and don't expect someone else (the amp) to do it for you."

The output impedance being higher raises the speakers Q, therefore requiring a larger box. He says that you should damp the enclosure properly, essentially making the driver see a larger box by adding stuffing. What you need to do is match (put into the equation) the output impedance in your amp and speaker cables with speaker design. Essentially you calculate the speaker's required box volume, by inputing all of the output impedance sources(amplifier, speaker cable, and series resistances fo xover components) into the equations used for calculating box volume.

There are other issues with his logic on other topics I perused. Not worth getting into!

Tim

[daryl]

http://img.villagephotos.com/p/2003-4/144711/CurrentSource0.jpg

Provided Link: Daryl's Speaker Stuff. (http://www.villagephotos.com/pubgallery.asp?id_=144711)


Hi Tim,

My guess is you are being lead astray by B.S.

Back EMF is a concept you learn in the beginning stages of electronics.

It is used to explain how devices such as inductors, motors, speaker drivers or any electromagnetic device can impede the flow of electricity.

A motor, speaker or inductors coil may have given resistance but apply a given voltage and less current flows than ohms law would predict at certain times.

Current is slow to start in the inductor, when the speakers diaphram moves it impedes the flow of current and same with the motors shaft.

The explaination is that the device generates it's own voltage in opposition to the applied voltage and that is why less current flows (because their is effectively less voltage applied to the circuit applied EMF - back EMF).

All fine and good.

That is the last you should hear or think about back EMF.

Moving on to the next level of electronics you move on to measuring impedance vs. frequency.

A motor operates at DC so you don't measure it's impedance vs. frequency but understand that it's impedance rises in proportion to it's shafts angular velocity.

An inductor has an impedance that increases with increasing frequency due to it's back EMF.

A speaker driver has an impedance curve with a resonant peak at Fs and rising impedance at high frequency due to voice coil inductance.

You probably have noticed that there are some 'experts' about who are trying to analize a system by following the signal through a system from the amplifier through a crossover to the driver and following back EMF from the driver back to the amplifier.

This is the scientific equivilant to chasing your tail.

Any time you see someone attempting this you can immediately ignore them as they are talking of things way over their head.

To calculate how a driver interacts with a circuit and amplifier you measure it's impedance and then solve the system as a circuit.

You can also model a driver rather than measure it's impedance by substituting an equivilant circuit like this simplified one shown below and then solve the whole thing (you can get more detailed with your equivilant circuit depending upon your needs).

Impedance fully describes back EMF and separate consideration need be given to back EMF.

In fact you cannot consider back EMF separately.

Back EMF has no place outside of beginning electronics.

Once you go beyond that you always use impedance vs. frequency.

So what you have is a drivers impedance and how the amplifier handles it.

Damping factor is a laymans term for output impedance.

The lower an amplifiers output impedance (higher damping factor) the less the ouput will be effected by varying load impedance.

Of course you have to remember that the loop impedance of your speaker wire adds to you amplifiers output impedance and is probably several times greater.

You want the combined output impedance of your amplifier and speaker wire to be low enough so that varying load impedance cannot change frequency response more than is acceptable.

Another issue involving load impedance and amplifiers is stability.

Certain impedances can cause an amplifier to go unstable.

This has to do with how conservative the amplifier is designed (phase margin).

You can design your crossover with conjugates so that you speaker has a flat impedance curve if you want to avoid the effects of varying load impedance upon your amplifier but the additional crossover componets increase cost and complexity.

Daryl

[paul_n_oneal]

Lot's of typing in that topic, thanks! Just to make sure for me are you saying that you can use either the impedance/phase of the speaker or the freq. resp. in the time domain separately to create a xover? I use both and trade off between the two. I have noticed that when the FR phase between two drivers in dang near spot on that the impedance phase at that point is close to 0 degrees. Not sure if the two are interrealted though. Anyway, I like this type of topic and thanks for typing all that.

Paul

[Jeff B.]

> Lot's of typing in that topic, thanks! Just
> to make sure for me are you saying that you
> can use either the impedance/phase of the
> speaker or the freq. resp. in the time
> domain separately to create a xover? I use
> both and trade off between the two. I have
> noticed that when the FR phase between two
> drivers in dang near spot on that the
> impedance phase at that point is close to 0
> degrees. Not sure if the two are
> interrealted though. Anyway, I like this
> type of topic and thanks for typing all
> that.

> Paul

I'm not Daryl, but you must have both impedance with phase and frequency response with phase to create an optimized crossover, and for different reasons. Your crossover's transfer function will be based on the load it is terminated with, and that's the speaker's complex impedance, and don't let anyone tell you that you don't need the phase data too, because you or it will change the transfer function. Secondly, this transfer function is then added to the frequency response to arrive at the final response for the driver. However, there is no relationship whatsoever between phase matching two drivers and the impedance phase at any given frequency - neither has any influence on the other. If you are observing what you describe it just means that your summed impedance must be nearly resistive at the same frequency your drivers are in-phase, but nothing says it will, or needs, to be this way.

[kene]

> Trying to learn as much as I can about how
> this affects sound and why some amps seem to
> do a better job handling it than others.

Given your response to Marlboro's cited article it seems you know plenty already.

Seriously good amplifiers will be able to continue *nearly* doubling their power as speaker impedances are cut in half. Merely mortal amplifiers cannot handle such loading safely/musically and internal resistance is used to protect them from excess current output in really low-impedance loads. This is why when you look at measured ratings into different impedances you see some amps nearly doubling their output into four versus eight ohms while others increase maybe only 30%. In the case of tube amps, the output power will actually drop into lower impedance speakers because of how high their output resistance is, and I'm sure you can run briefly through the math of this and see why for yourself.

In reality if we're talking about standard transistor amplifiers, most manufacturers would LIKE to peddle having perfect power doubling which would indicate zero internal reistance, but they fall short of this mark by similar degrees because of the material compromises they have to make. It's not just a matter of added output impedance, because I've calculated the supposed power outputs for given impedances of amps and they don't match up, so it is also about the power supply, etc. Getting closer to the ideal requires more materials and hence costs more. Diminishing returns sets in and you get the well-known phenomenon of exorbitantly higher prices offering incrementally smaller improvements. Everybody just chooses what price point their willing to work with.

On almost all speakers made DIY or manufactured you have variations in speaker impedance throughout the frequency range.

Daryl posted a thing about how you can design a cross-over to level this phenomenon out and he's mostly correct, though it tends to be too cost-prohibitive for most designers primarily in the low bass where impedances spike substantially at the resonant frequencies. These low frequencies require larger and costlier inductors and capacitors in their notch filters to level out the impedance. What also cannot be easily corrected for is that impedances can drop substantially well below resonance. It is not uncommon in fact for an eight-ohm nominal speaker to vary between 2 and 30 ohms. Run through the calculations and you will quickly see that different amps can vary by several decibels given a real-world speaker load. This is nothing new for some but I'm actually surprised by how few people (even around here) seem to be on to this fact. Stereopile's been measuring this paricular phenomenon for years though and at least for a good while was providing measured responses of amplifiers driving simulated speaker loads. It is no surprise then that many uber-costly amplifiers do in fact retain much more level responses, most notably exhibited in reduced mid-bass humps and superior deep-bass extension.

None of this directly answers your question, but it plays an additional and not-readily-distinguished role in perceived sound quality when you overlay that phenomenon on top of reverse EMF, because both phenomena increase and decrease according to the output impedance i.e.- you don't get one without the other. Reverse EMF in a driver simply causes the driver to stop moving more quickly when a signal is no longer applied. So an amp with a lower internal impedance will do a better job of allowing the driver to damp itself, so to speak. And the same rule applies: exorbitantly higher prices will confer incremental benefits.

So if you're wondering about what happens in practice, it really isn't a whole lot different than what high-end publications claim: as you go to uber-costly amps you get better bass transients and extention on almost any standard speaker. Bass hits harder, seems more authoritative, not only because the deep harmonics are more correctly respresented but because they dissipate more quickly according to the musical signal instead of breaking down into waffley cabinet resonses, etc. Finally the absense of resonances reduces smearing overlay on any following musical signal. Visually you can even see differences in the cone control of the woofers. When I got one of these amps in my system, I noticed power wires in the walls vibrating against the drywall that had never done so before, seemingly at lower volumes. Nothing magical in reality, but pretty impressive to hear. Similar benefits may be derived in other parts of the frequency range if a given loudspeaker has not already compensated for these impedance variations in the higher-frequency drivers, generally meaning better transient detail and so on.

[daryl]

Provided Link: Daryl's Speaker Stuff. (http://www.villagephotos.com/pubgallery.asp?id_=144711)


Hi Paul,

Not shure if I was following your post right.

I was interpeting it the same as Jeff and he pretty well covered that angle.

Daryl