Speakers: Components vs Coaxials

[low_bass_makker]

Originally Posted by santosh.s

Nope, I believe I am NOT confused. I think damping (both speaker or amp) and total Q are different ways to look at one and same thing. The contribution of amp's damping factor (more accurately its output impedance) is to increase electrical Qe of the speaker (and hence its total Q). In ideal case when amp has zero impedance, original Qe will remain unchanged. Even if Qe increases, it may not affect total Q much if Qm was dominating. And I suspect Qm might be dominating in tweeters while Qe in subs.

this is not the point the qms involes suspension, the VC mechincal properties and stiffness of the suspension and lots of other things affect it...

the qes is about the abiity of the speaker impedence and its other electrical properties....

the total Q of the speaker in a enclousre is again in the different thing....

which this you are asking is also confusing....

[MC Mayank]

I installed JBL GTO 607 components both in the front and rear of my Swift in November last year. My HU is a Pioneer 6850. I don't want to compromise on my boot space by fitting a sub. Will putting an amp really make a noticeable difference in quality?

Can I get a good amp for about 4k. And is there any by Alpine which might be compatible? I am totally unaware of amps and their quality but I really do like my music to have a punch without too much of a boom.

[santosh.s]

Originally Posted by DerAlte

I think so too - if you were to make the electrical equivalent circuit of all the elements in the chain, it makes sense.

Yes! It is not yet crystal clear, but here is what I have understood so far- Electrical equivalent circuit is the first level analysis where there is only one R, L & C equivalent for electrical and mechanical parts each (probably, electrical L & C can even be ignored I guess). Everything is obviously assumed to behave linearly. This is actually over-simplified model but it captures all primary factors. If I got it right, Thiele and Small seem to have worked hard to express this model in even more simplified manner (set of T/S parameters) so that any DIY enthusiast can design a good enclosure for a ready made driver unit, without necessarily having to understand any deeper.

In real life, things are more complicated as can be expected! So, there are next levels of accuracy in analysis. Let's not get into those here, but let me mention-

2nd level: the system is still linear but many more resonances can be accounted for. Each part of speaker has its own resonance- spider, cone, surround, enclosure, room... and the list goes on. The target is to make it (frequency response) as flat as possible.

3rd level: Non-linearities are also accounted for, read anything that contributes to THD, IM. It includes non-linear motor movement, cone flapping and the likes.

Even higher level?: Then solve the acoustic wave equation.... for a moving piston in a cylinder of infinite length to begin with, then proceed to more realistic spaces/enclosures....and so on. But that calls for (crystal clear) understanding of "curl" & "div" and things like that, and we are definitely not going to that extent

Sorry, it may look like a show-off, but actually the idea is to get things validated by knowledgeable experts rather than harboring myths!

[santosh.s]

Originally Posted by low_bass_makker

this is not the point the qms involes suspension, the VC mechincal properties and stiffness of the suspension and lots of other things affect it...

the qes is about the abiity of the speaker impedence and its other electrical properties....

To simplify mathematics of the whole system, everything is usually converted to either all-electrical or all-mechanical models.

the total Q of the speaker in a enclousre is again in the different thing....

which this you are asking is also confusing....

Don't get confused, I am talking about the final system, which should contain everything. It includes enclosure if the system has one. Remember, only one parameter is being changed here- and that is amp's output resistance!

[DerAlte]

Originally Posted by santosh.s

Electrical equivalent circuit is the first level analysis where there is only one R, L & C equivalent for electrical and mechanical parts each (probably, electrical L & C can even be ignored I guess). Everything is obviously assumed to behave linearly.

Incorrect assumption - electrical equivalent circuit can take you to any depth, provided you can keep breaking it down to basic components R, L & C. There are mechanical equivalents also (FEM) - but that is for mech engineers. Non-linearity is an 'observation', not a 'happening'.
The "ignore" part is for simplification - poison for analysis.
Even non-linear behaviour can be modelled, provided you can figure out the elements - check with the guys at CPRI. Beyond a certain point it is all Maths!
Shoot, now I will get nightmares all night !

[low_bass_makker]

Originally Posted by santosh.s

Yes! It is not yet crystal clear, but here is what I have understood so far- Electrical equivalent circuit is the first level analysis where there is only one R, L & C equivalent for electrical and mechanical parts each (probably, electrical L & C can even be ignored I guess). Everything is obviously assumed to behave linearly. This is actually over-simplified model but it captures all primary factors. If I got it right, Thiele and Small seem to have worked hard to express this model in even more simplified manner (set of T/S parameters) so that any DIY enthusiast can design a good enclosure for a ready made driver unit, without necessarily having to understand any deeper.

In real life, things are more complicated as can be expected! So, there are next levels of accuracy in analysis. Let's not get into those here, but let me mention-

2nd level: the system is still linear but many more resonances can be accounted for. Each part of speaker has its own resonance- spider, cone, surround, enclosure, room... and the list goes on. The target is to make it (frequency response) as flat as possible.

3rd level: Non-linearities are also accounted for, read anything that contributes to THD, IM. It includes non-linear motor movement, cone flapping and the likes.

Even higher level?: Then solve the acoustic wave equation.... for a moving piston in a cylinder of infinite length to begin with, then proceed to more realistic spaces/enclosures....and so on. But that calls for (crystal clear) understanding of "curl" & "div" and things like that, and we are definitely not going to that extent

Sorry, it may look like a show-off, but actually the idea is to get things validated by knowledgeable experts rather than harboring myths!

that is good but in real the speaker is not at all flat, also it depends of the driver whether it is a woofer or a tweeter.

here is a nice read in developing a driver.

Frontiers

[santosh.s]

Originally Posted by DerAlte

Incorrect assumption - electrical equivalent circuit can take you to any depth, provided you can keep breaking it down to basic components R, L & C.

Aren't you pointing to the same thing which I named as "level 2"? Yes, I agree, we can keep on extracting more and more R,L,Cs (and as the number of elements increase, their values usually keep reducing) and that is called finite-element method of analysis which is used in very diverse fields, including speaker design. I think we have certainly crossed the "scope" now

[santosh.s]

Originally Posted by low_bass_makker

that is good but in real the speaker is not at all flat.

I know, even so called "high-end" speakers are not flat, their THD and IMD figures are never published (and I suspect in many cases these may not even be measured!). But that only makes me so much more curious to understand what high-end really is. Yeh kaun si bala ka nam hai?

here is a nice read in developing a driver.

Frontiers

I happen to have read this site (and many others) as well as article mentioned by Navinji (the 1st one) quite some time back. Thanks for the pointers though.

[santosh.s]

Originally Posted by DerAlte

Incorrect assumption - electrical equivalent circuit can take you to any depth, provided you can keep breaking it down to basic components R, L & C.

I accept that my statement could have been easily mis-interpreted... actually I meant to say extracting only single lumped R, L, & C for the mechanical part is the first level of approximation. I hope we are on the same page now?

[navin]

Originally Posted by MC Mayank

I installed JBL GTO 607 components both in the front and rear of my Swift in November last year. My HU is a Pioneer 6850.
Can I get a good amp for about 4k.

I have not heard the 6850 but I have heard the 6950 so my comments will be based on this expereince.

The 6950 is a very good HU (I hope the 6850 is just as good) and if you are not into playing loud just stick to the HU itself. Maybe you can spend the 4K on damping.

The cheapest 4 ch. branded amp you will find is the Sony 554 (earlier called the 504). This amp is great VFM but it's sound is grainer than the 6950 (I dont know about the 6850). The amp sells for 5K but others here have posted prices as low as 3.8K (remember if you add an amp you will also need some nice RCA wire). so you have to budget for that too.

[low_bass_makker]

Originally Posted by santosh.s

I know, even so called "high-end" speakers are not flat, their THD and IMD figures are never published (and I suspect in many cases these may not even be measured!). But that only make me so much more curious to understand what high-end really is. Yeh kaun si bala ka nam hai?

I happen to have read this site (and many others) as well as article mentioned by Navinji (the 1st one) quite some time back. Thanks for the pointers though.

the high end are those who have maximum flat response and are the best in reproducing the best sound...or in genral which have good name in industry will charge you tons of money say this all go in R&D of this speaker so this is the cost or royalty you have to pay...

In any audio setup the weakest part is the speakers and the software part which is played whichin turn will effect in over all response of any setup. As you have stated before the speaker is having lots of variable involved.

[MC Mayank]

Thanks for the response Navin. So if I ain't playing loud music, what would you recommend I do to improve the sounds quality/clarity.

I really need to find out first how to get some good quality music for cheap though :(

[navin]

Originally Posted by santosh.s

Nope, I believe I am NOT confused. I think damping (both speaker or amp) and total Q are different ways to look at one and same thing. The contribution of amp's damping factor (more accurately its output impedance) is to increase electrical Qe of the speaker (and hence its total Q). In ideal case when amp has zero impedance, original Qe will remain unchanged. Even if Qe increases, it may not affect total Q much if Qm was dominating. And I suspect Qm might be dominating in tweeters while Qe in subs.

Sorry Santosh, there were some wires crossed earlier today. Now I get what you are talking about. I re-read all the posts and even the ones after the one quoted above and not that we are on the same page let me add....

Yes it is obvious that a woofer mechanical Q is much higher (huge spider/surround) in fact in many cases Qms = 10 Qes; and in a tweeter the electrical Q is higher and Qms < Qes or often Qms is 0.5 Qes.

However what happens here is that Qts in a woofer is almost = Qes (since Qms is much higher) and in a tweeter Qts is "about" 0.5 Qes (since Qes and Qms are about equal or atleast not of the order of 10x). Also when the speaker is put in a closed box (lets keep it simple and assume a closed box) the Qtc is always the woofer's Q in the box. The tweeter being sealed is not in the mechanical circuit.

So the damping factor (ratio of speaker's imprdance over amp's impedance) really affects the bass more.

FYI. remember because damping factor is dependant on the amplifer AND speaker in question it is never really just one number (so many manufacturers quote a damping factor approaching 100 as if 100 is some magical number). To a lesser extent DF will also depend on the type of music you play and how loud you play it. So where do manufacturers get this magical number from?
Seach me. I just use this number as a generic guide to explain things to my friends. A high dampoig factor really means a amp with a very low output impedance not nescarrily a good souding amp (just ask any valve amp aficionado).

Santosh, since you have read SL's website You must have also read that of John K. (Zaphaudio). John has tested a bunch of drivers for IM and THD and all sorts of stuff. Fostex's fared rather poorly in these tests.

Der Alte, Tube amps are so dependant on their "iron" it is their weakest link. Even today finding good iron is hard. Building a good OTL is harder still.

LBM, talking about just a maximally flat frequency response does not divluge the full picture. Impluse response, decay etc. are also to be considered.

[santosh.s]

Originally Posted by navin

Sorry Santosh, there were some wires crossed earlier today. Now I get what you are talking about. I re-read all the posts and even the ones after the one quoted above and not that we are on the same page let me add....

First of all no need to say sorry! and thanks for the elaborate response. I don't really see any contradiction now, for most part I can say that I did actually "validate" my understanding with you.

Yes it is obvious that a woofer mechanical Q is much higher (huge spider/surround) in fact in many cases Qms = 10 Qes;

That confirms my suspicion, and the fact that woofers usually need good electrical damping.

and in a tweeter the electrical Q is higher and Qms < Qes or often Qms is 0.5 Qes.

My suspicion is not exactly true here, but that can not be termed as disagreement or misunderstanding of the theory. With this information it still means that effect of electrical damping will not be as pronounced as in woofer. Firstly, for a good tweeter, it should be safe to assume here that Qms must not be far from the ideal 0.707 (as compared to how far it goes in woofers). Secondly, even in the worst case when Qms=Qes (Q=0.5Qms=0.5Qes), total Q can only be raised from 0.5Qms to Qms max. with a really bad amp having infinite output impedance (open circuit). In case of woofers too, Q can potentially be raised till Qms, but since its value is much larger than 0.707 it is possible that very bad amp (HU?) can make it quite resonant.

Also when the speaker is put in a closed box (lets keep it simple and assume a closed box) the Qtc is always the woofer's Q in the box. The tweeter being sealed is not in the mechanical circuit.

I think so far I was (indirectly!) talking about simpler systems with a speaker (either a woofer or a tweeter) directly connected to an amp. Things have already crossed the 1st level of approximation with inclusion of a crossover! I am not really inclined to go any deeper, but in this case instead of saying speakers Q is that of woofer in the enclosure, wouldn't it be best to treat them as two different speakers? How you divide the crossover between woofer and tweeter, whether you consider it as a part of speaker or amp etc. will add to the complexity. I would prefer to leave it at that

(just ask any valve amp aficionado).

I haven't researched valves much, thinking they are outdated technology with a lots of THD/IM and other disadvantages. I know however that certain harmonic distortions can actually make the original sound "sweeter" or more "musical" (have heard samples over net with exaggerated effects!). Many people say that the patterns of harmonic distortion generated by valves can actually appeal to valve lovers, due to this reason. In other words, they have developed and carried over weired tastes since valve days.

Another reason preached in favor of valves is cross-over distortion in transistor amps, dominant in class B and AB at low listening levels. It is supposed to be absent in class A or class D (PWM) amps (and it makes sense to me)

Santosh, since you have read SL's website You must have also read that of John K. (Zaphaudio). John has tested a bunch of drivers for IM and THD and all sorts of stuff. Fostex's fared rather poorly in these tests.

I vaguely remember to have visited Zaphaudio too, but I don't think I have read any THD articles there. Will definitely do that.

I have a few follow up questions about distortion, but I think it is better to postpone them for another time or thread.

Der Alte, Tube amps are so dependant on their "iron" it is their weakest link. Even today finding good iron is hard.

Just curious, what is iron here? (transformers?)

LBM, talking about just a maximally flat frequency response does not divluge the full picture. Impluse response, decay etc. are also to be considered.

Yes, LBM. Maximally flat FR is far from "the" qualifying factor for SQ. It is of course necessary but not at all sufficient. A very flat system can actually have significant non-linear distortion in it which is a different story. First of all, to be technically precise, FR constitutes of both amplitude response and phase response together (Most of the time only amplitude is considered). With a complete FR, it is theoretically possible to get impulse response, decay time etc (and the other way round).

[santosh.s]

A small addition regarding the last statement about converting FR to impulse response- it is valid for linear systems. I have no idea about how it is done in non-linear systems, if at all possible!