Jbl 4367

[dallasjustice]

Very cool.

If it makes the speaker's sound different, then you ARE changing the speaker's performance. IMO, you would need to know in what way you are altering the speaker's power response to know whether you improved the speaker's performance. I think crossover upgrades on a speaker designed like Harman designs their speakers is fools gold.

I wouldn't redesign the xo, just upgrade parts... Jupiter copper caps, foil inductors and Mills resistors. UPOCC silver or silver/gold alloy wire. Furutech binding posts, maybe Neutrik speakons to go from the external xo to the speaker cabinet.

I built these for a friend's AMT Rock Monitors with the big ESS Heil tweeters several years ago, the improvement over stock xos was massive. Yes, the middle inductor should be on it's side... The caps are Clarity ESA, more budget friendly vs the Jupiter copper caps.

 

[DaveC]

The goal is not to alter response, but to maintain the integrity of the signal so you get better resolution and tone. The parts I recommended are simply higher quality parts that will do less damage to the signal. I would not and have not recommended making any changes to the xo circuit its self. I'd trust JBL can design a xo, but being a pro speaker they aren't using the best parts quality possible. The slightly different specs of the foil inductors and copper foil caps won't make much difference vs the stock parts as far as the electrical characteristics of the xo and may not even be measurable at all. But the result WILL be a large difference in resolution and better tone/timbre.

I actually have experience doing this, I know what the result will be. If you want to experiment with it, as I said the first thing to do would be to simply replace a cap in series with the compression driver with a Jupiter copper cap. If you think that's an improvement it's likely you'll find value in upgrading all the parts and building external xos, if not then great, keep the speakers as-is and you're out $100-200 for the caps and then you'll have some experience hearing the difference between a run of the mill mkp cap and top end copper cap. Also, try getting rid of the steel push-connectors on the compression driver and solder the connection instead. These things make pretty big differences to a lot of folks, it's worth trying.

 

[Folsom]

Very cool.

If it makes the speaker's sound different, then you ARE changing the speaker's performance. IMO, you would need to know in what way you are altering the speaker's power response to know whether you improved the speaker's performance. I think crossover upgrades on a speaker designed like Harman designs their speakers is fools gold.

Michael, yes and no.

You're changing the output, yes. But unless you're changing the value of any of the capacitors, inductors, or resistors, you can nearly fearlessly swap them in and out with anything you want. So long as you stay with the same values nothing at all changes about the power. The crossover slopes, everything, stays exactly the same. You could try hundreds of different capacitors of the same value without a care in the world. But they're all going to sound different.

Upgrading the crossover isn't changing the design. It's more like swapping identically shaped kitchen knifes, one is from Target, the new one is a Wusthof. They could be the same weight, same shape, but you sure as hell can tell the difference when cutting and sharpening.

Contrary to popular belief they do in fact all measure differently. Again we're not talking about the values of them. We're talking about what a sine wave looks like when it goes in one end and comes out the other. Capacitors also have different ESR/ESL despite having the same value. (equivalent series resistance/inductance)

The capacitors on the JBL are cheap. I've got a small collection of them I've removed from other crossovers in order to be replaced with something good. The improvement was clear each time I upgraded them.

But if they came upgraded already you'd be adding $10k on the price tag after markup, given how many there are. The cost to performance isn't there for the average recording studio or such these speaker are originally intended to be used with.

I thought I already posted this, but please take a look at measuring differences between capacitors despite the fact the values and signals are identical. You can literally just scroll to the shots on the scope. This is more like hearing the differences in DAC's that for all intensive purposes measure identical, if we look at them the way a speaker plays, but we know none of them sound the same. But also you have to consider the crossover's distortion has an affect on any amplifier that uses feedback, so the interaction can vary for that too.

Fool's gold? Only so long as all of us audiophiles are fools for buying audiophile equipment to begin with. Upgrading a crossover can be done for a lot less post-purchase of the speakers. It can even be done on the existing board in it.

I understand your sentiment, when you spend that much on high quality speakers you'd like to predict it wouldn't have cheap capacitors (not the cheapest), and that there's no more engineering left to be done but that's just not the case. You're talking about something you just don't understand yet, in ways that make it clear that you're learning about it. It's ok if you just don't care, there's nothing wrong with being content. But I highly recommend paying some attention to the people that do understand this stuff; I wouldn't weigh a belief against people that can write out the math behind it.

 

[dallasjustice]

How do you know whether or not the replacement crossover changed the speaker's frequency response, without measuring the speaker's response?

If it's not measureable, then how do you define a design goal? Are we talking about swap-and-listen engineering?

The goal is not to alter response, but to maintain the integrity of the signal so you get better resolution and tone. The parts I recommended are simply higher quality parts that will do less damage to the signal. I would not and have not recommended making any changes to the xo circuit its self. I'd trust JBL can design a xo, but being a pro speaker they aren't using the best parts quality possible. The slightly different specs of the foil inductors and copper foil caps won't make much difference vs the stock parts as far as the electrical characteristics of the xo and may not even be measurable at all. But the result WILL be a large difference in resolution and better tone/timbre.

I actually have experience doing this, I know what the result will be. If you want to experiment with it, as I said the first thing to do would be to simply replace a cap in series with the compression driver with a Jupiter copper cap. If you think that's an improvement it's likely you'll find value in upgrading all the parts and building external xos, if not then great, keep the speakers as-is and you're out $100-200 for the caps and then you'll have some experience hearing the difference between a run of the mill mkp cap and top end copper cap. Also, try getting rid of the steel push-connectors on the compression driver and solder the connection instead. These things make pretty big differences to a lot of folks, it's worth trying.

 

[dallasjustice]

Well, let's see your measurements. I'd love it if you could demonstrate with real speaker measurements how your would be upgrade actually improves the speaker's performance. I'm guessing you can't do that. If you want to run and hide behind "it can't be measured", that's fine with me. But you wouldn't really know one way or the other since you haven't even tried to measure the loudspeaker.

The underlying assumption you make when claiming that you improved a loudspeaker design is that you also fully understand how the speaker was designed in the first place. Of course, you don't understand that. You don't work at Harman. You don't have access to an anechoic chamber or any of the design software. You aren't familiar with the speaker's polar plots. You don't know how each driver integrates with one another. The waveguide, for example, is designed using sophisticated software which only Harman can use because Voishvillo invented it. The drivers in this speaker are also totally unknown to you since they are 100% designed/manufactured in house.

As Mark Twain said:
"It ain't what you don't know that gets you into trouble. It's what you know for sure that just ain't so."

Michael.

Michael, yes and no.

You're changing the output, yes. But unless you're changing the value of any of the capacitors, inductors, or resistors, you can nearly fearlessly swap them in and out with anything you want. So long as you stay with the same values nothing at all changes about the power. The crossover slopes, everything, stays exactly the same. You could try hundreds of different capacitors of the same value without a care in the world. But they're all going to sound different.

Upgrading the crossover isn't changing the design. It's more like swapping identically shaped kitchen knifes, one is from Target, the new one is a Wusthof. They could be the same weight, same shape, but you sure as hell can tell the difference when cutting and sharpening.

Contrary to popular belief they do in fact all measure differently. Again we're not talking about the values of them. We're talking about what a sine wave looks like when it goes in one end and comes out the other. Capacitors also have different ESR/ESL despite having the same value. (equivalent series resistance/inductance)

The capacitors on the JBL are cheap. I've got a small collection of them I've removed from other crossovers in order to be replaced with something good. The improvement was clear each time I upgraded them.

But if they came upgraded already you'd be adding $10k on the price tag after markup, given how many there are. The cost to performance isn't there for the average recording studio or such these speaker are originally intended to be used with.

I thought I already posted this, but please take a look at measuring differences between capacitors despite the fact the values and signals are identical. You can literally just scroll to the shots on the scope. This is more like hearing the differences in DAC's that for all intensive purposes measure identical, if we look at them the way a speaker plays, but we know none of them sound the same. But also you have to consider the crossover's distortion has an affect on any amplifier that uses feedback, so the interaction can vary for that too.

Fool's gold? Only so long as all of us audiophiles are fools for buying audiophile equipment to begin with. Upgrading a crossover can be done for a lot less post-purchase of the speakers. It can even be done on the existing board in it.

I understand your sentiment, when you spend that much on high quality speakers you'd like to predict it wouldn't have cheap capacitors (not the cheapest), and that there's no more engineering left to be done but that's just not the case. You're talking about something you just don't understand yet, in ways that make it clear that you're learning about it. It's ok if you just don't care, there's nothing wrong with being content. But I highly recommend paying some attention to the people that do understand this stuff; I wouldn't weigh a belief against people that can write out the math behind it.

 

[reverendo]

this is definitely an interesting thread lots of interesting comments
will definitely stay around here.

 

[DaveC]

You're starting to get defensive and are making assumptions again. I don't appreciate it.

If you can open your mind and perform a simple experiment maybe you'll learn something.

 

[fas42]

I would make the simple "chain is as strong as its weakest link" comment here - for a comparatively expensive speaker to use the type of push on connectors shown here is equivalent to using fencing wire to attach parts of a Ferrari engine to each other, IMO ...

 

[Folsom]

Michael you don't understand enough about electricity to make the statements you're making. My knife analogy is a good one. You know the differences in the quality of the steel despite there's no outside factors that are clear measurement differences. When you use it it's very different. And either way you get your meal (sound).

You're not paying attention here that the speaker measurement differences are too small to give you a good impression of what a capacitor swap would do. It's not that it isn't there, it's that quantifying it is so hard that DaveC is telling you it may not be [relatively] measurable. I gave you a specific link to measuring capacitors because it can be measured and you're claiming I'm going to hide behind the recliner with a white flag sticking up that says, "it can't be measured" - the opposite of what I'm saying. That's very rude. My wish is to help you understand, not make you change your speaker's crossovers.

You're 100% correct that I don't have access to Harmon's stuff. What you don't understand is that no one needs any speaker at all to show the difference in capacitors - case in point already shared. I'm talking about the electrical signal. To the circuit the speaker's drivers are just a swinging impedance based on frequency that distorts the signal. This is pretty simple stuff. Your argument is pinged on the believe the capacitor must have a different value in order to sound different and that's not true for the measurement (as shown) or sound. What you must understand is a 10uf capacitor that's MKP and a 10uf capacitor that's Teflon will have the same crossover point when used in a speaker. The same frequencies will go through. But on a small level they'll be different, and while it's hard to see when measuring, or not possible with common speaker measurements with a microphone, the difference that small amount makes is very noticeable subjectively. But that doesn't mean someone will justify the costs when they hear the difference.

This isn't an argument of "will capacitors make my speakers sound different" it's if they make any speaker sound different. The reason why, again, is because we aren't changing the design of anything, we're changing the quality of parts in the design. The subjective appreciable difference will vary some from speaker to speaker, but that's not the question.

 

[Robh3606]

Whoever is recommending that he change the capacitors I have a question. Have you seen the schematic?? Without knowing the topology it isn't all that easy to make recommendations about part changes. The crossover is not your generic crossover type. The capacitors in this design are DC biased.

Rob

 

[sbo6]

Michael you don't understand enough about electricity to make the statements you're making. My knife analogy is a good one. You know the differences in the quality of the steel despite there's no outside factors that are clear measurement differences. When you use it it's very different. And either way you get your meal (sound).

You're not paying attention here that the speaker measurement differences are too small to give you a good impression of what a capacitor swap would do. It's not that it isn't there, it's that quantifying it is so hard that DaveC is telling you it may not be [relatively] measurable. I gave you a specific link to measuring capacitors because it can be measured and you're claiming I'm going to hide behind the recliner with a white flag sticking up that says, "it can't be measured" - the opposite of what I'm saying. That's very rude. My wish is to help you understand, not make you change your speaker's crossovers.

You're 100% correct that I don't have access to Harmon's stuff. What you don't understand is that no one needs any speaker at all to show the difference in capacitors - case in point already shared. I'm talking about the electrical signal. To the circuit the speaker's drivers are just a swinging impedance based on frequency that distorts the signal. This is pretty simple stuff. Your argument is pinged on the believe the capacitor must have a different value in order to sound different and that's not true for the measurement (as shown) or sound. What you must understand is a 10uf capacitor that's MKP and a 10uf capacitor that's Teflon will have the same crossover point when used in a speaker. The same frequencies will go through. But on a small level they'll be different, and while it's hard to see when measuring, or not possible with common speaker measurements with a microphone, the difference that small amount makes is very noticeable subjectively. But that doesn't mean someone will justify the costs when they hear the difference.

This isn't an argument of "will capacitors make my speakers sound different" it's if they make any speaker sound different. The reason why, again, is because we aren't changing the design of anything, we're changing the quality of parts in the design. The subjective appreciable difference will vary some from speaker to speaker, but that's not the question.

2x agree. The proposal is not to change any RLC values, but simply to upgrade to better quality components. Just swapping out the cement resistors will make a significant increase in clarity IME.

Also dallasjustice, congrats on your new speaker endeavor. I commend you on your adventurous spirit and willingness to try new speaker technology, especially when you already had a very good system.

 

[Folsom]

Whoever is recommending that he change the capacitors I have a question. Have you seen the schematic?? Without knowing the topology it isn't all that easy to make recommendations about part changes. The crossover is not your generic crossover type. The capacitors in this design are DC biased.

Rob

It makes zero difference other than extra capacitors are required. I've been thinking about possible exceptions for hours. None come up. Aside from unusually low voltage capacitors any film cap could be a replacement. Obviously some are upgrades, others are not.

In no way do I think Michael should feel like he needs to replace anything. But the possibility is there, and the differences whether or not subjectively worthy, are in fact there.

 

[sbo6]

It makes zero difference other than extra capacitors are required. I've been thinking about possible exceptions for hours. None come up. Aside from unusually low voltage capacitors any film cap could be a replacement. Obviously some are upgrades, others are not.

In no way do I think Michael should feel like he needs to replace anything. But the possibility is there, and the differences whether or not subjectively worthy, are in fact there.

As with almost all products, speakers have a budget and the crossover is a sub - assembly that is commonly an area with tight cost restraints. Looking at the xover in the pic provided this speaker is no different. Cement resistors, tin quick disconnect terminals mounted on a PCB = not high quality. If anyone thinks using higher quality Mills/Vishay resistors, changing to <enter your boutique caps here>, direct wiring of components w/silver solder eliminating the PCB, swapping speaker terminals with high quality WBTs and swapping internal wiring won't make a difference for the better I urge you to hear a speaker before and after with these changes. The folks at Parts Express and DIY audio will agree.

 

[Rodney Gold]

The speaker was measured and voiced using whats in the x/over..by saying you will get a "better" sound means it is different to what was the original designer had in mind
And what is better anyway .. exactly how will the proposed substituted parts change the sound to be
"better"

I would never buy a set of used G1's if they have been fettled by someone who thinks they know more than dickie..same way I wouldnt buy a Ferrari whose enthusiastic previous owner "had a go" at the engine....let alone "have a go" at the engine of my brand new one.

 

[spazmatron]

The speaker was measured and voiced using whats in the x/over..by saying you will get a "better" sound means it is different to what was the original designer had in mind
And what is better anyway .. exactly how will the proposed substituted parts change the sound to be
"better"

I would never buy a set of used G1's if they have been fettled by someone who thinks they know more than dickie..same way I wouldnt buy a Ferrari whose enthusiastic previous owner "had a go" at the engine....let alone "have a go" at the engine of my brand new one.

yea, with tweaks your leaving yourself open to making things different in a unwelcomed way.. its a risk, but if the x-over is simple and easy to get to then why not try a few of the 'better' components. Engineering based companies like JBL and Vivid design by controlling understood variables and use known, explainable engineering principles rather than 'unknown' audiophile tweaks.

this is also the approach michael has chosen to invest in as i understand it. good on him, from what i can make out he has been down 'the other path' and is having much more success with his current philosophy. explainable engineering by guys on the cutting edge like bruno and the JBL guys, coupled with the latest understandings of room acoustics.

keep on keeping on michael, i like the way your doing things

 

[Folsom]

The speaker was measured and voiced using whats in the x/over..by saying you will get a "better" sound means it is different to what was the original designer had in mind
And what is better anyway .. exactly how will the proposed substituted parts change the sound to be
"better"

I would never buy a set of used G1's if they have been fettled by someone who thinks they know more than dickie..same way I wouldnt buy a Ferrari whose enthusiastic previous owner "had a go" at the engine....let alone "have a go" at the engine of my brand new one.

Not correct. It wasn't voiced to those capacitors. They're just what Harmon uses; because they're adequate and cheap; they're in an awful lot of speakers. It was designed to work with the values of capacitors to achieve specific slopes, notches, etc, that's all. No designer would design with them in mind specifically unless the objective was to smooth out signal. Even Harmon knows they're not that good because they bypass the large caps with small ones. The one thing Harmon did do to "voice" the speaker was the DC bias on the capacitors for the tweeter which makes more even distortion for a slightly nicer to listen to treble. Harmon spends their money on incredible drivers and wave guides.

Interesting example. The G1's don't use expensive capacitors either. But they use an unusual amount of them. The price would increase by thousands to upgrade. Potentially as high as doubling the cost of the speaker. They're still probably better than the variety in JBL by a margin. And with them I'd believe that possibly they were chosen to tame the aluminum drivers a touch.

Changing capacitors, resistors, or inductors on the G1's isn't a statement of belief in knowing more than Dickie. Frankly that's an irrelevant statement given that if someone replaces the capacitors, resistors, and inductors with higher quality version, while using the same values, they haven't changed the design at all. Obviously someone needs to be competent at soldering to perform this action. If you can't quantify if the job was done correctly then check with someone that can. Will it be better? Matter of opinion; that generally points towards yes.

A better analogy for a Ferrari would be like going to store to get some very high performance tires to replace the stock rubber. They do the same function but you're going to have a preference. And beyond that you can have someone who knows how to mount a tire do it; you don't have to do it yourself but could learn to do so. You don't have to go to Ferrari to accomplish this and changing it isn't making a statement you know better than Ferrari. It isn't changing the design of the car, it's just giving a different feel to the same exact design so long as the tires are the right size.

 

[spazmatron]

Not correct. It wasn't voiced to those capacitors. They're just what Harmon uses; because they're adequate and cheap; they're in an awful lot of speakers. It was designed to work with the values of capacitors to achieve specific slopes, notches, etc, that's all. No designer would design with them in mind specifically unless the objective was to smooth out signal. Even Harmon knows they're not that good because they bypass the large caps with small ones. The one thing Harmon did do to "voice" the speaker was the DC bias on the capacitors for the tweeter which makes more even distortion for a slightly nicer to listen to treble. Harmon spends their money on incredible drivers and wave guides.

Interesting example. The G1's don't use expensive capacitors either. But they use an unusual amount of them. The price would increase by thousands to upgrade. Potentially as high as doubling the cost of the speaker. They're still probably better than the variety in JBL by a margin. And with them I'd believe that possibly they were chosen to tame the aluminum drivers a touch.

Changing capacitors, resistors, or inductors on the G1's isn't a statement of belief in knowing more than Dickie. Frankly that's an irrelevant statement given that if someone replaces the capacitors, resistors, and inductors with higher quality version, while using the same values, they haven't changed the design at all. Obviously someone needs to be competent at soldering to perform this action. If you can't quantify if the job was done correctly then check with someone that can. Will it be better? Matter of opinion; that generally points towards yes.

A better analogy for a Ferrari would be like going to store to get some very high performance tires to replace the stock rubber. They do the same function but you're going to have a preference. And beyond that you can have someone who knows how to mount a tire do it; you don't have to do it yourself but could learn to do so. You don't have to go to Ferrari to accomplish this and changing it isn't making a statement you know better than Ferrari. It isn't changing the design of the car, it's just giving a different feel to the same exact design so long as the tires are the right size.

not the case. the brand of x-over component was not chosen for this reason..dic has used bennic for years (at B&W too) and is happy with their performance vs audiophile caps and resistors i believe.

every one knows if you want to make speakers that sound good you have to have them blessed by god and sacrifice a goat on a bimonthly basis.. luckily both JBL and Vivid do this so michael wont have to do take any retrospective action... put the knife down michael, free the goat from the back yard.. its ok! the blessing was preformed at the factory.

 

[Folsom]

My comment with Bennic and Vivid isn't based on parameter.

 

[maxima95]

Not that it makes much difference to this prong of the discussion, but the crossover is in 2 parts (2nd part below).

That said, if these were my speakers, fiddling with the crossover would be a lot further down the road, if at all.
New speaker stands are on tap. There is a lot of listening, acclimatization and other things to do over time.

Subsequently, if interested, there may be some information available (search, forums, word of mouth, ...) on results of crossover modifications with this speaker. I would seek this out first.

 

[dallasjustice]

Since I've been told I don't anything about electricity, I've made an effort to learn about capacitor distortion in crossover networks. Believe it or not, there are AES papers on this topic. Here is a good starting place for those who wish to learn more:
http://www.aes.org/e-lib/browse.cfm?elib=14630

The title of the paper is nonlinear distortions in capacitors. The paper breaks down an extreme scenario where a capacitor used for a high pass filter can exhibit distortion. The things I learned from this paper are:
1. One would need to know the resonant frequency for the capacitor. Without this information, it's impossible to know whether a cap can be made better with a replacement.
2. One also needs to know the amount of signal current needed to make the capacitor distort.

The example used for this paper was a very extreme scenario so I wouldn't assume it relevant for any particular loudspeaker design.

As has been pointed out before, these crossver networks are DC biased. All you parts jockeys should also know that these drivers are very efficient. Therefore, very little current ever passes through the high pass filter, even under extreme scenarios.

Unlike the parts jockeys, I don't have all the answers. But I do know one should have a better understanding of how a speaker system works and be in a position to take distortion measurements before and after to know whether a capacitor change will have any positive effect on the loudspeaker's output signal.