Am I reading this graph correctly?

[nordhaven]

Jay was nice enough to model this crossover up for me. I am new to all this can someone tell me what the graphs mean. I'm guessing the speaker impedance is 6 ohms and the efficiency is around 87db? what else can extruded from these measurements? Jay is gone for a few days or I would ask him.

[Pete Schumacher ®]

Jay was nice enough to model this crossover up for me. I am new to all this can someone tell me what the graphs mean. I'm guessing the speaker impedance is 6 ohms and the efficiency is around 87db? what else can extruded from these measurements? Jay is gone for a few days or I would ask him.

You've got a firm grasp on the reading. You can also see that the "reverse null" in the summed response chart shows good phase tracking in the XO region.

Looks very promising and should make for a nice sounding design.

[nordhaven]

On the schematic of the crossover layout there is a + sign behind the two drivers. Witch end or line back to source is the + and witch is the -? Another way of asking is. If looking at the symbol for the source is the line coming out the top + or-?

[Pete Schumacher ®]

On the schematic of the crossover layout there is a + sign behind the two drivers. Witch end or line back to source is the + and witch is the -? Another way of asking is. If looking at the symbol for the source is the line coming out the top + or-?

The plus sign just means that the drivers are wired in positive phase. Usually, that means you connect the + side of the input signal to the + terminal on the driver, and usually, the crossover circuit is in between the + input and the + driver terminal.

[lhwidget]

Yep, just what Pete is saying, the short hand in the schematic above is that the drivers are connected in phase with each other, the bottom connection dots for the drivers are the drivers' - terminal connections, the top connection dots are the drivers' + terminal connections.

Following the conductor lines back to the source, the top connection dot at the source is the + speaker terminal on the amp, and the bottom connection dot at the source is the - speaker terminal on the amp.

Jay T

[nordhaven]

Thanks guys that clears things up!

[ajinfla]

Looks very promising and should make for a nice sounding design.

Hi Pete,

You can tell how a loudspeaker might sound from a 1m on axis simulation? Please enlighten me .

cheers,

AJ

[Pete Schumacher ®]

Hi Pete,

You can tell how a loudspeaker might sound from a 1m on axis simulation? Please enlighten me .

cheers,

AJ

Good components, smooth response. The makings of a nice system . . .

Come on now AJ. Haven't you been following DIY for the past couple years? Simulations with good methodology and accurate infinite baffle response plots can come very close, to within a dB or so, of the actual final system response. Close enough to tweak the rest by ear with small changes to one or two components.

[nordhaven]

Good components, smooth response. The makings of a nice system . . .

Come on now AJ. Haven't you been following DIY for the past couple years? Simulations with good methodology and accurate infinite baffle response plots can come very close, to within a dB or so, of the actual final system response. Close enough to tweak the rest by ear with small changes to one or two components.

Thats what Jay recomended.
The crossover occurs around 2.3 kHz, which should be optimal for this driver combination. If you're willing to (I usually recommend), you can fine-tune the tweeter network by your ear. In the crossover schematic, I suggested ranges of component values you may want to try. If not a perfectionist, I'd tweak only the tweeter padding (R9) and go with the default values for C1.

[ajinfla]

Come on now AJ. Haven't you been following DIY for the past couple years? Simulations with good methodology and accurate infinite baffle response plots can come very close, to within a dB or so, of the actual final system response. Close enough to tweak the rest by ear with small changes to one or two components.

Not sure what relevance the first question has and I'm not doubting that the measured (which probably won't happen) 1m on axis response won't be close to the sim. My question was how you determined what the system might sound like from a single snapshot in the huge photo album. From my experience that is inadequate.

Good components, smooth response. The makings of a nice system . . .

Fair enough, components seem reasonable.
Smooth response at the tip of a mic 1m directly in front of the speaker? Don't countless systems have that (as they were such designed)? This is sufficient? I certainly couldn't give you an educated guess as to how it might sound without much more data, so I was hoping to learn how you could. Thanks.

Cheers,

AJ

[Pete Schumacher ®]

Not sure what relevance the first question has and I'm not doubting that the measured (which probably won't happen) 1m on axis response won't be close to the sim. My question was how you determined what the system might sound like from a single snapshot in the huge photo album. From my experience that is inadequate.



Fair enough, components seem reasonable.
Smooth response at the tip of a mic 1m directly in front of the speaker? Don't countless systems have that (as they were such designed)? This is sufficient? I certainly couldn't give you an educated guess as to how it might sound without much more data, so I was hoping to learn how you could. Thanks.

Cheers,

AJ

It's really simple AJ. If you have heard the driver involved before, you know if you liked it or not. If the implementation in a new system doesn't put that driver in an untenable position, such as a tweeter crossing too low, the chances for success are fairly good. Tweaking the final response to suit the environment they'll be used in always varies. Getting a good starting point to begin that process is what really matters. Saying something encouraging like, "you have the makings of a fine sounding system there" or "should sound pretty nice when you're done" sure beats the heck out of "don't even try you noob, you've never built anything before, and if you don't measure, you'll never be happy." I really do think a number of folks around here have done just fine designing by ear. I know Maynard's latest Iowa display attests to that fact.

So, I'll continue to encourage someone who's attempting a system that uses quality parts with statements like, "should make for a nice sounding system" and not worry in the least that I'd not like the end result.

I like apples. But by golly, I like grapes too!!!

[jclin4]

jAy worked up a crossover for me based on simulations for a TB W5-704D / Vifa XT25 MTM. I went ahead and built a crossover with different cap values (based on my own errant simulations...but that is a different story).

I voiced the speakers for several weeks, tweaking the values until finally they ended up where jAy thought I should have started. His XO was right on the money.

I'll post project details later in a different thread (I'm dealing with a hand injury and cannot type too much), but I just wanted to add that jAy's simulation and design work is topnotch.

- John

[ajinfla]

It's really simple AJ. If you have heard the driver involved before, you know if you liked it or not. If the implementation in a new system doesn't put that driver in an untenable position, such as a tweeter crossing too low, the chances for success are fairly good. Tweaking the final response to suit the environment they'll be used in always varies. Getting a good starting point to begin that process is what really matters. Saying something encouraging like, "you have the makings of a fine sounding system there" or "should sound pretty nice when you're done" sure beats the heck out of "don't even try you noob, you've never built anything before, and if you don't measure, you'll never be happy." I really do think a number of folks around here have done just fine designing by ear. I know Maynard's latest Iowa display attests to that fact.

So, I'll continue to encourage someone who's attempting a system that uses quality parts with statements like, "should make for a nice sounding system" and not worry in the least that I'd not like the end result.

I like apples. But by golly, I like grapes too!!!

Gotcha. Hope the leg's feeling better. Been there myself more than once (still dumb enough to ride the dirt bikes but I'm done with the football and rugby )

cheers,

AJ

[Jeff B.]

Not sure what relevance the first question has and I'm not doubting that the measured (which probably won't happen) 1m on axis response won't be close to the sim. My question was how you determined what the system might sound like from a single snapshot in the huge photo album. From my experience that is inadequate.



Fair enough, components seem reasonable.
Smooth response at the tip of a mic 1m directly in front of the speaker? Don't countless systems have that (as they were such designed)? This is sufficient? I certainly couldn't give you an educated guess as to how it might sound without much more data, so I was hoping to learn how you could. Thanks.

Cheers,

AJ

Keep in mind that first arrival on-axis response is by far the greatest determining factor in how we perceive a speaker's sound.

[jkim]

Jay was nice enough to model this crossover up for me. I am new to all this can someone tell me what the graphs mean. I'm guessing the speaker impedance is 6 ohms and the efficiency is around 87db? what else can extruded from these measurements? Jay is gone for a few days or I would ask him.

Hey, I'm on a trip with my family in Chicago area. I thought I couldn't be online since I weren't bringing my laptop and didn't want to spend money on buying internet acess. But this hotel I got from priceline is awesome! Sweet suite style room with divided spaces with two TVs, indoor pool, very clean, luxurious feel, and computers with free internet! But only for 45 minutes at a time. But I can relogin after a session. All these are at $39 per night!

Yes, the speaker's system impedance curve is a typical, nominal 6-8 ohm speaker's. The Tang Band W6-789 is very sensitive for a nominal 8 ohm 6.5" hifi driver that can provide good bass. This is respectable even compared to 4 ohm drivers. Its 92 dB sensitvity on the spec is not that far from real. I know this because I designed a crossover for this driver and the builder told me that the tweeter-midwoofer balance as I designed was nice. This meant that the actual sensitivity of this driver must be at least 91 dB.

The final system sensitivity is about 87 dB in the mids. It's decreased from 91 dB because I used about 4 dB baffle step compensation. You may not understand this terminology very well if you don't have much experience in crossover designing. Speakers in a typical room, placed away from walls, are affected by what we call baffle step loss, which is gradual 6 dB loss as the frequency decreases (in fact, full 6 dB only in anechoic chamber). To compensate this, we gradually attenuate the midwoofer's midrange as the frequency increases. This amount of attenuation is called the amount of baffle step compensation. I used about 4 dB attenuation. A reason why I (and people) don't use full 6 dB compensation is that we have some amount of boost in the bass/midbass due to room boundary effect, which prevents full 6 dB loss in practice.

About the driver polarity, I already gave an answer in my email. So did people in the above.

Going back to hotel room...

-jAy

[nordhaven]

It's really simple AJ. If you have heard the driver involved before, you know if you liked it or not. If the implementation in a new system doesn't put that driver in an untenable position, such as a tweeter crossing too low, the chances for success are fairly good. Tweaking the final response to suit the environment they'll be used in always varies. Getting a good starting point to begin that process is what really matters. Saying something encouraging like, "you have the makings of a fine sounding system there" or "should sound pretty nice when you're done" sure beats the heck out of "don't even try you noob, you've never built anything before, and if you don't measure, you'll never be happy." I really do think a number of folks around here have done just fine designing by ear. I know Maynard's latest Iowa display attests to that fact.

So, I'll continue to encourage someone who's attempting a system that uses quality parts with statements like, "should make for a nice sounding system" and not worry in the least that I'd not like the end result.

I like apples. But by golly, I like grapes too!!!

Thanks for the encouragement!

[nordhaven]

Hey, I'm on a trip with my family in Chicago area. I thought I couldn't be online since I weren't bringing my laptop and didn't want to spend money on buying internet acess. But this hotel I got from priceline is awesome! Sweet suite style room with divided spaces with two TVs, indoor pool, very clean, luxurious feel, and computers with free internet! But only for 45 minutes at a time. But I can relogin after a session. All these are at $39 per night!

Yes, the speaker's system impedance curve is a typical, nominal 6-8 ohm speaker's. The Tang Band W6-789 is very sensitive for a nominal 8 ohm 6.5" hifi driver that can provide good bass. This is respectable even compared to 4 ohm drivers. Its 92 dB sensitvity on the spec is not that far from real. I know this because I designed a crossover for this driver and the builder told me that the tweeter-midwoofer balance as I designed was nice. This meant that the actual sensitivity of this driver must be at least 91 dB.

The final system sensitivity is about 87 dB in the mids. It's decreased from 91 dB because I used about 4 dB baffle step compensation. You may not understand this terminology very well if you don't have much experience in crossover designing. Speakers in a typical room, placed away from walls, are affected by what we call baffle step loss, which is gradual 6 dB loss as the frequency decreases (in fact, full 6 dB only in anechoic chamber). To compensate this, we gradually attenuate the midwoofer's midrange as the frequency increases. This amount of attenuation is called the amount of baffle step compensation. I used about 4 dB attenuation. A reason why I (and people) don't use full 6 dB compensation is that we have some amount of boost in the bass/midbass due to room boundary effect, which prevents full 6 dB loss in practice.

About the driver polarity, I already gave an answer in my email. So did people in the above.

Going back to hotel room...

-jAy

Jay thanks again for the help!!! I'm sure they will turn out great. Any tweaking would be subjective on my part. I don't have a perfect reference system nor do I presume to know perfect sound. I'm sure they will be better then anything I've bought at my budget. Thanks again for the help!