New super-budget 2 way: The Orient Express

[Paul Carmody]

I thought about these two drivers I had procured for pretty much no money... they looked really well-built, but how well could they perform--could they perform together? Well, why not build a box and try it?
The Woofer is the Silver Flute W14RC25 wool cone, the tweeter is the Vifa Logic AC25SG05. Both are super-cheap (and available only) over at that Wisconsin speaker store.
The box is 11 Liters, vented out the front. I'm building these for a friend, who wants some actual bookshelf speakers, he even took pictures of the bookshelves he wants them on. He wants them painted white to "disappear" into the room. I'm fine with that.

So the $11 question: how do they sound? Well, after measuring everything up close against the wall, I played around with several different topologies. Seems these two drivers have certain slopes they want to "fall into"--but finding a combination the two could agree on was difficult.

I won't post the crossover yet, because what Curt said the other day is right. But I can tell you at this point that the woofers have a satisfying bottom end, and a surprisingly low F3 that I didn't really even fully take advantage of. The Vifa tweeter, depending on how it's crossed, can sound harsh. But with the woofer carrying a bit more of the weight, the Vifa started sounding cleaner and more natural.

I had a really hard time getting an in-phase crossover to work with these drivers. So I finally resigned myself to phase quadrature, and I'm really surprised with how effortless and smooth everything started sounding. I've said this before, but I think a lot of us are uncomfortable with phase quadrature, because we can't flip the polarity and see that nice, satisfying giant null at the XO point. Instead, you flip the polarity and things still look flat, just in a yucky sort of way.

OK, enough rambling. So far, this project has definitely rounded 2nd base. I'll post the XO in a few, when I get a more time with it. By the way, this is definitely the fastest I've ever put together a speaker; even my wife pointed that out. I explained to her that I've learned a few tricks (cutting pieces slightly large then using flush trim bit = smart move, $20 pneumatic nail gun from Harbor Freight = BRILLIANT). So I figured the name Orient Express was fitting, given the production time, and country of origin for the drivers.

[shawn_a]

That works for me. Ya kept me up long enough to see what was so important and I'm glad. I've got a pair of those tweeters just wanting a home. I think I've found it now. Or that is unless someone wants to buy them from me. Still I might have to look in to this, Paul. Isn't it amazing what can be done when you get the tools to do the job more efficiently?

shawn

[Wolf]

That works for me. Ya kept me up long enough to see what was so important and I'm glad. I've got a pair of those tweeters just wanting a home. I think I've found it now. Or that is unless someone wants to buy them from me. Still I might have to look in to this, Paul. Isn't it amazing what can be done when you get the tools to do the job more efficiently?

shawn

As I recall, you have the woofers too!
Wolf

[Paul Carmody]

I'm not saying this is the final design. It's just that I have come across a strange situation and I don't know what to make of it.

As I said before, these drivers seemed to have certain acoustic slopes they wanted to do, but had a hard time agreeing with one another on compatible slopes. Here I have created an asymmetrical slope... but since that's a topic not really covered in books, I feel like I'm on my own figuring things out.

The woofer has a 2LR acoustic slope @ 2800 Hz
The tweeter has a 3BW acoustic slope @ 2800 Hz

They are very much in phase u pthrough the XO point, up to about 4 KHz, where I just can't keep them together anymore. Reversing the polarity creates a massive null at 2800 Hz. (so much for the phase quadrature idea) The result actually sounds really good; imaging is extremely accurate through the midrange and lower treble (I guess the graph sort of proves that).

So I guess my question is: is this a valid acoustic XO slope cobination? I haven't been around to have a sort of mental catalogue of various XO slopes (just this cookbook, which keeps things simple and symmetrical). I'd heard that asymmetrical acoustic slopes are sometimes worth a try, so that's what I did. Everything seems to sum flat--or is it maybe +a few dB? The power response seems to sum 3 dB above the XO point, and that's all I'm really worried about.

Electrically, I wanted to try to keep the Low Pass a 2nd order and the High Pass a 3rd order because I remember Jeff mentioning that the extra phase rotation inherent in the HP filter can help accomodate for acoustic offset. It sounded like a really good idea... hope I heard you right, Jeff!

So Curt, Jeff, or whomever is really good with passive filters, I'd love it if you could take a few minutes to look this over.

Thanks

[Wolf]

I'm not saying this is the final design. It's just that I have come across a strange situation and I don't know what to make of it.

As I said before, these drivers seemed to have certain acoustic slopes they wanted to do, but had a hard time agreeing with one another on compatible slopes. Here I have created an asymmetrical slope... but since that's a topic not really covered in books, I feel like I'm on my own figuring things out.

The woofer has a 2LR acoustic slope @ 2800 Hz
The tweeter has a 3BW acoustic slope @ 2800 Hz

They are very much in phase u pthrough the XO point, up to about 4 KHz, where I just can't keep them together anymore. Reversing the polarity creates a massive null at 2800 Hz. (so much for the phase quadrature idea) The result actually sounds really good; imaging is extremely accurate through the midrange and lower treble (I guess the graph sort of proves that).

So I guess my question is: is this a valid acoustic XO slope cobination? I haven't been around to have a sort of mental catalogue of various XO slopes (just this cookbook, which keeps things simple and symmetrical). I'd heard that asymmetrical acoustic slopes are sometimes worth a try, so that's what I did. Everything seems to sum flat--or is it maybe +a few dB? The power response seems to sum 3 dB above the XO point, and that's all I'm really worried about.

Electrically, I wanted to try to keep the Low Pass a 2nd order and the High Pass a 3rd order because I remember Jeff mentioning that the extra phase rotation inherent in the HP filter can help accomodate for acoustic offset. It sounded like a really good idea... hope I heard you right, Jeff!

So Curt, Jeff, or whomever is really good with passive filters, I'd love it if you could take a few minutes to look this over.

Thanks

What do your targets look like in comparison? It looks like 2nd order rolloff on the tweeter too, regardless of filter. If you shift your values a hair, the phase will lock in place, and they will coincide. For me, as long as the 2 vertical phase-wrapped lines are close, I let it go, but you are close enough to fiddle just a hair more. If your physical offsets aren't figured, you'll throw it off majorly anyway after they are entered.

Asymmetrical slopes are fine as long as the summation works. Your power response *looks* fine, but the bump at the XO means there will be a bump off-axis, and that is not going to be favorable. I'm no guru, but I'd fiddle with it some more.
Later,
Wolf

[Jeff B.]

I'm not saying this is the final design. It's just that I have come across a strange situation and I don't know what to make of it.

As I said before, these drivers seemed to have certain acoustic slopes they wanted to do, but had a hard time agreeing with one another on compatible slopes. Here I have created an asymmetrical slope... but since that's a topic not really covered in books, I feel like I'm on my own figuring things out.

The woofer has a 2LR acoustic slope @ 2800 Hz
The tweeter has a 3BW acoustic slope @ 2800 Hz

They are very much in phase u pthrough the XO point, up to about 4 KHz, where I just can't keep them together anymore. Reversing the polarity creates a massive null at 2800 Hz. (so much for the phase quadrature idea) The result actually sounds really good; imaging is extremely accurate through the midrange and lower treble (I guess the graph sort of proves that).

So I guess my question is: is this a valid acoustic XO slope cobination? I haven't been around to have a sort of mental catalogue of various XO slopes (just this cookbook, which keeps things simple and symmetrical). I'd heard that asymmetrical acoustic slopes are sometimes worth a try, so that's what I did. Everything seems to sum flat--or is it maybe +a few dB? The power response seems to sum 3 dB above the XO point, and that's all I'm really worried about.

Electrically, I wanted to try to keep the Low Pass a 2nd order and the High Pass a 3rd order because I remember Jeff mentioning that the extra phase rotation inherent in the HP filter can help accomodate for acoustic offset. It sounded like a really good idea... hope I heard you right, Jeff!

So Curt, Jeff, or whomever is really good with passive filters, I'd love it if you could take a few minutes to look this over.

Thanks

Yes, you heard me correctly. I've mentioned this a number of times, and so have several others. I think I even sent you a link to my 2000 article that talks about using asymmetrical slopes and why.

Not only are asymmetrical slopes a workable idea, it is actually what everyone HAS to arrive at if there is much acoustic offset between the two drivers. Targeting symmetrical slopes with drivers that have 1" of offset and a crossover at 2kHz or higher will never result in a flat summation. You have to change the slopes in order to adjust the phase relationship so the summation is correct.

If people model their designs without offsets then their model will arrive at symmetrical slopes and flat summation, but if their speaker actually has the woofer and tweeter mounted flush on the baffle with an offset then there will be a dip at the crossover. They may like the sound of this dip though, and feel their design was successful.

On the other hand, any one targeting flat response and entering offsets, and basically ignoring target slopes, will simply adjust their slopes to get flat response and nearly every time will relax the slope on the woofer to make it asymmetrical. You can also increase the slope on the tweeter to accomplish the same thing. The point is to adjust the phase rotation to bring them to alignment. As for me, I ignore target slopes and don't usually think in these terms while designing, except to check to see what I am arriving at toward the end.

Your phase relationship looks excellent, you can not do much better than this with offset drivers. And, with offset drivers you can not maintain a perfect phase match over a wide bandwidth either, the key is in the crossover region. You have hit that fine.

Jeff

[Paul Carmody]

Yes, you heard me correctly. I've mentioned this a number of times, and so have several others. I think I even sent you a link to my 2000 article that talks about using asymmetrical slopes and why.

AH! That's where I read it! See, I was paying attention.

Your phase relationship looks excellent, you can not do much better than this with offset drivers. And, with offset drivers you can not maintain a perfect phase match over a wide bandwidth either, the key is in the crossover region. You have hit that fine.

Thanks. I really appreciate you taking the time to look this over and answer my question. it's really nice to have a knowledgeable set of eyes around. I had a feeling I was at the crux of a teachable moment!

Thanks also to Jay for the mini-essay. It made a lot of sense. You should post that up on your website.

[jkim]

Here's my write-up about this issue: dealing with driver offset on flat baffle. Someone asked this question a while ago and this was my reply to him.

-------------------------------------------

... As far as I know, there's no book or article that explains in detail how we can deal with driver offsets on a flat baffle in a crossover design, and why the method works. For acoustic LR4 designs, the method is really simple. In short, on a flat baffle, we don want to use theoretical LR4 rolloffs due to the driver offset (i.e., drivers are not time aligned; the midwoofer is farther from the listenter than the tweeter). What we normally do is "relax" the drivers' rolloff rates.

To see how this works, consider the following fact. When a midwoofer is low-passed, there occurs an increasing "phase lag" over a frequency range. This delay begins far before the crossover frequency and gets larger as the frequeny increases. A fact is, as the rolloff rate gets higher, the phase lag rate also gets higher. For the theoretical LR4 low-pass filter rolloff, the delay amount reaches 180 degrees at the crossover frequency. The opposite happens with a high-pass filter. When a tweeter is high-passed, there occurs a "phase lead." This phase lead also begins far before the crossover frequency and gets larger as the frequency decreases. As the rolloff rate gets higher, the phase lead also gets higher. For the theoretical high-pass rolloff, the lead amount reaches 180 degrees at the crossover frequency. Then what's the phase difference between low and high-pass rolloffs at the crossover point? Yes, 360 degrees. That is, the sound wave from the woofer is exactly one wavelength slower than that from the tweeter. That's why we have a full, 6 dB gain (double the SPL of each driver) at the xover frequency.

Now, what will happen if the midwoofer is a little farther from the listener than the tweeter? Then, there will be an additional amount of phase lag for the midwoofer, or equivalently, an additional amount of lead for the tweeter. As a result, their phase is misaligned. The solution is simple. If we relax the rolloff rate of either the midwoofer or the tweeter a little bit, the corresponding lag or lead will decrease. That's why we call this technique "asymmetric LR4." But this is a misnomer because, sometimes, we can relax both drivers' rolloff rates. In pratice, however, people often want to relax only the woofer's rolloff due to the power handling issue of the tweeter. But this is not always the case.

What will happen if we don't use this technique but use theoretical, exact LR4 rolloffs on a flat baffle? Then, the "in-phase" listening axis that ensures phase alignment around the xover point will be located below the speaker, often times below the woofer axis. Why? Because that is a point where the tweeter and midwoofer are time-aligned for the theoretical LR4 to work.

Lastly, this technique usually don't work for LR acoustic 2nd order crossovers. The reason is, 2nd order rolloffs are already very shallow, not leaving a room for further rolloff rate reduction due to the drivers' naturall, limited bandwidth rolloffs.

Note that we can use a flat baffle without worrying much about this phase alignment between woofer and midrange in a 3-way. The reason is that at a lower frequency wavelength is long enough so that the driver offset won't cause a large phase misalignment.

Hope this explanation helps you to better understand this issue.

-jAy

[trevorg]

As I recall, you have the woofers too!
Wolf

As do I... Man i need to do something with em... Well my car comes outa the shop on Monday, and is about 200$ dollar away from running, new job hopefully comming next week... May be time to get back into things...

[nick29498141]

Split-Slot Port?

Looks nice, but what's the reasoning?

[Paul Carmody]

Split-Slot Port?
Looks nice, but what's the reasoning?

As they say in the accidentally-walked-in-on-the-other-person-cheating scene in the movies: "it's not what it looks like."

It's really just a regular slot port. I was using a piece of masonite to hold up the top part of the slot while I tried out different tunings. When I'm ready to finish the box, I'll glue the top of the slot in and take out the masonite brace.

turns out that this woofer is pretty obstinate in terms of box tuning. With this size box (11 Liters), basically much no matter what I did to the tuning, the F3 remained at 52 Hz. The only thing that changed was the hump just before it rolled off. The only way to get lower extension (which this driver will do) is to build a bigger box--which, unfortunately, sacrifices some power handling. But this is a nice-sized box, and it's a satisfactory amount of "honest bass" for jazz and rock music (except maybe metal, since these speakers won't do the ubiquitous "sonic booms"). The other nice part is that the bass is there, regardless of how quietly or loudly you listen.

[nepaeric]

Those wool cone Silver Flutes are some great offering by Madisound, and so are those Vifas.

My brother swears by the Flutes, and uses them in numerous autosound applications, to much customer satisfaction and great request. The price is right, as well.

Best Regards,
Eric

I thought about these two drivers I had procured for pretty much no money... they looked really well-built, but how well could they perform--could they perform together? Well, why not build a box and try it?
The Woofer is the Silver Flute W14RC25 wool cone, the tweeter is the Vifa Logic AC25SG05. Both are super-cheap (and available only) over at that Wisconsin speaker store.
The box is 11 Liters, vented out the front. I'm building these for a friend, who wants some actual bookshelf speakers, he even took pictures of the bookshelves he wants them on. He wants them painted white to "disappear" into the room. I'm fine with that.

So the $11 question: how do they sound? Well, after measuring everything up close against the wall, I played around with several different topologies. Seems these two drivers have certain slopes they want to "fall into"--but finding a combination the two could agree on was difficult.

I won't post the crossover yet, because what Curt said the other day is right. But I can tell you at this point that the woofers have a satisfying bottom end, and a surprisingly low F3 that I didn't really even fully take advantage of. The Vifa tweeter, depending on how it's crossed, can sound harsh. But with the woofer carrying a bit more of the weight, the Vifa started sounding cleaner and more natural.

I had a really hard time getting an in-phase crossover to work with these drivers. So I finally resigned myself to phase quadrature, and I'm really surprised with how effortless and smooth everything started sounding. I've said this before, but I think a lot of us are uncomfortable with phase quadrature, because we can't flip the polarity and see that nice, satisfying giant null at the XO point. Instead, you flip the polarity and things still look flat, just in a yucky sort of way.

OK, enough rambling. So far, this project has definitely rounded 2nd base. I'll post the XO in a few, when I get a more time with it. By the way, this is definitely the fastest I've ever put together a speaker; even my wife pointed that out. I explained to her that I've learned a few tricks (cutting pieces slightly large then using flush trim bit = smart move, $20 pneumatic nail gun from Harbor Freight = BRILLIANT). So I figured the name Orient Express was fitting, given the production time, and country of origin for the drivers.

[chrisn]

Looks Great I see I'm not the only one using GE caps

[Paul Carmody]

First off, my apologies to Andy. I'm ashamed to be using so many parts.

I was listening to the v1.1 (posted earlier) and while it "got the job done," I found it a bit too "dippy" to call it a closed case. For most rock or jazz or instrumental, it sounded good--nice and warm. But with female vocals, it was like sibilance city. That's when you'd start to notice the dips in response: all the consonants were e-nun-ci-a-ted, but the voice itself felt cold and bodiless.

I decided it was a chance to try out some new filter ideas--stuff I'd seen Roman do and Dennis Murphy do, but could never figure out why. So I experimented with parallel contour filters and found the slopes more flexible than I'd ever imagined.

So what to do with these new slopes? Well, why not try what Jeff suggested, and "relax" one or the other, in order to get better phase coherence and flatter summation? Well, that started working well, too... as you can see that the region between 1000 - 10,000 Hz smoothed out a lot. Unfortunately, I couldn't exactly tell you what the acoustic slopes ARE anymore... I mean, they're akin to a 2LR and a 3BW, but not exactly. And they're "spread apart" somewhat, to help with phase alignment and flattening FR. But they cross exactly at 3 KHz, and with reverse polarity show a deep null.

Anyone know why the phase is rotating slightly before the XO point? Does this matter? Does this signify anything?

I can't say for sure if this is the final version, but I could honestly endorse it at this point and recommend someone build it without regret. For reasonably-small bookshelf speakers, they have a BIG sound. Even my wife was surprised, and said, "You got $10 drivers to sound this good?" My only point of contention at this point is that I'm still hearing some sibilance--albeit pretty subtle at this point--what's frustrating is that I can't figure out where it's coming from. Am I overlooking something, or could it be the fact that it's a $10 tweeter?

Thanks to any and all who have input.

[curt_c]

My only point of contention at this point is that I'm still hearing some sibilance--albeit pretty subtle at this point--what's frustrating is that I can't figure out where it's coming from. Am I overlooking something, or could it be the fact that it's a $10 tweeter?

Thanks to any and all who have input.

You are really coming along with your crossover design knowledge Paul. Kudos!

I'd guess your sibilance is coming from that bump at 6K on the silver flute's response, and I'd suggest notching it out to see if that helps.

A trick I use to save on components is to simply parallel a small value cap across the woofer inductor. Try a .47 ufd to start and vary its value to center it on the 6K peak. Adding a resistor in parallel with the cap may be necessary to smooth out the response of the notch (lowers the circuit Q)

The added components will not affect the basic transfer function of your LP filter, but only alter the response around the notch frequency. -And should also smooth out the bump in summed response at that frequeny as well...

C

[Paul Carmody]

You are really coming along with your crossover design knowledge Paul. Kudos!
I'd guess your sibilance is coming from that bump at 6K on the silver flute's response, and I'd suggest notching it out to see if that helps.

A trick I use to save on components is to simply parallel a small value cap across the woofer inductor. Try a .47 ufd to start and vary its value to center it on the 6K peak. Adding a resistor in parallel with the cap may be necessary to smooth out the response of the notch (lowers the circuit Q)

Thanks, Curt. Coming from you, that means a lot. I've seen you mention these "elliptic"-type filters before, but didn't understand them. But now I'm starting to understand--so let me ask for clarification. Is the idea sort of like creating a parallel RLC filter, but without a resistor... thus creating a steep roll-off, which can be used as part of the basic filter topology, and not just an "added notch filter"? Then, if the roll-off is too steep, you can vary the resistor to alter the Q to taste?

If this is what you mean, I have been playing with this but as a series RLC filter (without a resistor) on the shunt path of some LP filters and gotten some great-looking results. I know PCD wasn't necessarily meant to do it, but you can find back doors. Unfortunately, I don't think it can model the kind of filter you're talking about (paralleling a cap to an inductor) in PCD... at least there's no backdoor I've found. Jeff, I love your program, and I don't think I could have managed my way into this hobby without it, but if you ever figured out a way to model this, that would rock!

But thanks again, Curt. Your suggestion was great. I DID model it in LSPCad 5, and it definitely works. I only wish I knew more about how to adjust components better in that program, so I could get a better understand of what's going on. I still feel a bit lost in LSPCad--although I use JustMLS all the time, because it is wonderful. At this point, I CAN model a working speaker in LSPCad 5, but when it comes to adjusting components, I get so confused flipping through windows that I forget what I'm trying to do. I don't know if I'm using the program correctly, or if there is an easier way to tweak small changes and see the effect to the overall system like you can so quickly in PCD. Yes Jeff, your software is THAT good.

[Jeff B.]

Thanks, Curt. Coming from you, that means a lot. I've seen you mention these "elliptic"-type filters before, but didn't understand them. But now I'm starting to understand--so let me ask for clarification. Is the idea sort of like creating a parallel RLC filter, but without a resistor... thus creating a steep roll-off, which can be used as part of the basic filter topology, and not just an "added notch filter"? Then, if the roll-off is too steep, you can vary the resistor to alter the Q to taste?

If this is what you mean, I have been playing with this but as a series RLC filter (without a resistor) on the shunt path of some LP filters and gotten some great-looking results. I know PCD wasn't necessarily meant to do it, but you can find back doors. Unfortunately, I don't think it can model the kind of filter you're talking about (paralleling a cap to an inductor) in PCD... at least there's no backdoor I've found. Jeff, I love your program, and I don't think I could have managed my way into this hobby without it, but if you ever figured out a way to model this, that would rock!

But thanks again, Curt. Your suggestion was great. I DID model it in LSPCad 5, and it definitely works. I only wish I knew more about how to adjust components better in that program, so I could get a better understand of what's going on. I still feel a bit lost in LSPCad--although I use JustMLS all the time, because it is wonderful. At this point, I CAN model a working speaker in LSPCad 5, but when it comes to adjusting components, I get so confused flipping through windows that I forget what I'm trying to do. I don't know if I'm using the program correctly, or if there is an easier way to tweak small changes and see the effect to the overall system like you can so quickly in PCD. Yes Jeff, your software is THAT good.

You mean like this crossover (which hasn't been made public until now, sorry PE, for making an alternate crossover)?




I model this all the time in PCD. You simply select a parallel RLC notch filter, either before or after the crossover, and set the R value to 9999. That's all it takes.

Jeff

[Wolf]

Thanks, Curt. Coming from you, that means a lot. I've seen you mention these "elliptic"-type filters before, but didn't understand them. But now I'm starting to understand--so let me ask for clarification. Is the idea sort of like creating a parallel RLC filter, but without a resistor... thus creating a steep roll-off, which can be used as part of the basic filter topology, and not just an "added notch filter"? Then, if the roll-off is too steep, you can vary the resistor to alter the Q to taste?

If this is what you mean, I have been playing with this but as a series RLC filter (without a resistor) on the shunt path of some LP filters and gotten some great-looking results. I know PCD wasn't necessarily meant to do it, but you can find back doors. Unfortunately, I don't think it can model the kind of filter you're talking about (paralleling a cap to an inductor) in PCD... at least there's no backdoor I've found. Jeff, I love your program, and I don't think I could have managed my way into this hobby without it, but if you ever figured out a way to model this, that would rock!

But thanks again, Curt. Your suggestion was great. I DID model it in LSPCad 5, and it definitely works. I only wish I knew more about how to adjust components better in that program, so I could get a better understand of what's going on. I still feel a bit lost in LSPCad--although I use JustMLS all the time, because it is wonderful. At this point, I CAN model a working speaker in LSPCad 5, but when it comes to adjusting components, I get so confused flipping through windows that I forget what I'm trying to do. I don't know if I'm using the program correctly, or if there is an easier way to tweak small changes and see the effect to the overall system like you can so quickly in PCD. Yes Jeff, your software is THAT good.

I usually just use 1000 ohms, but (as Jeff said) 9999 is more accurate I'm sure. You are thinking about this correctly. I also agree that targets are starting points before fine tuning. I usually hit them, and remove them from the simulation. You can change components to align the vertical (180 degree) points up better, and it isn't that hard to accomplish. Keep it up!
Later,
Wolf

[curt_c]

Thanks, Curt. Coming from you, that means a lot. I've seen you mention these "elliptic"-type filters before, but didn't understand them. But now I'm starting to understand--so let me ask for clarification. Is the idea sort of like creating a parallel RLC filter, but without a resistor... thus creating a steep roll-off, which can be used as part of the basic filter topology, and not just an "added notch filter"? Then, if the roll-off is too steep, you can vary the resistor to alter the Q to taste?

If this is what you mean, I have been playing with this but as a series RLC filter (without a resistor) on the shunt path of some LP filters and gotten some great-looking results. I know PCD wasn't necessarily meant to do it, but you can find back doors. Unfortunately, I don't think it can model the kind of filter you're talking about (paralleling a cap to an inductor) in PCD... at least there's no backdoor I've found. Jeff, I love your program, and I don't think I could have managed my way into this hobby without it, but if you ever figured out a way to model this, that would rock!

But thanks again, Curt. Your suggestion was great. I DID model it in LSPCad 5, and it definitely works. I only wish I knew more about how to adjust components better in that program, so I could get a better understand of what's going on. I still feel a bit lost in LSPCad--although I use JustMLS all the time, because it is wonderful. At this point, I CAN model a working speaker in LSPCad 5, but when it comes to adjusting components, I get so confused flipping through windows that I forget what I'm trying to do. I don't know if I'm using the program correctly, or if there is an easier way to tweak small changes and see the effect to the overall system like you can so quickly in PCD. Yes Jeff, your software is THAT good.

It is a notch filter, we just ‘borrow’ the inductor from the LP filter as one component of the notch, so it provides two functions. As the resistor value decreases it reduces the effect of the notch, but also broadens it bandwidth. (and lowering the filter Q) As Jeff has pointed out, you can do this easily in PCD. Select a series notch before the LP filter, and change the inductor value of the notch to your LP value. Change the LP inductor value to ‘0’.

I find the component value manipulation in LspCAD quite intuitive and user friendly, but I’ve been using LspCAD 5 so long that my opinions in this respect are most certainly biased. Anyway, try this:

-Open up the summed frequency response window, network window, and for easy reference, the network schematic window.

-Highlight the component you want to vary in the network window. Use the left and right arrow keys to change the value of the component up and down. The change to the summed response will be immediately displayed. (If a specific target response is desired, this same procedure can be used with any component selected in the optimizer window.) Alternately, double click on a value, and type in a new value in the pop up box.

For convience, the component values in the schematic window can also be changed. They won’t arrow up and down, but new values can be entered and the other windows will be updated.

The Snap function can be selected for larger incremental changes to standard values, but I usually leave it off, as often the values available to us will not be an industry standard value.

I’ll be glad to answer any other questions you might have about LspCAD.

C

[jkim]

Paul,

A minor point about your crossover in post #4. As you have already noticed, the tweeter's acoustic rolloff is close to 4th order and the woofer's rolloff close to 3rd order although you used a 3rd order electrical filter for the tweeter and a 2nd order electrical for the woofer. This is a result of the drivers' natural acoustic rolloffs combined with the crossover's electrical rolloffs. Note that when you estimate a filter's rolloff rate on a graph, you need to see the slope up to one to two octaves far from the xover point because that's the range defining the curve's rolloff rate. The SPL range of your graph is a bit limited to do this.

Technically, your crossover is an acoustic LR4 (not LR2) design because the total phase wrap is 360 degrees with two drivers connected in normal polarity --- that is, the two drivers' responses are 360 degrees out of phase around the xover point versus 180 degrees out of phase in an LR2 design that would require reversing the tweeter polarity. The woofer's rolloff is "relaxed" from the textbook fourth order slope to accomodate the phase lag due to its offset on the baffle plane. There is another, easy way to view this trick. The method is essentially to come up with something between textbook acoustic LR4 and BW3 filters. Why? When there is no driver offset, a textbook LR4 filter exibits symmetric lobing at the axis in the middle of tweeter and woofer. Also, when there is no driver offset, a textbook BW3 filter with drivers connected in phase exibits asymmetric lobing, tilted upward. What will happen if the woofer's acoustic center is located behind the tweeter's? Yes, the lobing of LR4 is now downward and that of BW3 is usually still a little upward but much less so than textbook BW3's. This is why an LR4 with either tweeter or woofer rolloff "relaxed" works for drivers on a flat baffle. You can also achieve this by relaxing both drivers' rolloffs a little bit (both are steeper than 3rd order but less steep than 4th order rolloffs). In this case, "asymmetric LR4" is a misnomer. But in most cases actual results are asymmetric rolloffs.

-jAy