Zaph ZD5

[Lindahl]

Been browsing for a while now, read through JonW's Modula thread, and read about the Zaph ZP5. I love the slim profile of Totem speakers, and the 5" Revelator looks like the only way to get that slim profile and not give up too much bass. Unfortunately it costs an arm and a leg. The only problem, I hate the look of Vifa tweeters - I know, I could be somebody's wife for sayin something like that, but I just don't like it. I was thinking of going with a 27TDFC replacement. It looks like, however, this won't work for a 2nd order crossover. The 27TDFC looks like it drops off too early? It does have lower distortion in that range, however. Either way, the 2nd order crossover in the ZD5 is a ladder delay network, which, considering it's trying to shift the drivers into phase, sounds very tweeter specific. So it looks like I'm kinda stuck, here.

To sum up, I'm looking for a design with the Scan-Speak 15W8530 and the Seas 27TDFC. Someone's probably gonna get mad at me for using a low-cost tweeter with a $200 woofer - but I can't seem to find a viable replacement for a small mid-woofer that has a decent low-end response but will let me keep a 7" or so baffle width. Any ideas? Trying to save cost here, since the woofer eats my budget for breakfast.

The description of the 27TDFC Modula MT sounds right in line with my taste on tonality - the warmth through the midrange, but with airy highs with a bit of sparkle (from the slopeless distortion of the 27TDFC?). The low-end response of the 15W8530 should keep that warmth, right?

Thanks,
Brian

[jdybnis]

The new scanspeak 6600 is a viable replacement for the XT25. It should be shipping again soon. No help on the budget though.

BTW: CJD's RS150 MTM have a slim profile and will give you decent bass. Especially if you make it a floorstander.

[Jonasz]

The Dayton RS28A may have decent enough extenson for a second order filter?

[Chris7]

You can mount the 27TDFC in the MCM waveguide as Zaph does with his Waveguide TMM project and get a clean second order LR crossover with just a single cap on the tweeter. The waveguide would fit on a 7 inch wide baffle.

[Lindahl]

You can mount the 27TDFC in the MCM waveguide as Zaph does with his Waveguide TMM project and get a clean second order LR crossover with just a single cap on the tweeter. The waveguide would fit on a 7 inch wide baffle.

Excellent idea, thanks. I'll start looking into it. So, for the crossover, the woofer has the same two notch filters, and it's inductor, and the tweeter has the single capacitor? Also, is there any way to contact Zaph? I'm not seeing the reason for the 950hz filter, though the 5khz one makes perfect sense. If he's correcting the slight dip, the dip looks to be centered above 1khz, not at 950hz. I also was wondering how shallow I can make the cabinet (while still keeping the same volume, of course). I'd like to try and make it around 10" deep.

[Jed]

You could probably do LR2 with the SEAS tweeters but it will require more parts and a complete redesign of the crossover. Also, why not just go 4th order? Zaph's crossover won't work if you change out the tweeter, so don't even bother asking him.

That said, you could work up a new design given your parts no problem using the FRD tools and speakerworkshop. Ideally, take your own measurements and go from there.

Jed

[Lindahl]

You could probably do LR2 with the SEAS tweeters but it will require more parts and a complete redesign of the crossover. Also, why not just go 4th order? Zaph's crossover won't work if you change out the tweeter, so don't even bother asking him.

Chris's suggestion of the waveguide won't work to get the phase right? The notch filters used in the ZP5 are just to smooth the woofer's response. The waveguide eliminates the complexity in the tweeter network, by being phase-correct, mechanically. You don't have to worry about the XT25's 500hz distortion. All you're left with is getting the slopes correct with the right capacitor and inductor, and some attenuation(?) - using reverse null modeling. Maybe I'm oversimplifying.

[Chris7]

Chris's suggestion of the waveguide won't work to get the phase right? The notch filters used in the ZP5 are just to smooth the woofer's response. The waveguide eliminates the complexity in the tweeter network, by being phase-correct, mechanically. You don't have to worry about the XT25's 500hz distortion. All you're left with is getting the slopes correct with the right capacitor and inductor, and some attenuation(?) - using reverse null modeling. Maybe I'm oversimplifying.

Well, basically, but good crossover design is not cookbook. You need to simulate the new drivers on a baffle to understand the baffle effects, simulate the new crossover you're developing, etc. This is not a minor effort if this is your first speaker project. And it's no longer the ZD5; for time reasons Zaph does not take questions about what are effectively completely new designs.

If this is your first speaker project and you want high performance, a very narrow baffle, and unusually deep bass for a 5.5 inch driver, I'd suggest building the ZD5 as Zaph designed it. You're not going to match his twenty years of experience with your first design without a serious investment of time and money (measurement gear, etc.) and it doesn't seem to me that there is any cost-benefit payoff here. You're using a lower performance tweeter and investing a lot of effort just to avoid how the original tweeter looks. From a seating position you'll hardly ever really notice it.

[Jed]

Maybe I'm oversimplifying.

Yup. There is a lot more involved here. Phase is one part of the puzzle. You'll have a completely different tweeter circuit to reach an LR2 target. If you know how to design/measure that, then you are all set.

[Lindahl]

Yup. There is a lot more involved here. Phase is one part of the puzzle. You'll have a completely different tweeter circuit to reach an LR2 target. If you know how to design/measure that, then you are all set.

Is that something LspCAD can do pretty easily? From the description, it looks like it can.

Well, basically, but good crossover design is not cookbook. You need to simulate the new drivers on a baffle to understand the baffle effects, simulate the new crossover you're developing, etc.

Isn't this just LspCAD? At least for a design were most of the complex crossover work is already done (woofer control, phase-alignment, and all-but-eliminated diffraction)? There isn't a lot of new design going on here... at least it doesn't look like it.

Taken from the description of LspCAD:
"A paradigm that makes the initial prototype modeling simpler is the parameter mode that allows the user to modify parameters such as crossover frequency and Q value rather than having to fiddle with component values."

It doesn't sound like it's difficult to design a low-pass/high-pass - you don't have to mess with components, just the slope and crossover frequency.

This is not a minor effort if this is your first speaker project.

Granted, but I'm not sure how building existing designs is going to get me closer to my goal. I wouldn't be learning anything.

You're not going to match his twenty years of experience with your first design without a serious investment of time and money (measurement gear, etc.) and it doesn't seem to me that there is any cost-benefit payoff here.

I don't expect to match his design, just come out with a decent one, based on his work. I have no problem investing time, but not money for measurement gear. I have an Radio Shack SPL meter, and RoomEQWizard, but I'm not sure how far that gets me - I could see myself upgrading to a decent $100 mic, though. LspCAD should get me pretty darn close, though, shouldn't it? At least for a crossover that's simplified by having the woofer already controlled (existing design), already being phase aligned (thanks to the waveguide), and baffle diffractions all but eliminated (again, waveguide, and the woofer size to baffle size ratio)?

You're using a lower performance tweeter and investing a lot of effort just to avoid how the original tweeter looks. From a seating position you'll hardly ever really notice it.

It's not just that, but it's also the air of the tweeter that I'd like, probably from the upward tilted response and it's distortion in the upper highs. Looks is still a big thing, though. I know... I know... :roll:

Sorry if it sounds like I'm being combative - I'm attempting my best to learn, even if though I'm being stubborn with my goals. I learn quickly and don't give up, so as long as I don't start getting ignored, I'm hoping I can pull this off with some help, here.
:T

[Jed]

You're on the right track but without accurate FR, impedance, and phase data of the tweeter measured (or simulated in your baffle with minimum phase) in your baffle, you can fiddle with LSPcad all you want but the results won't be anywhere near accurate. Not saying its impossible, you just need accurate files to start off with before designing the crossover. Either you find/simulate FRD and ZMA files or measure your own. My preference is obviously for measuring.

[Lindahl]

You're on the right track but without accurate FR, impedance, and phase data of the tweeter measured (or simulated in your baffle with minimum phase) in your baffle, you can fiddle with LSPcad all you want but the results won't be anywhere near accurate. Not saying its impossible, you just need accurate files to start off with before designing the crossover. Either you find/simulate FRD and ZMA files or measure your own. My preference is obviously for measuring.

I see. So you really can't get information to do accurate modeling unless you're actually measuring the driver (or someone's done that for you). The tweeter is pretty commonly used - though I'm guessing the addition of the waveguide changes everything. Zaph has used it directly in his TMM Waveguide design - is there any chance he'd make those sorts of files available to a wannabe?

[Chris7]

Is that something LspCAD can do pretty easily? From the description, it looks like it can.
...
Granted, but I'm not sure how building existing designs is going to get me closer to my goal. I wouldn't be learning anything.
...
I don't expect to match his design, just come out with a decent one, based on his work. I have no problem investing time, but not money for measurement gear.

LspCAD is fairly expensive on its own. These packages take a while to learn, and it sounds like you expect to use it more like a wizard than actually delving into the fundamentals of speaker design. You will still learn something by starting with a build of someone else's design. That's how I got into this hobby, anyway. If you want to start from scratch, I'd choose a tweeter that's easy to use (the TDFC is a good one) and a non-metal midwoofer that isn't too expensive and learn the basics with that rather than starting with $200 woofers. With the money you save you could even get a good measurement set up (they start at around $100).

Sorry if it sounds like I'm being combative - I'm attempting my best to learn, even if though I'm being stubborn with my goals. I learn quickly and don't give up, so as long as I don't start getting ignored, I'm hoping I can pull this off with some help, here.
:T

I would probably suggest starting a new thread with a different title. You'll attract more readers that way. Say you want help designing a speaker from scratch with the TDFC and a narrow midwoofer with good bass.

[Jed]

I see. So you really can't get information to do accurate modeling unless you're actually measuring the driver (or someone's done that for you).

I imagine Roman can better explain what you are asking in the first 2 links at this address: http://www.rjbaudio.com/Audiofiles/audiofiles.html

You can process "accurate" data with different methods. Simulation using FRD tools, or measuring- with measuring as the best choice.

[jkrutke]

Also, is there any way to contact Zaph?

Zaph has used it directly in his TMM Waveguide design - is there any chance he'd make those sorts of files available to a wannabe?

I hate to let you down, but nope and nope. My email FAQ will be back up later explaining why.

But, I will take a moment here to shoot a few holes in your ideas. :B Will a waveguide 27TDFC work with a 15W8530? Sure, it could be nice, but there's not a single part of the ZD5 design that you would be able to copy. The waveguide + single cap LR topology only works for a 2kHz crossover, something the ZD5 is not. Raising the cap value higher to match the ZD5 would cause the resulting tweeter rolloff to stray from LR2. And of course, lowing the crossover on the 15W8530 to match a 2kHz LR2 is a complete redesign. Then there's the issue that the tweeter's acoustic center in the waveguide is likely too far back for an LR2 with a 5" woofer, unless you don't mind the listening axis pointed upwards or if you design in asymmetrical slopes. Additionally there's the issue that a waveguide on a baffle next to the woofer will change the woofer's response due to cavity effect.

The bottom line is that you need to get your own measurements and go from there with nothing less than a complete design from scratch.

Going off topic somewhat, I was thinking of pairing a waveguided 6600 AirCirc with the 7" Revelator. Possibly on a dipole baffle. Those are a couple drivers that have been in my inventory for a long time and I really want to use them.

I'll leave it up to everyone else to continue this discussion. Here's where I make a quick exit.

Regards,
John

[Paul Ebert]

Going off topic somewhat, I was thinking of pairing a waveguided 6600 AirCirc with the 7" Revelator. Possibly on a dipole baffle.

Yet another enticing dipole design that I cannot afford

[Lindahl]

...

Glad I could have the man himself tell me to toss it in the can. Thanks for the explanation as well. It helped me visualize the reasons why it wouldn't work.

LspCAD is fairly expensive on its own. These packages take a while to learn, and it sounds like you expect to use it more like a wizard than actually delving into the fundamentals of speaker design.
...
If you want to start from scratch, I'd choose a tweeter that's easy to use (the TDFC is a good one) and a non-metal midwoofer that isn't too expensive and learn the basics with that rather than starting with $200 woofers. With the money you save you could even get a good measurement set up (they start at around $100).

I would probably suggest starting a new thread with a different title. You'll attract more readers that way. Say you want help designing a speaker from scratch with the TDFC and a narrow midwoofer with good bass.

This all sounds like good advice. I've taken the time to familiarize myself with a few possible drivers and plan on posting soon. I assume your suggestion about avoiding metal woofers is to avoid complexity in crossover due to the huge break-up problems they have? Since I'm starting from scratch, I'm taking some time away from here to do some serious reading on crossover design, so that I'll have plenty questions to ask when I return. Wish me luck!
:B

[cjd]

LSPcad is a TOOL

It does not do the work for you.

It can model, and you can tweak and get things that LOOK good from one particular perspective, but it's all too easy to NOT make a note of something. For example, very early on in my run of this hobby, I had a nice flat response modeled - looked great. Except that it only worked due to negative summing, and it would ONLY have worked from just exactly perfectly one position - move a millimeter and you'd have a nasty bump in response right where many people prefer a small dip.

It's like saying a table saw can make your speaker boxes for you. Almost.

C

[Jonasz]

When the XT25 and SS9800 are more or less interchangeable, would the 9800 be an upgrade in this design or does it have to poor extension down low?

[SoundEngine355]

The new scanspeak 6600 is a viable replacement for the XT25. It should be shipping again soon. No help on the budget though.

BTW: CJD's RS150 MTM have a slim profile and will give you decent bass. Especially if you make it a floorstander.

If you using the 66000 what changes to the crossover are needed?