Optimo on a diet?

[crapitecture]

I’ve been trying to work out a design for my new mains. My current system consists of 3 NHT SuperZero’s for the front channels, surrounds based on Zaph’s Hi-Vi B3S design.

I want to eventually replace all the channels with designs that use all of the same drivers, but I’m going to have to piecemeal it as I get the money and time. Although I’ve been very pleased with the NHT’s, I’m ready to shift them into my office, and try something a bit better and perhaps a tad more exotic. While searching for designs I might be interested, I was intrigued with the Optimo design at Humble HiFi. I really like the way it looks, and I was interested in pulling off a transmission line speaker. The drawback for me was the cost of the drivers he uses. I just wouldn’t be able to justify the expense to my wife. I’ve looked at utilizing the work done on the Modula MT and porting it over to the TL design, but I’m just not too enamored by the gyrations needed in the crossover to tame the RS180. So I’ve done some preliminary homework on a design that would instead utilize a TB W6-1721 and the Seas 27TDFC. They seem like reasonable quality drivers for the price. Also I haven’t seen any projects using the W6-1721 yet so I thought this might be an opportunity to contribute another design to the community. I eventually want to develop a center channel and surround design as well.

My problem is that although I have a few DIY speakers under my belt that have been successful, the TL and crossover arena is still like black magic to me. So I’m here to ask for some help in developing the design further. Also, there seems to be a serious lack of speaker design software for the Mac, so I’m stunted in that area in developing a crossover.

I’ve used MJK’s tables and Zaph’s measurements to arrive at the basic parameters of a TL design. The basic design is pictured below. Here are my questions:

Is the W6-1721 the right driver for a TL design? Are there better, less expensive speakers that I should be considering.

When you fold a TL line how do you measure the length accurately? I’ve not been counting the additional length (measured at the centerline) at the 180 degree turn, is that correct? What about at the opening at the end of the TL? Also, I am assuming that dimensions regarding the placement of speakers, openings, etc are measured at the center line.

Considering my sub, is a 30hz tune the right direction to go? I like the idea of creating a full range speaker that I can use elsewhere as I continue to trade up speakers, but perhaps a different tuning might provide a much better midrange character. Thoughts on this?

I’ve been reading on ML TL designs and there seems to be some advantages to using that kind of design as it reduces the length of the line. Would one of the MathCad MJK Sheets gurus be interested in running the numbers for an ML TL design.

Has anyone had experience developing a crossover with the 1721? Any help in creating a good crossover for this design would be much appreciated.

Thanks in advance for your help!

[crapitecture]

Ok, I've partially solved the software problem as I was able to get an older PC from my office's surplus. I have been trying to learn Speaker Workshop the last couple of days. I used the FRD tools to trace the Frequency response and impedance measurements as well as the T&S data. I've tried to create a crossover but I'm not getting it to work. I found a tutorial on creating an XO and have followed it, but I'm still getting abnormal outcomes. I'll keep hacking away at this for a while and any help getting me on the right track would be greatly appreciated.

Is Speaker Workshop the right tool for this?

Am I already too far off the mark that I should just scrap this idea?

[crapitecture]

More work on the XO's

I've stumbled through with the initial crossover. This is what I have so far. It is not modeled in the enclosure because that is up in the air at the moment, pending some feedback. There is a bit of a lull between 300-1500hz, is there a way to address that, or is this a minor thing. Any suggestions?

I guess I'm a little surprised that I haven't had any responses to my questions yet. Is there a noob hazing period I'm not aware of?

[ThomasW]

I guess I'm a little surprised that I haven't had any responses to my questions yet. Is there a noob hazing period I'm not aware of?

When questions go unanswered, it typically means it's not a project or subject that interests people, or providing an answer would be so time consuming it's too much of a hassle.

TL's are rather a pain to design and only a few people are interested in them because there's not much benefit over a ported box.

We're not big proponents of the designing loudspeakers using only mfgr's published frequency response plots. Reason being it completely ignores unit to unit variability between drivers, and the effects of the baffle the drivers will be mounted in.

[augerpro]

Not a lot of TL builders on this forum so for that particular aspect you may want to try another source.

Couple ways to do your crossover. Measure in box and model it or use FRD tools to sim it. I would decide first before going any further with your modeling, simply because you'll have to do all that work again when you get the drivers on the actual baffle.

SW kind of sucks IMO but you can use it if you want. You can use this program for either method I mentioned. Measuring with the driver in box will lead to the most accurate result. Using the FRD tools to trace and then sim can work, but you must have an understanding of the method for it to be accurate. Main things are making sure you model the baffle diffraction and sum that with response you traced from the manufacturers info, and accounting for the relative acoustic offsets of the drivers. Have you done this correctly?

[crapitecture]

I was using Zaph's measurements in SPL Trace as the basis for my info. I have downloaded most of the FRD tools so that I can sim it. I certainly understand that this is definitely not optimal, but I suspect that it is better than just using the mfg's info. I was thinking that I could do this step by step. Get the basic crossover, then address the baffle diffraction, etc. considering that I don't have the enclosure fully designed yet.

From seeing Zaph's and other's projects, it looked to me like you could design pretty much everything in software and get fairly accurate results that they correlated by measured results. Certainly, there is variation between units, but does every driver need to be measured? If they did, wouldn't it make kits a bit useless?

As you can probably tell, I'm dipping my toes into this hobby so it is difficult for me to justify the cost of proper software, jigs, microphones, etc. that would allow me to measure etc. I was just hoping to be able to make something better than my NHT's that I could continue to tweak as I learned more. Unfortunately, I just don't have a ton of money to throw at this right now, but I do want to learn. I hope that I don't have to get all the measurement tools before realistically going forward, that would suck.

ThomasW: I understand why people don't respond. I was being a little facetious I was just surprised that no one was interested, considering the discussions I have been reading for the past year or so that I've been lurking here. No big deal.

Thank you both.

[augerpro]

Keep in mind Zaph uses actual measurements of the drivers in the box to model with.

You can do it the SPL Trace way, you just have to take care to do it correctly. Building a driver file that is an accurate representation of reality is absolutely the first and most important step. Screw up here and all the rest will be wrong. So start at the beginning. Trace the response, then add in the baffle diffraction, adjust for AC. Once you've done this and have an accurate driver file, THEN move on to crossover modeling.

Post your progress...

[crapitecture]

Thanks Brandon, I will take a step back and proceed as you suggest.