Current Project: Zaph ZRT 2.5 - A Change of Methodology

[bmwman91]

This will be a build thread, complete with pictures and a bunch of pedantic commentary that people are free to skip over. The ZRT components are en-route from Madisound & I should be receiving them on 11/2. I will be going to a lumber yard later this afternoon to get enclosure materials.

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

My previous speaker building projects have been entirely of my own design & construction. For me, that is much of the enjoyment: balancing all the various aspects of a design to try to get the optimal compromise of performance in various areas. Admittedly, I am probably mostly in the "newbie" class of speaker builder, at least as far as it seems to be defined on the Zaph audio site. Maybe some of the "newbie" classification comes from being an engineer, and assuming that I can always improve on a design that didn't originate as my own, which is why I designed my own in the past (arrogance, the most common folly of just about anyone in any field).

Since I get so much fun from the optimization & analysis aspect of speaker design, it is sort of a challenge then to embark on this ZRT project. The fact is that John K was probably designing loudspeakers since before I was born (not calling him an old-guy or anything), and at this point I have only a few designs under my belt. So, no matter how much I want to tweak this or that, I am not going to so that I can see how well an "off the shelf" design that originated with a seasoned designer can do compared to my efforts. Hopefully, my builds' performance will seem totally feeble once I listen to the completed stock ZRTs. Being that the ZRTs will use far more expensive and lower distortion drivers than stuff I have used in the past, I think that there's some automatic performance improvement right there.

I have done my best to analyze all of the aspects of this design that I can in a computer simulation environment. Of course there is a bunch of stuff I am likely missing, but I am not an audio professional and can't justify the cost of some analysis packages at the moment. There is a surprisingly good amount of free software out there, and the dedication of the DIY audio community has never ceased to amaze me. Anyway, these have been my findings.

- Baffle design: Seems to be optimized (duh). I ran the dimensions through the impressive baffle diffraction simulator sheet provided by the FRDC, and really, there isn't anything I can do to improve it using a rectangular box. MAYBE I could use some sort of TransLam box, but that is a lot of work, and it seems rude to stray from what John K intended in his design. I guess it is hard to accept that such a simple box design can be optimal...gotta tell the inner engineer to shut up & accept it...a more experienced engineer already did the hard work. Besides, I can always build other boxes & swap in these drivers to test for fun.

- Crossover design: At the moment, I have a DCX2496 + 3xA500 amp setup for my 3-way tri-amped system. From a technical perspective, active filtering has all sorts of advantages over passive. In practice, the difference seems more minimal. Since I already invested in the gear, I really wanted to try to use this for the ZRT 2.5s & set the filter orders / cutoff frequencies to match those of John's passive design. Really, I did. But, as he says so plainly in his FAQ, "newbies" messing with the filter designs are only going to pervert the design & mess it up. Although the passive crossovers are not inexpensive, I feel that I must at least TRY them. Eventually I will probably screw around with the DCX2496, just because I can. Still, I am going to force myself to stay true to his design since he put the work into making it sound excellent. If anyone here HAS tried a tri-amp setup with these, I'd be interested in hearing your experiences. By no means do I expect to really improve performance with an active filter setup, except maybe by obtaining better cone control by reducing output impedance seen by the woofer. Mostly, I figure I have all the gear and I would hate to have it sit unused (seems like a waste). I assume that some of why he always uses passive crossovers is that he is a big "value" guy, and designing things that require people to buy multiple amplifiers & fancy active crossover units seems pretty contrary to his mission to bring great "value" to people. That, and a well designed passive unit still performs damn well.

- Enclosure volume: It is tempting to increase the volume above 65L to get lower bass extension. However, at 65L the group delay of the ported box, tuned to ~30Hz, is more or less the flattest you can get it. According to WinISD, tuning to 32Hz seems to get GD a wee bit flatter, but probably not enough to notice. Using 0.75" MDF and the bracing he shows in his drawing, accounting for the ports, cutouts & drivers, the actual internal volume comes out to be 62L. The only way to get it to 65L without touching the baffle is to make it a little deeper. I am not doing that because a) 62L...close enough, just adjust the port length to keep the right tuning for it, and b) I can build the boxes with absolute minimal waste from 4 24"x48" MDF boards. If I wanted these to be just a little deeper, I'd end up having to buy 6 of these boards and wasting almost one whole board. Waste is sort of abhorrent to me, and I can adhere exactly to John's design spec without waste. I will play with the ports a tiny bit since he seems a little less strict on that, as long as the baffle dimensions remain the same.

Anyway, I am sure that a lot of people here are going to read this & think, "well duh, of course you can't find anything wrong with the design, newbie!" Yup. Part of this adventure has been personal in that I am forcing myself to trust in someone else's design and accept that someone else did a better design job that I could. SAYING that you know that there are smarter people out there is really easy to do. Actually accepting that and truly realizing that are is another thing (particularly for technical individuals..."but, but, but given enough time and analysis, surely I can make this BETTER!" is a tempting assumption). And heck, I want to see just how good I can get some DIY speakers to sound, regardless of who designed them.

Phew, that is a lot of text! If I clicked into this thread, I probably wouldn't read it, either! Hopefully nobody took the above text as me trying to "sound modest", or somehow demonstrate technical prowess. I have my limitations, and this project has helped me realize how far they seem to extend. Technical hobbies are a fun way to challenge yourself, and some of the most interesting challenges come when you realize you don't really know what you are doing. Besides, coming to that realization helps to pave the way for someday "knowing what you are doing." Sometimes it is just fun to put down the details of things that are floating around your head & see what other perspectives you get from people.

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

The enclosures will be constructed of 0.75" MDF. The aesthetics will be something along the lines of the enclosure shown on the kit parts page at Madisound. This design by Lee Taylor & Co. is absolutely gorgeous!


I will NOT be copying the design, as gorgeous as it is, since would feel akin to stealing (and probably is, legally). The paint/trim will be along the same lines though, with satin black sides and stained wood for the top/front. At this point I am still debating between veneer, or maybe just using solid hardwood for those panels. Workability seems like it can be an issue with a lot of exotic hardwoods (sort of thinking about Cocabola, Zabrawood, Tigerwood or Padauk). Trimming veneer has been problematic for me in the past...although it is probably worlds cheaper to redo messed-up veneer than it is to replace a piece of solid exotic hardwood!

Construction will begin this evening. Future posts will have more pictures and less uninteresting commentary!

Here are the previous projects that I have posted about on here. These are MOSTLY pictures, which is more interesting than a wall of text.
First 2-Way's
Excessive Subwoofer

[bmwman91]

And now for something completely different. Pictures (sorry about the quality, this was from my phone)!

All of the basic panel cuts are done now. The little pile to the right is the waste. It's the least waste I've ever had after cutting panels...I wonder if this was deliberate in the design of the 65L enclosure.


This is "my" workshop. My parents live about 25 minutes away from me, and I routinely go over there to make a mess in their garage since I don't yet have my own. Someday...stupid Bay Area housing prices!


I also picked up a rabbet bit kit with bearings in all the sizes I need for this project (and more).


One other silly project I have been playing around with is a DIY aluminum smelter. My dad is interested in building functional internal combustion engine models, and I have no idea what I might do with this, but it seems like a good idea anyway.


It is made from an old helium tank. We are still working out the right mixture to use for the refractory cement that will line the smelter & lid. We have a size-B4 graphite crucible that it is sort of being designed around. Anyway, I probably won't dwell on this project too much since it isn't really speaker-related. It is fun & kind of nerdy so I figured I'd share anyway.

[reactivepower]

As another ZRT builder, I can't wait to see your results. I decided on a sealed box and am interested to hear (no pun intended) how your ported version turns out.

As far as the crossovers go, I would say to try the DCX2496. I wouldn't think there is any reason not too. You can't really go wrong with the active crossovers except watch out for DC on your tweeters. You may want to use a cap in series with 6600 to protect your investment. I would also recommend trying to find a way to get access to the series inductor meant to be used with the lower woofer. You should try to use this even with the active version.

Good luck. Any updates?

[tuxedocivic]

Huh, just read all your commentary and found it very interesting. I'm a lot like you.

[bmwman91]

Originally Posted by reactivepower

As another ZRT builder, I can't wait to see your results. I decided on a sealed box and am interested to hear (no pun intended) how your ported version turns out.

As far as the crossovers go, I would say to try the DCX2496. I wouldn't think there is any reason not too. You can't really go wrong with the active crossovers except watch out for DC on your tweeters. You may want to use a cap in series with 6600 to protect your investment. I would also recommend trying to find a way to get access to the series inductor meant to be used with the lower woofer. You should try to use this even with the active version.

Good luck. Any updates?

Yeah, my thoughts were that running active shouldn't hurt anything, but I wanted to give the stock design some justice first. At ~$300 for the XOs from Madisound, I might just hang onto them lol.

What is this about a series inductor for the lower woofers?

I have made a lot of progress on them, but have been lazy & have not put pics up. I am away on business travel, but I think I have some pics on my phone I can put up.

Originally Posted by tuxedocivic

Huh, just read all your commentary and found it very interesting. I'm a lot like you.

Thanks. I was afraid that I got way too pedantic about "personal experiences" and whatnot...I was still on Hong Kong (+16hrs) time when I wrote all that out. The rest will be pictures & less commentary, I swear!

[bmwman91]

OK, so here are some pictures. A bunch are on my home PC and I will try to remember to post them when I get back.

All of the joints used non-locking rabbet grooves. All of my previous enclosures used interlocking rabbet joints, but I have found that it gets cumbersome to adjust width, depth & offset. I seem to get tighter tolerances with just width & depth as variables. Initially, I was being dumb and routing everything with no outer guides, so I inevitably had to re-do them all with some equal-thickness guides and 1/32" deeper to knock out my mistakes. To do that, I used the lathe to turn out some Delrin bushings to fit over the stock rabbet bushing/bearings to get the right width.



Skipping way ahead...

I have lined the inside of the panels with 5/8" open-celled foam. In the past I have used a spray adhesive, but I didn't feel like killing brain cells this time, so I just stapled it in place. No panels are glued together yet. The braces WERE glued & screwed.


I am still figuring out where the crossovers will go. Since the baffle will not be removable, they need to go somewhere accessible. I am thinking of sticking them on the sides up around the drivers. Will that be problematic? I don't know, but it seems unlikely. If I decide to electronically filter everything, I can always drill new homes for binding posts & remove the passive XOs.


Speaking of binding posts, the ones I received didn't quite fit into the plastic spacers I had. Well, that is not a problem when you have a lathe in the garage! Damn I love brass...I think it is even more nice to work with then Delrin.


Here they are all fitted nicely.


This is where I left-off before heading out for my trip. I am actually in Redmond, WA for an acoustic measurement seminar being given by a consultant my employer brought in for a couple of days. The audio fun never ends!


This will probably be the last project that uses MDF. I hate the stuff...what a mess! I rigged the shop-vac to attach to the router, and it helped, but it is a real nuisance to work with the stuff. Maybe it is nice and dense, and cheap, but I don't care. Baltic Birch is the way I am going in the future.

I have decided to use some 0.025" Zebrawood veneer for the front & top. The chamfer bit mashed up the woofer bores, so I can't just use their profile to use a flush-trim bit, and the counter bore is too shallow to get a bearing into. So, I will have to make a jig and veneer, trim and mask the baffles prior to gluing 7 screwing the walls. The tops will just have to be done after assembly because of how the panels interlock.

What are people's recommendations for finishing the Zebrawood veneer? The natural colors are nice, although a TINY hint of red might be cool, but I think I will avoid taking chances with stains on expensive veneers that don't usually need it. Some sort of satin finish to seal & protect them is also preferable. Wax maybe?

I will have a LOT of screw countersinks to fill & sand too. Once I have done that, I will apply a R0.125" round to all edges (except veneered ones), and a R0.5" one to the front-side edges to attempt to soften baffle diffraction ripple. After some sanding sealer, I will be painting the non-veneered sides some sort of satin black. I have an HVLP gun...does anyone have recommendations for paints that work well in those? I used black tool enamel thinned with mineral spirits on my subwoofer, and that gave it a nice slightly textured finish isn't matte, but not satin either. I really like that, but it was a huge pain to pass through the gun!

[sreten]

Transfer functions


Hi,

FWIW for once it does seem you can very near the designed
passive c/o using some fairly simple active c/o functions.

rgds, sreten.

[reactivepower]

I was in the same boat with you for the crossovers. I ended up putting them on the back wall behind the drivers. They actually fit through the driver cut outs nicely. Mind you, your enclosure is deeper than mine since it will be ported.

You may have different experiences, but I have always found that a few braces should be on each wall for the deadest enclosure. Have you thought about adding any going front to back?

[bmwman91]

Thanks sreten. I tossed the XOs into LTSpice to get an idea of the cut-off frequencies. Since the drivers themselves are a part of the filtering in the passive design, I figured I would see how that worked out. Maybe I will play around a bit with my measurement mic. Since the DCX2496 is so darn easy to use, I can probably take some measurements with active and passive setups. Since the DCX lets me also control delay, I might be able to get the main vertical response lobe close to where it is with the passive XO.


reactivepower, I am sort of thinking of mounting them on the back. We'll see what I think up. I thought about using the back wall, but the ZRT page says that the back wall can be a reflection issue (hence 2x the foam) so I wanted to keep the XOs out of that path. As far as bracing goes, I just decided to keep it simple. I went way over-board with bracing in my subwoofer, and I don't know if that is of any real value. Since the largest panels are braced in this design, and the front/rear are about 25% as wide as those (almost the same width as the brace spacing) I figured it would not be too big a deal. Plane resonance in the sides seems like the biggest realistic issue to me (although eery little bit helps). As I sort of mentioned in the (overly wordy) original post, this is an exercise for me to let go of some of the obsessive-compulsive tendencies I have when it comes to fabrication.

[schmalex]

I built an amplifier enclosure with zebrawood. To finish, I rubbed it with teak oil. The more coats you apply, the shinier it gets. It's really easy to work with. Zebrawood smells really funky, but that seems to go away with some finish. If you put a tint in the finish, I'd be worried that it would take away the contrast between light and dark, which is the most redeeming characteristic, to me.