Introducing the Dayton HiVi RTS181.3 Two-Way Bookshelf Speaker

A few months ago I decided to redo a pair of bookshelf speakers I had built almost 15 years ago. Click here if you want to check them out. This speaker makeover consists of swapping the HiVi Research W6 driver for a sleeker Dayton Audio RS180-8 7″ woofer and replacing the vanilla Vifa dome tweeter with a HiVi research RT1.3WE planar ribbon tweeter. And the results turned out awesome. The speakers sound much better, have a smoother, flatter frequency response, have more bass, more treble (thanks to a much more complex crossover) and I think they just look a lot better, more modern, more high-end then they did before. This blog post is meant to serve as the condensed version of a forthcoming (and much longer) post that goes into a lot more detail surrounding the design aspects of this speaker with a special focus on the crossover. For this post, I’m just going to show off some pictures and kind of walk through the process of taking this old pair of speakers and making them shiny and new again. I’ll also share some of my thoughts about this speaker project overall for those looking to embark on something similar.

With a retrofit build like this there were two basic criteria I had to stick to when picking out a new woofer: it had to model/sim well in the existing enclosure volume of 0.57 cubic feet and it had to physically fit in the cutout of the old woofer. This actually narrowed my search down quite a bit. Parts Express has a great selection of 6″-7″ woofers that would have fit the bill nicely, but when I ran the numbers for the Dayton Audio RS180-8 aluminium cone woofer, I knew I had a winner on my hands. I’ve used several Reference Series drivers from Dayton Audio and they never disappoint. It modeled great in the volume I had, and even with the original tuning frequency of 42 Hz, I wouldn’t even need to change the port length, and being that it is a 7″ woofer, it actually fit the opening almost perfectly compared to the old Hi-Vi W6 woofer. And with the woofer all picked out, I moved onto the tweeter selection.

This proved to be a much more difficult task. The original tweeter cutout matched a fairly standard 104mm diameter faceplate, common with a lot of tweeters, especially Vifa tweeters. It would have been easy to pick up a XT25TG30 or a DX25BG60 and they would have dropped right in, but I really wanted to do something different, something more challenging and more of a departure from the original tweeter, which was an older Vifa DX25TG soft dome. I wasn’t exactly sure how I was going to flush mount a new driver with a different diameter faceplate onto a speaker that already had a hole cutout in the middle of it though. So that was my original hesitation to selecting a completely new tweeter. But once I figured that out though (which I’ll go into below) I ventured beyond identical tweeter sizes. That’s when the HiVi Research RT1.3WE planar ribbon tweeter caught my eye. This tweeter looked sweet! A planar ribbon type tweeter with a nice and flat frequency response, a flat impedance curve and a huge 4.7″ aluminum faceplate. I quickly sketched up some drawings of these two drivers mounted into my pre-existing enclosure and they looked absolutely fantastic. Driver spacing was perfect on the front baffle and the dark gray surround color of the Dayton woofer matched the dark gray aluminum faceplace of the HiVi tweeter nicely and I just loved that little hint of copper peeking through its metal grill. Also the black metal grill actually matches the black aluminum cone really well. And that grill would help keep kids’ poking fingers at bay. I had found my tweeter. I added a pair of woofers and tweeters to my shopping cart and week later these beauties showed up at my doorstep. Let the speaker project begin!

With drivers in hand and the speaker enclosures already done, I had to figure out just how to get these new speakers to work with the existing holes of the original speaker boxes. For starters, I really wanted to flush-mount both the woofer and tweeter. Which is really what made this whole process just that much harder. With holes already cut in the enclosure, there’s no way to get a router in there to make a perfect circle to create a recess for the driver basket because there’s just a big open hole there. Normally you would use a circle jig with and a center guide pin in the middle of the location where you want the driver and you router the recess portion first, and then go back and cutout the driver hole. Or in some cases you can add another thinner board on top of the baffle that is cut out separately from the woofer hole. Both options seemed like non-starters. The speaker hole was supposed to come second, not first.

That’s when I came up with the idea to glue a strip of MDF inside the cabinet to the backside of the front baffle that had small blocks of MDF glued to it that were equal to the thickness of the baffle. This basically put a piece of wood just big enough for the guide pin of a circle jig to fit right in the middle of each driver cutout and at the same height as the baffle. As if the holes weren’t even there. And with that, I set the circle jig to the exact size for the woofer and tweeter, set the depth to match the thickness of the drivers, and cut out with perfect precision the recesses required to flush mount each driver. After that was complete, I ripped out that temporary MDF strip and small wood blocks and was left with a perfect recess to fit the new drivers. Well, I still had to take a jigsaw to the tweeter cutout because it was literally fitting a square tweeter in a round hole. The HiVi tweeter housing has a rectangular shape so I just had to cut little corners out of the existing round hole from the previous tweeter and I used a paper template as a guide that that matched the tweeter dimensions for doing so. Piece of cake. Let’s move on!

Alright so this is where we get to the abridged version of this blog post, the crossover design. This seriously deserves a complete post of its own and I’ve actually already started writing it, and it’s long, and wordy, and goes on and on about concepts of baffle step compensation and notch filters and discusses in detail response plots from REW of literally hundreds of different crossover combinations that I tried before settling on the final design you see here. So I won’t rehash those details again. Suffice it to say, I ended up with a ~2,800 Hz crossover point and 24 dB/octave slopes, a full 6 dB of baffle step compensation (which after two weeks of really listening in my actual room, I revamped it to be only 3.5 dB of BSC and to me they sound way better, see updated schematic to the left) and notch filters on both the woofer and tweeter to help clean up some driver resonances. In all, the crossover has 15 total elements, which is bigger and badder than any crossover I have designed to date. They were so massive in fact, that I didn’t even end up putting them inside the cabinet, I built separate little wooden enclosures to house all of the capacitors, inductors and resistors. This certainly is not everybody’s cup of tea, but I had more room to store the crossovers outside than I felt like I had space inside the enclosures. So I just went for it and made the entire crossover its own little box, painted it black, stained it dark brown and said to myself, well isn’t that fantastic!

My wife’s first comment when she saw them was, “Where do you plan on putting those?”. I resisted the urge to say, “They mount above the speakers on the wall, like picture frames!” Showcasing my beautiful crossovers next to pictures of our beautiful children. Instead I replied, “They’ll go inside or underneath the media console, don’t worry, you’ll never see them.” She made a grumble and walked off only half satisfied with my response. To be fair to her, at this point in the whole speaker process, it had been months since I started. I had taken over half the house with all my speaker measurement gear, laptops, amps, mixing consoles, wires galore, not to mention the speakers themselves half-finished, everything just sitting in whichever room I had commandeered for the week to do all of my speaker testing. So when I took over the kitchen table to solder the crossovers into these little boxes, I could understand her curiosity. I think she was happy though that the crossovers would not become a permanent part of our family room decor. As you can see in the finished pictures, the crossovers fit nicely into the bottom cabinet of our media console and sit up against the sides, with the door closed, they are completely out of the way, out of sight, out of mind.

Jeesh Dan, that’s the abridged version? Yup, just dumb stories and no technical substance, for now. I’ll let most of the pictures do the talking. All anybody really wants to see is pictures anyway. The crossovers feature all air-core, low-DCR inductors, metalized polypropylene capacitors and non-wire wound resistors. I used Passive Crossover Designer 7.0 to simulate and design the crossovers and used REW (Room EQ Wizard) to take all the measurements, 533 measurements in total, once all was said and done. I finally realized the limit to the number of graphs you can display at once in REW – and it’s 99. And you know what? Sometimes, it just wasn’t enough. Moving on.

Once I had settled on the crossover, I pulled the drivers out and actually finished the speakers. Also, I went ahead and ordered up all the crossover components from Parts Express, and while I waited for those to arrive, I spent the next couple of weeks sanding, staining and painting the enclosures. So just to go into how I did this part really quickly, without going into too much detail. I sanded down the entire speaker, all six sides including the stained oak panels with 150 and 220 grit sandpaper and an orbital sander. I wasn’t sure I would be able to sand off the stain completely and though I did get most of it, in the end I was not able to completely sand away some of the deeper penetrating spots in the grain. My original plan was to stain the speakers in a dark gray stain, similar to my last pair of speakers, but after putting one coat on over what was still left of the old dark walnut stain, it didn’t look right. I thought for sure I would end up with some odd hybrid stain color that was going to look wrong. So I buckled and re-stained the oak sides with the same dark walnut stain as what I used 15 years before. They turned out different though, the grain has a lot more detail now, more pop, more texture, so even though they are basically the same color, they have a lot more personality now. So I’ll take it. Plus I finished them in a semi-gloss polycrylic instead of a flat clear coat like I did last time, so they have a lot more depth just because of shinier sheen. After about 4 coats of stain, the wood was looking nice and rich so it was time to mask off the sides and paint the rest of the speaker. This is where I also debated painting the speakers white. I’ve always wanted a pair of high-gloss white speakers. I don’t know why, I will do it someday, perhaps on the new studio monitors I’ve got in the works for my recording studio, something akin to the Yamaha HS5 perhaps? But for this speaker setup, I just went ahead with the traditional black. I knew it would look good with the dark stained sides, the all-black drivers and more importantly, wouldn’t stand out too much in my family room as the mains for our home theater setup. Now that they’re done, I kind of wish I had done them in a different color, because they look so close the original speakers, and yet I spent soooo many hours sanding and staining and painting, that it would have been nice for them to actually look different, but oh well, they are what they are and if I really don’t like them one day in the future, I can always pull the drivers out and redo them all over again. I’m going to have to dip them into a bath of paint and stain remover though first.

Once the crossover parts arrived I spent some time figuring out where to place each component and building the actual boxes that would house all the parts. I drew everything up 1:1 scale using DeltaCAD on the computer and determined that I could pretty much fit every part on a 9.5 x 11.5 inch “circuit board”. And by circuit board I mean a 1/4″ piece of MDF from Home Depot. I drew up the placement of each inductor, capacitor and resistor on the computer and then shifted and moved things around for a few days until I got something that I thought looked good and complimented the schematic well enough as to prevent unnecessary crossovers of wires while keeping the inductors at a reasonable distance so as to not inductively couple to one another. My main goal for good inductor placement was to keep the closest inductors at 90 degrees from each other within the same frequency range (so the tweeter’s inductors are closer together but rotated at 90 degrees) and then the woofer’s inductors are not at 90 degrees from the tweeter inductors but are far enough away as to not cause a measurable effect on the driver’s response. I did do a bunch of tests in this regard by measuring the coupling effects of inductors at various distances with and without rotating them 90 degrees and true to every design guide you may have ever read, two inductors whose fields are at 90 degrees from one another can literally be placed almost touching and it will not have a measurable effect on the driver frequency response. However the worse effect I saw was not keeping the tweeter’s inductor away from the woofer’s inductor which causes the woofer’s low-pass filter response to not perform as well, thus letting back in some of those terrible cone break-up modes we are trying so desperately to knock down. That having been said, so long as the inductors were about 5 inches apart, even without rotating them 90 degrees, their effect was negligible on the frequency response of the speaker. So that’s what I attempted to do, which is in part how the crossovers ended up being so big. And you should know that if you’ve ever tried to locate four inductors all in close proximity to each other, you physically can’t mount every one of them at 90 degrees to every other one. At least two of them will always violate the rule and want to couple. At least I couldn’t find a way to do it.

Alright, so enough crossover layout talk. I printed the final parts placement as a 1:1 print and taped it to the 1/4″ piece of MDF and drilled about 100 1/4″ holes which would become the thru-holes for the leads from the caps/inductors/resistors as well as for a pair of zip ties to secure each part. That’s the main reason for doing the entire layout on the computer. There’s basically no way to know where to drill all these holes and attempting to do the layout on-the-fly with real parts and then drilling holes would have just been a mess. And as someone who’s done some real PCB layout work in the past, the best way to do this is to spend the time on the computer doing the complete layout, placing the parts, routing the wires, figuring out where everything is going to go before you start actually cutting/drilling the board. And even though I had measured and drawn up every part to scale on the computer, I still messed up from which side the inductor leads came out – they were actually 180 degrees out from how I drew them, so when I went to place them, the holes were a 1/2″ off from where they were supposed to be. I had to go back and drill a couple of new holes to rectify that. But hey, no one’s perfect. And if that had been the only thing I messed up on these speakers, I would have been in good shape (I won’t even go into the actual mess that was routering those speaker baffles, oh man, so many things went wrong there, but I won’t go into it).

So I guess that brings me back to finishing the crossovers, I painted the “circuit board” in a high-gloss black paint, I mounted all the parts using zip ties to hold them in place and then I soldered all the component leads with some 60/40 tin-lead solder. I used my trusty old 30-watt Radio Shack soldering iron that I have had for the better part of 20 years and even though the company is no more (rest in peace old friend) that $7 soldering iron lives on. One trick I discovered with this crossover build that I thought worked really well was I used 14/2 Romex copper wire to connect all of the part leads anywhere there wasn’t enough length to just go point-to-point. And if you don’t know, Romex is the de-facto brand of high-voltage electrical wire used in most homes, is made up of beefy solid copper wires and can be bought for relatively cheep by the foot at The Home Depot. This stuff is awesome, it holds its form, it eats up heat and solder just loves to flow around this stuff nicely making strong, low-loss connections. I got one good picture of the back of the crossover that shows this off. I recommend it over typical stranded-type speaker wire for making crossovers. I built a small box out of some 1×2 furring strips (97 cents at Home Depot) and then stained the box the same color as the speakers, threw on a clear coat and then installed the crossover board using some 10-32 hurricane nuts and 3/4 screws I bought from Ace Hardware. And last but not least, I installed six pairs of banana jacks (or binding posts) from Amazon to provide an easy connection point from the amp to the crossover and the crossover to their respective woofer and tweeter in the speakers. And with that I think we are done!

I set up the speakers in their final resting spot on either side of our HDTV and called this project done. I ran MCACC on my Pioneer receiver and let it do its thing and then sat back and enjoyed watching a few TV shows and initially just the intro to a couple of movies. I’m still giving them a good listen and haven’t really had a chance to even break them in or get a feel for what they sound like as home theater speakers. Up until now all of my listening tests were comprised of music samples and REW sweeps. So it’s a little different watching TV shows and movies and hearing how dialogue, conversation and sound effects sound coming from these speakers. But so far I am impressed. They aren’t the most efficient speakers in the world and you can tell by how high the volume knob has to go before you get to reasonable level, but most amplifiers should have no problem running these to ear-screeching levels. They have plenty of bass when sitting up next to a wall and sound very neutral overall. That planar ribbon tweeter still has some sparkle in it, even though it’s been tamed to match the efficiency of the woofer. The MCACC setup made a couple of small corrections in the EQ and still boosted the treble by just a couple of dB. I went back and made a couple of minor adjustments to my liking however. I mean, what does MCACC know anyway? Alright and that’s about all the time I have for one night, please check out the pictures below and let me know what you think of this home theater speaker makeover in the comments below! I’d love to hear your feedback. Thanks for reading and have a good one!

Lots of Pics

Christmas-Themed Pics with New Center Channel

About Dan

For a complete biography, stop by and click around for a while.
This entry was posted in DIY Audio and tagged , , , , , , , , , , , , , , , . Bookmark the permalink.

3 Responses to Introducing the Dayton HiVi RTS181.3 Two-Way Bookshelf Speaker

  1. Don says:

    Hello Dan,
    I found this blog while looking for info on HiVi RT1.3 tweeters. Thinking of using them in a Sony APM speaker rebuild. I noticed the crossover was pretty complex with a lot of parts for a 2-way. Just curious if you modeled any of them in a digital crossover to check how they sounded before settling on crossover transfer function/design?


  2. Dan says:

    Hey Don, thanks for the comment! So I did not do any modeling with digital crossovers. Everything was measured and tested using physical parts with the speakers mounted in the enclosure and in my room. I basically played with multiple cap/inductor combinations and performed measurements of each an every one. I don’t show most of the tweaking in this post, but I started another post that I just never published that goes into much more detail as to my method for creating these specific crossovers. But it basically comes down to having a whole bunch of inductor, cap and resistor values on-hand and then adjusting them in small increments either up down until the desired response is achieved. I do with just the woofer first, then just the tweeter and then the combined responses of both. Also, in order to have some idea of what to do, I took raw driver measurements in the enclosure and imported them into Passive Crossover Designer (PCD). From there I could play around in simulations with different inductor/capacitor values as a starting place for 2nd, 3rd and 4th order slopes. Once I got something that looked close to what I wanted in the sim, I built it and measured it with multiple values of caps and inductors in series or parallel as needed. I tweaked from there as needed based on measurements and listening. When all was said I and done, I built the crossovers using the final stacked/parallel values from my makeshift crossover with single-value parts from Parts Express. As was the case here after about a month of listening I just wasn’t happy with the baffle step compensation and made one final tweak and they’ve been this way ever since. And they still sound fantastic. But for me it’s just lots of measuring and listening and playing with values while understanding to some degree what changing the values will do to the response (go up or down, etc.). The 4th order slope here is not too complicated, but could be made simpler if desired. I was pushing the lower limit of the tweeter, so I wanted a sharp roll-off. If you cross higher, I think you can get away with a lower-order slope. Then L-pad is just level adjustment and then the notch filter was just to make the response look better in the plots really. I couldn’t really hear the difference with or without it, but it does tame the peaking around 15 kHz. Anyway, hope that gives you an idea of what I went through. I’m sure there are other ways, but this works for me!

  3. Don says:

    Thanks Dan, I appreciate the explanation.

Leave a Reply

Your email address will not be published. Required fields are marked *