Latest Project - The DM-4 ARS (Active Reference Series) Loudspeaker

4/15/2017 - Update - The speakers are complete. Stop on over at www.danmarx.org/blog for details.

9/27/2016 - Update - This project is alive and in full force! The details can be found over on my Blog for Whoever located at www.danmarx.org/blog. Stop by and check it out to see how it's coming along. Read below for the introduction to this project and how it all began.

Sometime in the Year 2014

It's been several years since I've designed and built a new full-range speaker, so needless to say I've had a lot of time to consider what I might throw together if I ever found some time. As I researched the web searching for ideas I came up with several designs ranging from small 2-ways, to floorstanding 2-ways, to slim 3-ways, to monsterous line arrays to just about everything in between. I jumped on the forums to see what the latest speaker craze was and while there are certainly no shortage of great speaker designs out there, nothing really got me excited enough to take the plunge and start buying parts. I started looking into commercial designs, from the low-end to the high-end, to even the crazy high-end just to see what's out there. Once again, there is no shortage of ideas for a simple pair of speakers (or how much they can cost!). It's quite amazing really.

But as I thought about what I really wanted to build, what in my mind is the ultimate in two-channel high-fidelity type listening, I kept going back to my days in college when I first was introduced to what one would consider the absolute high-end in audiophile gear. There was a stereo store called Auditions up in SLC Utah where one could audition, hence the name, haha, systems that ranged from the thousands, to the tens of thousands to the hundreds of thousands of dollars. Honestly something I had never even considered, since I'd been buying speaker from Parts Express since I was 13 years old for around $20 a piece and that was for the nicer stuff! I thought I knew what good sound was, but what did I really know?  Except what I had been exposed to up until that point in my life. Well I sat in front of some incredible setups, the kind with amps the size of a ottomons, speaker cable the diameter of a python and speakers with finishes so glossy and shiny you could see yourself in them and the face you made as your jaw dropped when the sales guy told you how much they cost. 

And did they sound good? Of course they did, they sounded amazing. I remember one setup in particular had a pair of what were called Watt/Puppy made by Wilson Audio. A local favorite since they were based out of Provo UT right where I lived. The speakers sounded fantastic and of course being the DIYer that I am my first thought was, I could build a pair of these. I recognized the tweeter immediately as a Focal product, easily available from Madisound, the midrange looked like the classic Usher paper woofer and the two 8"drivers I couldn't make out but figured I could pick something comparable. But the look was classic, the enclosure design simple, yet elegant. After listening to a few tracks (of some music I'd never heard before), I came away thinking that one day I would have a pair of speakers just like those, one day, even if I had to build them myself. Over the years I became more and more fond of the Watt/Puppy design. Not so much because I wanted to buy those exact speakers, but I thought that that style, the sloped baffle, the lower separate cabinet to me just emminated high end.

Maybe it's because the way they looked in that audition room that day, surround by the big ottomon-sized power amps, that made them just seem, I don't know expensive would be one way to put it. But also just like a nice pair of speakers, that anyone who owened thosespeakers cared about music and the way music sounded and spent time just sitting around listening to music. Well it's been 17+ years since that college boy got his first listen of the truly insane and in reality, for me, the untouchable world of audio. So instead I dabble in inexpensive T-amps and Parts Express speaker buyouts to keep me happy and still somewhat in tune with the world of audio and music. Obviously since then the Internet has gotten huge and information on everything is right at your fingertips. So as I ponder my next speaker project for me the ultimate speaker would be to do a Watt/Puppy clone and again mainly just for reasons I've already mentioned - it was just that one speaker that sort of started my interest in higher-end audio, even if I could never afford it. And for me it's not about owning a Wilson Audio product - it's the look, the style, that sort of classic, almost vintage look nowadays of the two separate cabinets, the dual 8" drivers, the angled front baffle and sloped sides, that awesome Focal inverted titanium tweeter that I always wanted. So I think now is the time to make that simple dream a reality and start working on bringing these beasts into the more affordable world of DIY audio.

As it turns out, I'm not the only one who fell for that Watt/Puppy look. As I searched the Internet I came across what I would call the ultimate Watt/Puppy clone made by Albert VonSchweikert called the VR-5 SE speakers. While there is much less info about these speakers I immediately saw the resemblance as others had previously noted, probably why these speakers showed up while searching for the Watt/Puppy. But I was immediately impressed by a few things about his design that I liked better than Wilson's. One of my biggest issues with the Puppy was trying to fit two 8" drivers into cabinet the size of the Puppy without it being a compromised design, either high f3 or misaligned or both. Why is the Puppy so small? And that port in the rear is like 2" in diameter? Unibox will tell you out right that that port diameter is way too small to not cause chuffing due to high air speed. So I needed a bigger box and a bigger port! 

That's where the VR-5's design nailed it. While I don't have the exact dimensions (update: 29"h x 11"w x 24"d), that massive flared port in the front certainly looks big enough to keep the port air speed low enough to not be audible and at 24" deep, there's plenty of room for even the most volume-hungry 8" drivers to be happy. While the driver selection is completely different from Wilson, what appear to be mostly SEAS Excel drivers, the most unique aspect of the design in the rear-firing ambience tweeter. Uh oh, I can hear the DIY crowd moaning already. Hear we go again with some Bose-esque reflecting sound mumbo jumbo, can't make a real speaker have great imaging and depth of sound, just drop in some speakers in the back and boom, instance ambience. Right? Well, let's hear the guy out. It's not the first speaker to have a rear-firing driver, in fact it's quite common in a lot of commercial speakers these days. I've never built any speakers like that but certainly could imagine how a driver firing at the wall behind the speaker could give that sort of live effect. Maybe it's psuedo and forced or maybe it sounds great and so what if I have two different tweeters in my speakers? I've seen stranger designs with all the speakers on the front baffle. Besides, you throw an L-pad on it and give it a fancier name, call it the "Ambience Level" selector and if it sounds like crap you just lower the little knob to all the way to the left and you've got a plane jane front firing speaker. What's to lose?

Anyway, I've gone on, so after remissing over the Watt/Puppy I'm afraid my speaker sytem is going to have to replicate more of the elements of the VR-5's because to me they just suit the look and style that I'm going for that much better. Especially in the upper cabinet with the mid and tweeter, it's not so boxy as the Watt/Puppy and should be easier to build. Though in all fairness, they basically look like the same speaker. And once I'm done with it it will really look like neither speaker. I have no intention of tyring to copy the exact volume, finish, driver types, crossover, etc. of either speaker. To do so would be futile on my part and not my intent. So on looks alone from the outside, sure it will be close, call a clone or a copy if you will, but aside from some basic inspiration from these other two designs, the entire DM-4 speaker system will be entirely my own design and creation from top to bottom.

It's funny because as I envision starting this project, obviously I'm pretty stoked because I'm anxious to build a pair of speakers since it's been so long, but I feel as though this is one of those speaker project that sort of gets no respect. No respect from the DIY crowd because it's a copycat of a commercial speaker, and an expensive one at that and no respect from the audiophile crowd because, well, it's a knock-off and good luck getting it to sound as good as the originals. I remember reading about a guy who posted a bunch of info about a Watt/Puppy clone he was excited to start working on but everyone on the forums was so negative about the whole project that he scrapped it, sold the speakers he had and never pursued it again. And that's too bad really. Probably one of the reasons I won't be spamming the forums with this project. I'm afraid of the negativity I might receive. Even though I would love input from the general population of fellow speaker builders, as I truly respect their input and their experience, I might have to venture this one on my own. This probably won't be a project others will duplicate and I that's fine with me. It won't stop me from blogging about it on my personal site and for the casual reader it will be nothing more than a stop along the way about just another speaker project. And with that, here's what I've got so far for the design.
  

I have drafted up some plans. They are much less Watt/Puppy and much more VR-5 SE but after looking at the two designs, I think I like this look better. Not an exact copy but they don't need to be. I was mainly going for something similar from an aethetics standpoint. These speakers look awesome to me and when I picture a pair of these in my living room and just can't wait to get started. I think they'll be awesome.

Here's a rough CAD drawing of what they would look like:

So I'm just going to get some thoughts down on paper before I begin this project: 

• Combined visual replica of the Watt/Puppy and VR-5 SE Loudspeakers – combining the best of both speaker designs into one unique speaker design 
• Budget of about $1000 including amplification, crossovers, drivers, wood, etc.
• Use of the latest Dayton Audio RS Paper Cone series for all drivers with phase plugs (for that VR-5’s Seas Excel look)
• Dual 8” drivers in low-frequency cabinet – Dual 8 ohms in parallel – total 4 ohms with the RS225P-8
• 7” Driver in upper cabinet – 4 or 8 ohm? Still don’t know yet RS180P-4 or 8 - probably 4 ohm
• 1-1/8” Soft Dome Silk Tweeter – only the 4 ohm version is available (not a ScanSpeak Revelator or a Focal tweeter but will have to do for the budget I’m working with). Plus it has the metal dome grill which, as a parent of young children, I can't do without.
• VR-5’s use an ambience tweeter in the rear of the cabinet – I’d like to copy that design somehow possibly mating the Dayton with a Tymphany H26TG06-06 1" Silk Dome Tweeter. It is a horn loaded tweeter than looks just like the one in the VR-5’s. Might actually be the same one. It’s is 6 ohms with a super high SPL at 1W/1M. Will need padding so the impedance in parallel with the Dayton tweeter shouldn’t be a problem as it will also most likely need padding. Would be cool to put it on a switch or an L-pad so it could be toned down or removed all together. - After further consideration, I think I will lose the rear-firing ambience tweeter. I have no way to really model the effect of the tweeter  and I am afraid it may degrade imaging. Too many unknowns with such a setup. Best just to avoid it for this build. So instead of the DM-5 (which had 5 drivers), I'm calling it the DM-4, which has only four drivers.
• The Puppy uses a small rear-firing port that is way too small for the amount of air that can be displaced by the dual 8” drivers according to Unibox. The VR-5 improves that design with a large 4” port with a massive flair. Looks just like the flared ports from PE. I don’t like the fact that it is front-firing, as you can see into the box and disrupts the clean look of the front baffle. I’d rather it be in the rear like the Puppy. That port diameter models well for air speed and anything less would be a compromise.
• The Watt’s have a trapezoidal/pyramid shaped enclosure that looks nice but will be difficult to make. While the VR-5 cabinet looks like it's square with a chamfer the mimics the Watt look but would be easier to make. I may compromise and do both – a more subtle trapezoidal/pyramid look with the chamfered front together. One that could be built as box and then chamfered with a table saw. The non-parallel walls should be considered. The main focus of course being a true time-alignened design with the acoustic centers of each driver physically in the same plane. This would be ideal when used with an inexpensive active crossover that does not have a time alignment feature.
• The orginal Puppy enclosure size just isn’t optimal for the two Dayton drivers. Each driver wants to see about 42L each, or 84L totalf or a standard QB3 alignement. The exact Puppy volume with bracing will only provide about 57L total for both drivers. So with the Dayton drivers with that volume, the F3 suffers, and the tuning can’t be much lower than 40 Hz before the alignment becomes more and more mistuned from a standard design. I’d like to go bigger without changing the visual aesthetics of the lower enclosure too much
• The VR-5 speakers are super deep – 24” deep as compared to the Puppy’s 18”. So the idea would be to keep a similar width/height dimensions as the Puppy and increase the depth until I reach the desired volume. Which should be about 21” deep. But dang, that is a deep speaker box. In the end, the box is going to be as big as needed to support a standard design for the dual 8" woofers. I've got some initial numbers from Unibox that I'll be writing about in detail as the project progresses.
• The Watt/Puppy is constructed of a proprietary composite material that there is no way any DIYer is going to replicate but the VR-5 speakers use a type of 4-layer damping method against standard 18mm MDF. They have a write-up on it and some pictures which looks like it can be replicated to some degree. While the exact materials may not be available, PE sells an acoustic foam that looks like the 3rd layer of dampening, while the 4th layer is probably just acoustic stuff. It’s the 1st layer of damping that is glued right against the MDF that will be tricky. Dynamat may work in this case, although they say it has a high Q and is very rigid, some sort of crushed rock, so I imagine something stiffer but don’t know what just yet. Then the 2nd layer is also some unkown material. Will take some digging. But I agree with the conscept of multiple material types with varying Q glued to eachother to the MDF to reduce resonances. 
• Otherwise I may just do standard MDF all around with lots of window bracing, foam/egg crate and polyfill to reduce cabinet resonsances and absorb internal standing waves.
• The Watt uses an interchangeable port in the upper cabinet that could easily be replicated with some of the PE adjustable ports. I would like to duplicate this even though the VR-5s appear to have a sealed upper cabinet. I’d like the option of multiple tuning choices or plugging the port and making it sealed. Or just making it sealed. Sealed would be easier and with the crossover, I'm not sure the port even does anything if the box is crossovered well above fB. Though it would be nice if the enclosure could stand on its own, without the woofer cabinet. So porting it would be ideal in that case. 
• I think I have decided to use a sealed enclosure, since the crossover frequency will probably be in the upper 300 Hz region and I really don't think I will ever use these as a standalone speaker without the lower woofer/cabinet.
• The Watt/Puppy has the option of adjusting the angle of the cabinet so allow the user the tweak the sound. The VR-5 just use a set of toe spikes and the angle is fixed. I think I will go with the fixed angle as it looks cleaner. Besides a properly aligned crossover and the time-aligned design shouldn’t need to be tweaked. If it sounds better up or down a couple of degrees, I might just get some adjustable toe spikes so I can adjust it a few degrees up or down.
• Finish is up for debate. Might be high gloss piano black. Might be veneer. I will see about taking the cabinets to a local body shop and having them paint them for me. That’s the only way to get a truly flawless finish. But I may just paint them in my garage too as it will be cheaper but more time consuming. I might veneer them as well.
• The crossover will have to be custom for the drivers in the cabinets. I intend on doing FR and impedance plots of all drivers in the cabinets and designing a crossover to match. There’s no free ride with this one. Just need to get out there and get the measurements and do the design. Passive Crossover Designer 7 will be used.
• After further consideratio on the crossover, I just might go with an all-active 3-way design. Right now I'm looking into active crossover options, maybe from the pro audio world (like a dBx DriveRack PA+) or a MiniDSP. There's a lot of great info out there on active speaker setups. And honestly part of the appeal for me right now is just the cost. An active crossover and a 6-channel amplifier honestly are not going to be that much more expensive than buying really good air-core inductors and audiophile grade metallized polypropyline capactors for a passive setup. Plus the ability to tweak the crossover points and levels, and depending on the type of crossover, the type of slope and everything else real time is really appealing to a tweaker like myself. I might start out simple though with an analog 3-way active unit like the dBx 234s. I've also read about just using a computer and a good sound card with a software-based active crossover just to get into it for cheap and see if I like the concept or not. But it's farther from a set it and forget it approach. But I just don't want to botch the entire design with a crappy passive crossover that wasn't designed right because I don't have the tools or experience to really do it well. It doesn't mean I won't botch an active setup either!
• If I go active, the amplifier I am looking at is the Emotiva UPA-700. For a limited time, Emotiva is letting these amps go for $399 until they are gone. Supplies are limited, but this amp, dollar per watt is approaching T-amp pricing so it's a real steal for what should be a decent amplifier. At least as good as anything else in this setup I think. Plus it's just a good looking amp with some decent specs to back it up.
• I’d like to invest in some accelerometers and the measurement software so I can do acceleration plots on the cabinets. It would mainly be for my own experience. Taming cabinet resonances can be a science in its own right.
• All data initially will be taken off PEs website, once I get the drivers in hand, I intend on measuring them in a "similar" cabinet/baffle configuration to get FR plots of each speaker. Then I can model the crossover more accurately. Since I'd like to go active, I can deterimine the crossover points and levels in PCD, run REW on just the crossover to check the transfer function of the crossover slopes/points/gains, run each crossover to each driver, measure again, capture FR plots, and then run the entire speaker through REW and tweak each level/crossover point as appropriate to achieve the desired FR. At least that's the theory anyway. I know there's some timing elements in this flow as well, but just not sure how to do that part yet.

9-28-2014 - Speakers Have Been Purchased!!!

When I built the massive (4) TC sounds subwoofers for my brother's home theater, we had agreed on a rather simple form of payment - speakers, instead of cash. He agreed to fund my next speaker project so long as it was reasonable and within an agreed upon amount. So tonight I placed my order from PE for all the goodies I would need to get started on this project. It doesn't represent an entire list of things to completely finish the project, but consistutes the main components and biggest expenses.  Here's the parts list below for the DM-4 Reference Speaker System:

Item	Qty
	Dayton Audio RS225P-8 8" Reference Paper Woofer 8 Ohm Part # 295-367	4
	Dayton Audio RS28F-4 1-1/8" Silk Dome Tweeter Part # 275-140	2
	Dayton Audio RS180P-4 7" Reference Paper Woofer 4 Ohm Part # 295-375	2
	Dayton Audio BPP-G Premium Binding Post Pair Gold Part # 091-620	8
	#6 x 1" Deep Thread Pan Head Screws Black 100 Pcs. Part # 081-440	1
	#8 x 1" Deep Thread Pan Head Screws Black 100 Pcs. Part # 081-425	1
	Wired Home SKRL-14-50 14 AWG OFC Speaker Wire 50 ft. Part # 100-020	1
	Dayton Audio DBPP-BK Double Binding Post Plate Black Anodized Part # 091-612	4
	Precision Port 4" Flared Speaker Cabinet Port Tube Kit Part # 268-352	2
	Jasper Circle Jig Model 240 Part # 365-265	1
	1/4" (16-14) Female Disconnect Crimp Terminal Blue 50 Pcs. Part # 095-280	1
	Speaker Gasketing Tape 1/8" x 1/2" x 50 ft. Roll Part # 260-542	1

I have decided to run this setup as a completely active speaker system. After spending many hours with PCD7 (Passive Crossover Designer) tweaking with both passive and active designs and considereing the costs associated with implementing either design, I really think that going active for these speakers is going to be the best for several reasons which hopefully I have the time to discuss in detail. Oddly enough, the active crossover models so well and doesn't even require any fancy compensation networks, notch networks, unique slopes etc to get a nice flat FR. I read a comment from somone that said since you can't drop in an off-the-shelf passive crossover into a speaker and get it to sound good, what makes you think you can drop an off-the-shelf active crossover into a speaker system and get it to sound any better? I thought, there is some merit to that statement. Would a textbook active crossover behave or sound any different than a textbook passive one? Do active crossovers require the same level of complexity in their shaping abilities as a passive system for the same set of drivers? Stay tuned to find out....

9-29-2014 Update: The amp has also been purchased! For this setup I chose the Emotiva UPA-700. Once I deciced to go with an active crossover, I think the amplifier choice sort of  just fell into place. Not many multi-channel amps out there with the specs like the Emotiva at this price point. A friend at work had turned me on to Emotiva a while back, so he kept me in the loop of sales and factory renewed deals they had from time to time. When a UPA-700 showed up in their list of products they were discontinuing, and therefor offering some huge savings, I really couldn't pass up the opportunity to own one of their products. It fit the bill nicely and should match well with this speaker arrangement. This amp provides 80 watts to all 7 channels driven simultaneously into 8 ohms, or as the case with the DM-4 speakers here, 100 Watts into 6 channels at 4 ohms. You can read the rest the datasheet on their website. But dollar per watt/per channel, it's not much more expensive than a decent 2-channel Topping or Lepai class D amp once you figure in that you would need to buy three of them. But this thing comes in an all-in-one aluminum package, with "fully discrete, dual differential, high current, short signal path Class A/B topology", a bandwidth of 10-80,000 Hz, THD of <0.01%, a masssive toroidal transformer, big ole' capactitors, triggered input/outputs, LED status lights, plus I think it just looks freaking aswesome. Hey, looks count too right? They better, because if this whole setup ends up sounds like complete garbage, the only thing to fall back on will be its good looks! Plus the amp shares a single power supply, so the four bass drivers should be able to capture some of that extra reserve on two channels that the midranges (and especially the tweeters) aren't using on the other four channels. Plus the seventh channel isn't being used at all. I suspect even at running at max output, this system should have very low audible distortion. Plus the amp's power output matches nicely with the rated capabilities of the Dayton Audio drivers. Any more power than this would likely damage the speakers. So there wasn't much sense in getting some crazy multi-thousand watt amp. Even under the notion of "more headroom" is always better. My wallet disagrees.

10-11-2014 - Crossover Options and General Thoughts and Ramblings

So onto some initial thoughts about the crossovers. There are several options when it comes to commercially available active crossovers. The main criteria for choosing one over the other for me will be two things - required options and price. Based on the driver selection and some initial simulations I think I can get away with a very simple stereo 3-way crossover. At least that's what I would like. Since the sloped baffle driver arrangement already corrects for properly aligning the timing of the drivers, I'm not convinced I need time correction capabilities in the crossover. It might be nice to have as an option, to be able to tweak that parameter on the fly as needed, get everything dead nuts on, but if the drivers are physically time aligned already, the crossover shouldn't need to do anything extra in that department. Secondly, the drivers are essentially a matched set - all Dayton Audio Paper Cone Reference Series matched with the RS-25F Silk Done tweeter. This combiniation (also with the aluminum series) has proven to make a great speaker in many DIY setups. One of which I have in my own home theater. So I don't believe the crossover will need notch filters or other unique shaping capabilities. Each driver in its own right has a very nice, very flat, frequency response. (With the exception of baffle step compensation which I'm still debating on implementing or not).

For example the 8" RS225-8 drivers should be flat (in the enclosure) from 40 Hz to over 2,000 Hz before cone break-up modes dominate. I'll probably cross over the woofer/midrange between 300 - 500 Hz, well below the region where cone breakup modes are a problem. And with 24 dB/octave slopes, which are typical for most active crossovers, I can expect the problematic region to be well below 72 dB down (3 octaves) or even 96 dB (4 octaves). The midrange is a little trickier, as cone break up modes for that massive 7" woofer are a lot lower than say with a 6" or even a 5" midrange would be. Most people would argue that a 7" woofer is hardly a midrange at all as it requires either a pretty low low-pass filter or a high-order filter to ensure the cone modes are supressed enough to not be audible. Which in turn requires an equally very low high-pass filter or high-order filter on the tweeter which puts the tweeter at risk of damge or distortion from the low crossover point. I get it, I don't see too many speakers with 8" woofers and 7" midranges. Part of what I love about the Watt/Puppy and VR-5 SE speakers however is that massive midrange. It's the look, that classic high-end look (and yes, it's totally subjective, I know). 

I think this is where active crossovers really shine because the high-order crossover is taken care of.  Those 24dB/octave slopes are pretty steep which aren't as easy to design properly in a passive setup. Plus they get expensive, requiring at least double the components than a 12 dB/octave one. I believe that the RS-28F tweeter should have no problem operating down to 1,600 Hz if needed with an active crossover. That's just about 3x the fs of the tweeter and well within the recommended limits. In fact, Parts Express even claims you can run these tweeters down to 1,000 Hz (with the appropriate filter of course) which they don't say exactly what that is. So assuming they don't think the average person has a 36dB or 48dB/octave crossover up their sleeve, I'm going to bet that a 24dB/octave slope will be sufficient. 

Which brings be back to the RS-180P 7" midrange driver. How high can that driver be crossed over before cone breakup modes become audible and problematic? (not to mention beaming problems as well). Well according to the datasheet, the region to avoid is anyting above 4kHz. And more realisitcally, you want to be below 3kHz for the crossover depeding on the order of the crossover. Again, with a 24 dB/octave crossover, I'm guessing that anything below 2 kHz is going to be acceptable and while dropping all the way to 1 kHz might technically be achieveable, I'm not sure it's required for this setup. But so what? Why do I even have to pick a single frequency anyway? Active is so easy, you can choose almost any crossover point you want and listen/measure do whatever and then tweak it until it sounds to your liking. There is no worry about picking the perfect point, it's all adjustable with active crossovers. Very little commitment. And for those of you with committment issues, active crossovers are the perfect outlet for letting your desire to run free go wild. Possibilities are really endless.

Which brings me to the last point of the purpose of the crossover, it's not just to cross over the broad freqency range of human hearing perfectly into 3 different drivers but is to ensure that the phase and power level of each of these drivers is properly matched for a consistant and audibly pleasing experience by the listener. With passive crossovers, padding or leveling is accomplished with resistors (or RC or RL network in the case of shelf filters). With an active setup all it takes is a knob or a dial or a setting to select any range of individual driver level usually anywhere from -12 to +12 dB. No need to worry too much about pretermining how bright your want that tweeter, or how forward you like your midrange, or whether or not you've got too little or too much bass drowning out everything else. Indepent control over the level of each driver make adjustments easy and on the fly. And that's another reason why I like active crossovers. The last item I mentioned was phase consistancy. Which for now I will have to defer to another time. More on that later.

Of course this doesn't mean you get away scott free going active. Active setups still require meticulous measurements with a calibrate microphone of each driver to be able to "see" what everything looks like. Otherwise it's a guessing game and you're relying solely on your ears to tell you what sounds best. My plan basically goes like this:

1. Measure the drivers, individually, near-field, in the enclosures, in the room, without the active crossover, direct (maybe a very low crossover point on the tweeter, I don't want to damage them while measuring them).
2. Import measured FR and impedance plots for each from driver from REW into PCD7 
3. Design the crossover points and define the levels required for a flat in-room response 
4. Evaluate the transfer function of each crossover
5. Set the crossover and measure the actual transfer function with REW (loop-back through each channel of the crossover into my laptop)
6. Tweak the crossover as needed to match the simulation in PCD7
7. Connect the crossover to the amp and speakers
8. Re-measure each driver indivually with the predefined crossover points and levels in-room near-field
9. Tweak as needed the levels and points until they match the simulation
10. Then measure all the drivers together, to ensure they sum properly at the crossover point. 
11. Again tweak as necessary until in-room response matches desired response
12. Then call it day, sit back and listen to my favorite tunes!

Decisions, decisions, decisions! Almost any of these options should work perfectly fine for my application so it really depends on how much I want to tweak and how much I want to spend. I'm basically in the research mode for the crossover and am reading everything I can about the pros and cons of the various units listed here. I've still got some time though, since I haven't even begun the speaker boxes yet but I hope to norrow it down in the next few months. Both the active digital approach and analog approaches appeal to me for different reasons. I could be swayed on either one. Time will tell.

Next on the adgenda is to post some screenshots of the simulation data for the enclosures and the crossovers. I've got everything pretty much to a point that I'm satisfied for now, before getting better measured data on all my speakers, just with the catalog data from Part Express. But it's a start and at least gets me in the ball park.

And that's all the upates for now. I also will post some nice high-res pics of all the drivers and my sweet new Emotiva UPA-700 amp. I've been listening to the amp on a set of older speakers and it sounds awesome. For just two channel listening, nothing is bi-amped or tri-amped yet, that amp has a lot of power. I can crank it up way louder than the speakers can handle at the moment, but it sounds clean and unaltered. But my current speakers lack quite a bit of the sound I'm really looking for, so it makes me even more excited to get started on these new speakers. I know they will sound sooooo much better. Later!

Picture of all the goodies from Parts Express!

10/16/2014 - Thoughts on MiniDSP

So I've been spending a lot of time over at HomeTheaterShack reading up on the capabilities of MiniDSP. While inititally I thought there was some inherenent benefit to keep the crossover entirely in the analog domain, I can't help but be intrigued by the ultimate versatility of MiniDSP. I was a little turned off originally by comments made by posters in the forums about the need to keep the ADCs "fully saturated" but "just below clipping" in order to take advantage of every bit, or every LSB of the ADC, otherwise the sound would be compromised, high noise floor, and not sound at its absolute best. But in order to connect your system in this fasion, requires a pre-amp set to this magical limit, and then six adjustable outputs that drive the amps, essentially required a 6-channel pre-amp and a 2-channel pre-amp. It just didn't seam practical. I like to follow the KISS principle, Keep It Simple, Stupid. But after reading Wayne P's post over at HTS about 24-bit processing and the whole arguement over using every last bit, he pretty much debunked, in my opinion, this idea of using needed to utilize every LSB when you've got a 24-bit processor. That the depth of a 24-bit signal is so deep that any degradation in fidelity or dyamnic range due to not using every bit is so miniscul that it's negligible making the whole argument moot. 

But in order for me to fully convince myself, I wanted to take a look at the levels present in my system and see just where the voltages are set so that I could take full advantage of every bit of a 24-bit processor where most of those bits aren't used simply due to the level of the signal. The original MiniDSP in kit form can be configured in two flavors - with 0.9Vrms input or 2.0Vrms input. (Note that in either case, the output is always 0.9Vrms). So how does one choose which input level they should use? I grabbed the data sheets for both my pre-amp and my amp to see just what kind of light they could shed on the subject. For my pre I'm actually using just a Denon AVR-1801 receiver. While technically yes, it's just a run-of-the mill 5.1 receiver, it actually doubles as a perfect pre-amp. Not to mention in keeping in tune with my goal of having this actually be a "budget" 2-channel setup, I bought the Denon off a guy on Craiglist for $50. Barely used, with the remote. It was in perfect working condition. The datasheet for the PRE OUT specs the response from 10-100kHz with 100 dB S/N ratio, THD of 0.008% and a rated output of 1.2V. My first though was, based on that voltage, I should set the jumper to the 2.0Vrms setting. 1.2 V is clearly higher than 0.9 and therefore the dreaded clipping would surely occur. Nevermind worrying about keeping the ADC's saturated, they'd be oversatured at max output.

So the second piece to this puzzle is how much input level does the UPA-700 need in order to reach maximum output power? So with a quick check on the datasheet, it would appear that in order to achieve full rated power of nearly 110W per channel, with 29.5 dB of gain, the input voltage should not need to exceed 0.705 Vrms. Perfect! So that settles it, I set the jumper of MiniDSP to the input sensitivy of 0.9 Vrms. That way I can guarantee that at low listening levels, if it mattered at all, I will still be utilizing as many LSBs as can be expected, but that at full volume, should I ever should to listen to music that loud, I won't exceed the max input of the MiniDSP and should never experience the dreaded clipping. This assumes that the input/output ratio of the MiniDSP is 1:1.

Denon AVR-1801 Specs

Emotiva UPA-700 Specs

The MiniDSP in kit form actually provides one of the simplist and cheapest, yet most-versatile, crossover option, since it's much more than just a fancy crossover. It offers programmable delays, PEQs, response shaping, shelf filters, notch filters, adjustable filter types and filter slopes - all per channel/driver. Plus a master PEQ which can be used for taming room responses. Plus set it and forget it operation. What's not to like? I like the idea of the kit version because then I can build a really nice chassis/enclosure for it so that it integrates nicely with all my other gear. Maybe something like this or this. From MiniDSP's own site, they give a great example of how to create a 3-way crossover.