12/27/02 -- FINAL Dayton Budget MTM update

    This is the final DB-MTM update because you can stick a fork in them -- they're done.  They sound great and the full write-up will be posted soon.  I'm on vacation, but I'm working on a few projects, so the final write-up may take a few more days.   

    For those of you who have been asking, here is a link to the Dayton Budget MTM parts list. The total cost is about $162, for what I would call the "frugal" version. This, fyi, is the version that I am building for a friend, based on listening test comparing non-polar electrolytic caps with polypropylene caps in the lowpass (in other words, I can't really hear a difference on average equipment -- a Denon DRA 375RD receiver and similarly priced NAD CD player with a combined cost of about $550). On higher-end equipment (300b SET tubes amps, Musical Fidelity A3.2CD player, Conrad Johnson 17-LS preamp -- total estimated retail equipment value of around $10,000) I can hear the difference... and it's not enough to make me want to spend and extra $40 on all PP caps in the lowpass (and remember, this is from a guy who would actually buy a $4,000 preamp... on Ebay, of course, for $1600 - NIB!). For those of you who want to upgrade the parts, feel free (I recommend going to a 14 awg, .5 mH inductor as the first upgrade -- everything after that is up to your own personal preference) :-)

    Personally, I think the electrolytics sound a little warmer, although a hair grainier, on the high-end system. It's certainly nothing you would pick up on without having heard both versions on both an average and a "high-end" audio front-end.  I recommend all electrolytics in the lowpass, but you're free to go with polypropylene caps.  The highpass uses all polypropylene caps, however, and I don't recommend using electrolytics for the DB-MTM highpass network.

    A few other things -- the "notes" column shows combining caps. Don't attempt to reconcile the math, because it doesn't work out (on purpose). For example, it might say to combine a 17 uF cap with a 6.8 uF cap to make a 23.5 uF cap. That's intended to suggest which caps to combine on the parts list to make the cap shown on the schematic for this project.

    If anyone finds any errors or has any questions, please let me know.

 

12/18/02 -- Quickest Dayton Budget MTM update

    I couldn't wait for the parts to arrive to build the prototype crossover for the DB-MTM, so I built one out of parts I had in my big boxes 'o crossover parts.  I unwound a few inductors to make the values needed (which is what you should do, rather than try to use "stock" components).  More on this when the crossover is finalized.

    I took a few measurements and things matched up nicely (not perfect, but nicely).  I think the fact that I wasn't able to use the components specified in the crossover (the lowpass caps are off by a few μF here and there) leading to the slightly larger than intended midrange depression shown in the response graph below (about 1 more db than I wanted, but certainly nothing that causes a problem -- I really like the way this puppy sounds, as I'm listening to some Dire Straights "Brothers In Arms" through it in mono as I write this).  The midrange dip really relaxes the midrange in a way that I think most people will like, even if it doesn't lessen when I get my parts shipment and build more exact prototypes. 

    On the subject of this midrange dip, according to Ed Frias, he intentionally included a 3-4db dip in this same region in his AR.com DIY kit that so many people have built and enjoyed.  The famed "BBC" dip, as it's called when you depress the midrange by a few db in the 3khz region, is a psychoacoustic phenomena that many designers intentionally build into their speakers to avoid "glare" that many people find objectionable.  Remember, flat response isn't the ultimate objective, good sound is.  But all that aside, for those of you who just want the bottom line and are ready to fire up those soldering irons, go for it.  Most likely, when I remeasure the crossover with the proper components, the response will be closer to the target.  Even if it doesn't, I'm encouraged enough by the preliminary tests to feel comfortable that you early-adopters out there are going to be happy.

    I plan to finalize the crossover, schematics, and write-up after my parts arrive and I get some time, over the holidays, to listen to these guys in stereo... but I won't be going back to the drawing board.  I'll experiment a bit with the BBC dip, but I don't anticipate any significant changes, if any, from the crossover published yesterday (which had an error in it and has been fixed in the current version).  One thing I must say is that the added peak filtering in the lowpass really pays off -- the midrange on these is quite good for such an inexpensive woofer, and the all-important power region really plays nicely.  Enjoy!

 

12/17/02 -- Quicker Dayton Budget MTM update

    I know, I know... you're getting itchy to hit the order button to get those parts on the brown truck and have some toys to play with on your vacation (hey!  I'll be placing my order tomorrow, too!)  But before I get to the good stuff, why is it that so many of you are chomping at the bit to break out the power tools and build another set of speakers, when you know as well as I do that the thought of "some assembly required" appearing on the Barbie Dream House that's going to be under the tree on Christmas morning is sending shivers down your spine?  Take a quick break and ponder that in individual discussion groups.

    Ok, you're ready?  Then here goes... the first "semi-official prototype" crossover that's ready for prime-time.

 

    But before you get too excited, here's the catch -- I'm ordering the parts to build this one tomorrow, too.  That means that I'm not promising that this is a final, perfect, or otherwise Rev-approved crossover.  What it means is that I've played this game enough times to know when to shut-up and let Calsod do the talking... and the graph on the right shows what it said.  Response looks good (4th order LR crossover at 2650 hz, solid 4 ohm nominal impedance, good phase tracking in the crossover region -- in other words, all the usual suspects were witnessed at the scene).  That's good enough for me to get started, so those of you who don't want to wait around for the final version can get going.  Also, before you ask, get the closest component values you can find -- the final, optimized version will include commonly available or easily created component values.  Feel free to ask questions on the message board.

    So what about this crossover?  Why so dang complicated?  Why 4th order slopes?  Ok, fair questions.  Here's my thinking.  First, the crossover was designed with the primary objective of flattest possible response in the critical midrange region (200hz-1500hz).  Quite often, you see a nasty little dip that results from a baffle step compensation circuit.  While full BSC is included, I don't often use LR networks -- and in this case, each attempt resulted in poor impedance characteristics.  More importantly, it was easier and more accurate to use series notch filters and inductors to compensator for BSC and tame the peak in the woofer's response at around 4500hz.  Also, if you really want a clear illustration of baffle step rise, take a look at the frequency response for the Dayton 295-300, below.  You clearly see the transition from 2π radiation (half space) to 4π radiation (full space) as the frequency approaches 500hz.  Note that the response was obtained at 1.5m (@2.83v).


(click to enlarge)

    So, why 4th order slopes?  Simple.  Impedance and sound quality.  I definitely preferred the sound of this system, in initial testing, with the stronger suppression of the peak at 4.5khz and more rapid rolloff -- probably because MTM designs, in my opinion, tend to glare a little in the upper midrange, so they need to be voiced a little on the laid-back side (which, as those of you who have built any on my previous designs know, is my general preference for voicing).  The transition to the tweeter was cleaner and less upper-midrange "grunge" was noticed.  Also, it allowed me more flexibility in the response in the 2-5khz range (it took a while, but I managed to get a nice little dip at 3khz, also called a BBC dip, and about 1db, or so, or recess in the upper midrange).  This, to me, provides optimal soundstage presentation and a less fatiguing speaker (it also adheres to Paul "Peppy" Holsopple's "smile profile" theory).  I hope my initial impressions are borne-out in this crossover implementation, but like I said earlier, I've been down this road enough times to know where it leads and I'd be picking through a box of crossover parts, rather than just ordering them by the pair, if I didn't have a high-degree of confidence in my simulation -- and I did cheat a bit and use the virtual crossover simulator in SoundEasy 6 to get an idea of how this will sound.  For those of you who don't know what that is, it's a feature of SoundEasy that's akin to using an active crossover to listen to a crossover in a speaker, using your computer as the active crossover).

    Lastly, why so complicated?  Simple... it's not all that complicated.  The resistors in the highpass were put there to control the resonance peak resulting from the interaction of the high and low pass filters and to maintain an acceptable impedance minimum (3.1 ohms).  The dual notch filters in the low pass were designed by hand and then optimized, and coupled with the rest of the circuit, to carefully control the response in the critical portion of the midrange region and to tame the aforementioned 4.5khz peak through the crossover region.  They also allowed me to obtain a more benign impedance profile.

    That's all for now.  Stay tuned.

12/13/02 -- Quick Dayton Budget MTM update

    Me again.  Now that we have all that "welcome back" stuff out of the way, it's time to get back to business.  I'll be getting an archive up and running soon, but with the holidays you can expect more hit 'n run updates than full-blown articles and projects, so lets get right to it, shall we?

    Quite a few of asked for some box dimension info, so you can start hacking up that MDF in your garages and get started on your Dayton Budget MTM's.  That's the first version of the new project that I'll be posting.  A few of you have suggested that you'll wait to cut the baffle until I post which version is best.  More likely, though, I'll post what I think is the optimum crossover for each version and I'll let you guys sort out which is best.  Truthfully, though, I don't think there will be a "best", there will just be a few versions that you can pick and choose from, based on your own tastes.  If I say one version is better, someone else will tell me I'm wrong :)  Speakers are like wine -- there are good ones and bad ones and some that people like better than others, but if there was a "best" wine, why are wine lists so damn long?  The best answer I can give is that everyone has a different taste and what's best to one is merely mediocre to another.  My wife likes white zinfandel and I like Chianti.  Which one of us is wrong?  Neither you say?  Then there's your answer to every question ever posted on any message board asking which [insert audio component of interest here] is "best".

    Cutting to the chase, I didn't get a chance to finish the mechanical drawings on the enclosures, get them scanned, etc., etc., so here's the deal

:
(click to enlarge)

●  External dimension: 8.5"x24"x12.5"
●  Internal dimensions: 7"x22.5"x11"
●  Tuning frequency: 47hz (that's a 3" ID vent @ 6.75")
●  Line the sides, top, and bottom with acoustic eggcrate foam (1 1/2" peak to peak stuff from PE).  We'll talk about why I want you to leave the back untreated later.
●  One "h-brace" (any sort of brace will do here, just make sure enough material is removed from the piece to ensure clean air flow).  Location isn't all that critical, so I put it behind the lower woofer.  If you're definitely doing the MTM, put it behind the tweeter.  There's no rocket science on the brace, but if you have a question, please post in the forum.

    ok... of all that, the only really important thing is the front baffle dimension of 8.5" -- much variance from that will affect the overall frequency response. You can reshape and resize the box, if you want to change the shape or tuning frequency, but much more the +/- 1/2" in the baffle width may result in undesirable frequency response in the critical 500hz-1500hz region (I said "may", not that it definitely "will").  Not following these directions to the letter will not result your banishment from The Reverend's Church of the Screaming Tweeter...  although you may be required to sit way in the back, near Lefty, if you really get whacky and then bitch to me that something's mucked up.

    Here's an important design factor -- all drivers WILL be flush mounted with a center-to-center spacing of 4 7/8" (this is for the MTM with the dayton silk dome tweeter and 295-300 woofers and will only leave about 1/8" clearance between the frames of the woofers and the tweeter).  The drawing shows 4.75" CTC, but that puts the frame in contact.  The updated drawings are sitting on my desk at work, so I'll try to get them scanned and posted over the weekend.  Where are the obligatory enclosure photos?  Glad you asked (click to enlarge). . .

   

    On another subject, someone asked me about the Eros MKII crossover.  Here's a picture of one of the ones I built.  You might notice my technique for preparing the foil inductors for use.  What I do is enclose the leads in heat shrink tubing (to avoid shorted connections) and put spades on the end (the ones shown are gold plated ones from Radio Shack -- these are pretty nice spades, even though they're a little expensive).  The connections are all point to point and many connections are encased in heat shrink tubing for long-life, good looks, and because we all know that's the best way to get chicks to dig us.  If anyone has a burning desire to hear more on this subject, let me know and I'll be glad to provide you with more crossover construction information that's bound to make you a hit with the ladies that will make your crossovers more slick.  If you really wanna be a hit with the ladies, you need to get a cool website and a shaggadelic nickname like "The Rev" be rich and good looking.

    Ok, that's all for tonight.  You can read the previous update here until I get the archive up and running (not a big priority with only one archived update).  Also, I don't intend to post the updates from before the hiatus, so if there's something you remember seeing that you want, let me know and I'll post a copy of the old stuff.  If anyone's looking for the suggested crossover page, it's here, but the Dayton 8 MTM crossover is here (I guess I didn't archive the suggested crossover page after I added that Dayton 8).