Austin DIY HT Projects - Page 9

Scott: careful some things are way too cool to be handled by the inexperienced, and could be VERY dangerous

So far it has been very fun and rewarding. It will be REALLY rewarding to have some nice speakers, but I place a good amount of value on the HUGE amount of knowledge I've gained so far. I've learned so much not only about speakers and how audio works, but also lots about woodworking and the fundamentals of how electronics/electricity. After all, the learning experience is really why I'm doing this in the first place. I'm looking forward to starting my XO design. I imagine that will be my favorite part of the whole project.

My wife is awesome. The Shiva sonosub really sold her on the whole HT idea, and DIY in particular. Now we're both just plain hooked. The first time we fired that sucker up, we knew we'd be taking the plunge with the rest of the audio setup. She even helped me build the flexi rack!

Ultra-Cool. It's a fake pack of cigarettes that I roll up in my sleeve so I can look, well, cool. No wait, it's that deep-heating muscle rub stuff. No, it's actually a grooming product for that perfect coiffure.

No, not really.

It's a quiet PC case cooling fan. I did some research, and it should be among the least noisy 80mm fans you can buy. And it was only $5! COOL!!! I'm going to build a hood contraption to sit on top of the receiver and suck the hot air out. The built-in fan in the receiver is LOUD and quite distracting when it kicks on. Not sure if the idea will even work.

See here...
http://www.hometheaterforum.com/htfo...hreadid=131944
...and here...
http://www.hometheaterforum.com/htfo...hreadid=132607

Wood dye? I was planning on thinning some oil based enamel to a wiping stain consistency or something. Is that the same thing? Anyway, I'm not going to finish the rack until I finish the speakers. That probably won't be for quite a while because I'm going to want a weekend or two _without_ a garage project. So bare+bondo will be the motif for the short term. Bondo, baby! Yeah! Fender-bender chic! Mark can have his velvet, you can have your leopard print, we'll stick with our BONDO.

By the way... thanks for the quick disconnects. I'll need to order more, so I'll be stuck with way too many in the long run. I'll give you plenty of my leftovers to restock your supply.

Methinks this boy is hooked. Jeff, we can talk stain, dye, etc next time we meet. Speaking of MEET - I wonder what month we should target for a meet...hmmmm.

May?

Hank -

Where did you get that "spray lacquer" you've used for some DIY experiments? I have some unfinished wood bar stools that could use that kind of a finish.

By the way, I think I still owe you money for food you bought on my behalf after the "Lawrence of Arabia" showing at the Paramount. I meant to mention this to you at the last HTF meet but got caught up in the thrill of four rooms jam-packed with HT equipment. Next time I make it to a meet or movie night that you're attending, please remind me...

Colin: I don't use spray lacquer. My finish of choice is the one I demoed at the big DIY meet at Johathan's: hand-rubbed Danish oil finish - you know, the labor intensive one. I would definitely NOT recommend that finish for any furniture that would be exposed to liquid spills. Oil finishes, while giving a very deep, rich, warm look that really brings out the grain, offer almost no moisture protection. For your bar stools, I strongly recommend polyerethane finish. Spray would be quick and easy, but subject to *runs* if you spray a bit too thick per coat. I'd use brush-on poly. You have your choice of gloss, semi-gloss or satin. Poly gives excellent moisture protection and is tough. For ultimate tough, you might consider Behlen's Rock Hard varnish. I built a baby highchair and finished it with that stuff and it has survived the world's most brutal abuse: 2 babies.

You owe me nothing. I enjoy your HT and lust after your Maggies. Okay, a high-end beer will do.

Jeff, don't keep us in suspense, how did the crossover modelling go for your speakers? Based on those tweeter measurements we took I'm anxious to find out if they sound as good as they measured.

It's coming along slowly. I seem to be able to either creat a flat frequency response or a nice phase null, but not both at the same time. I also find it odd that when I add my 4th order LR network to the tweeter, the impedance drops. When I optimize it, it drops even more.

This being my first project, I'm not exactly sure what to shoot for.

One question... here is an overlay of the MT and MTM woofer response measurements (blue line is MTM). Shouldn't I expect to see the MTM putting out 6 dB (or at least 3 dB) more than the MT across the board rather than just below 400 Hz?

Thanks!

When the impedance of the tweeter drops, what's the value of the first parallel inductor (usually put into net 2 I think)?


Mark

EDIT: I whacked the component values and graphics because my LspCAD skills are evolving and things are getting flatter and less expensive. I'll post more graphics when I _really_ think I have it right.

EDIT: I whacked the component values and graphics because my LspCAD skills are evolving and things are getting flatter and less expensive. I'll post more graphics when I _really_ think I have it right.

BUT...

I'm worried about the level of tweeter padding because I'm unsure of the level of the woofers (see MTM vs MT image). I remember you guys hitting the volume knob at some point due to the sound card beeping. If that was done after some other measurements were already saved, could it explain the different SPL calibrations I'm seeing in the FRD files? Or do I just not know how to import these suckers properly?

The level is the same because we didn't do an actual SPL calibration (that's why everything is referenced to 0dB instead of its 2.83V sensitivity). Don't worry about comparing the two woofer files. It only matters that a) the volume wasn't touched between the woofer and tweeter measurements on each speaker, and b) that you use the MT tweeter measurement with the MT woofer measurement and likewise for the MTM. If the volume control was fiddled with in between then there may be some problems (I can't remember if I touched it or not, I may have - sorry. This is why I have Scott do my measurements )

Tips:

I usually put the tweeter attenuation resistor last in the network. Try it in either place and see what works best for you.

The low tweeter impedance is caused by the parallel L that's only .1 to .2mH. Kick it up to at least .3mH.

Try not to use any inductor over 3mH.

Nix L3 in the woofer net and combine the parallel caps into one larger cap.


HTH,

Totally unscientific suggestion--go into Driver 1 or 2 settings (whichever you set for the woofer) and in the Scaling field under the driver frequency response file name, enter 6dB. Since we only took relative level measurements that may be causing the problem. Changing the volume on the test amp won't make a difference, LAUD automatically adjusts for that.

Also, I find it useful not whack crossover designs, I just flip to another memory setting (the R1/S1 buttons). That way you can compare, overlay, take snapshots, etc. of multiple options. Especially useful for emulating.

My latest design is crossed over at 3k. I'd prefer crossing at 2.5k (since if you're not crossing at 2.5k, you're doing it wrong), but I get a smoother response between 1k and 2k this way. So.... am I doing it wrong?

I'll post more screenshots and component values soon now.

There is a 2 dB discrepancy between the MT and MTM response below 300 Hz. If I go in and add 6dB to the MTM woofer response to get the expected results above 300 Hz, then what should I use for my baffle-step baseline?

Thanks for the help!

Given the extension of the Vifa woofer you're using, I'd say 3KHz is just fine, except that by pushing your crossover point higher you're making it even more critical to be on axis when listening to the MTMs. Just make sure you place those on little platforms that put the tweeter at ear level.

Grr, relative measurements. Wish I had been set up to do absolute measurements Saturday. This is where emulating comes in, I could tell you by listening if you're off.

I think your 2dB difference below 300Hz is a measurement artifact. You tend to get continuously downward sloping responses at the limits of the time window. If you look closely in Stereophile you'll notice that their measurements show the same effect, you can pick out the frequency at which the nearfield measurements were spliced in. I have yet to figure out why this happens.

If you're not comfortable ordering parts off of these results we can always re-measure with absolute values. If you can install a second set of binding posts in the cabinets (needed for an external crossover anyway) and wire up everything in advance, we can do a complete re-measure in about 30 minutes.

Here's my latest MT design which is back down at 2.5 again. Not as flat as my 3Khz one, but it's close enough. Jonathan: Thanks for the warning about the sweet spot height changing with frequency. Mark: putting the resistor first or last makes a big difference! Thanks!

All components (below) are values that can be purchased from PE. I'm looking at the 18 ga air core inductors, the Dayton metallized caps, and the Dayton non-inductive resistors.

Here's the response, the woofer network (baffle step), and the tweeter schematic. (10W series resistor is fine for tweeter attenuation, right?)



Here's the response with inverse tweeter polarity. I think this is a good phase null... as if stabbed by a steak knife. The schematic is the woofer network... 2nd order electrical, 4th order acoustic, right?



Does this look good? Am I on the right track? Would you make any changes to this, or should I move on to the MTM?


If the <300 Hz measurements cannot be relied upon, how will I pick my baffle-step baseline for either speaker? I know the curve is supposed to span 6 dB, but it's hard to find a reference point at the top of the lumpy woofer frequency response, too.

I guess the bigger concern is getting the tweeter level match right. With the current measurements for the TM, I only needed 3 dB of tweeter attenuation to bring the 91 dB tweeter down to the level of the 87 dB woofer with 4 dB of baffle step comp. That strikes me as not-quite-right...

If you think the situation warrants re-measuring or emulating before I order parts, that's cool with me. I had hoped to order parts in time to receive them by Friday, but there may still be time if I order them via overnight delivery. If you think the measurements we have are fine, I'm okay with that, too. I'll defer to your wisdom and experience.

What do you think?

EDIT: sounds like re-measuring it is. I'm installing the other pairs of highway-robbery binding posts tonight.

http://www.partsexpress.com/pe/showd...ID=14966&DID=7

I thought I was getting a good deal when they were on special for $8 a pop. :b

Jeff, I've got some similar (long post) to those and I hate to tell you , but they cost $1.25 each from Madisound:

YG-POSTB
YG-POSTR

NEW!
New from Yung. Gold plated 3-way posts w/ extra long 45 mm shaft, 5 mm diameter. 9mm knurled diamond-pattern head. 6mm cable entry hole. 4 mm banana plug hole. Red or Black.
$1.25

p.s. Resist the testosterone temptation to buy thier G25-cup 1" diamater posts that young Mark H. is addicted to.

Yeah, I got taken on those I guess. I should have looked around, but I saw they were on deal of the day (or some similar about-to-expire special) so I just went for it. I bought some less expensive binding posts for the shiva sub, and they were are a pain in the butt to install and are a pain in the butt to use. I wanted something easier and nicer this time around, so I just shrugged off the cost as a worthwhile convenience factor thing. If only I had looked at Madisound, I could have had my cake and eaten it too!

I'll certainly be ordering my speaker spikes from Madisound. I need 2 for the front feet of my center channel tower to tilt it back so that listeners' ears are on-axis. YBSC = $3.95 each. Pretty groovy.

I think the rest of my shopping list is priced pretty well at PE, though.

Hank: you'll be disappointed in me. I saw some of the PE cherry vinyl at Jonathan's, and I liked it. Looks like i'm going to wimp out after all.

VINYL??!!!

[rant]I'm disappointed in you and I'm reeling in shock that Jonathan has that vile stuff! You go to the expense and excruciating mental torture to learn crossover design, and you short-change the art of cabinetry finishing be using v____?? Tsk, tsk, tsk [/rant]

I bought the vinyl for Mark. You won't find that stuff on a Mirus speaker.

Jeff, you seemed satisfied that the 3M #90 spray telegraphing was minimal enough to be a non-issue. If that's the case, then I'd say go with the veneer. It's more expensive, but the 3M spray eliminates the hardest part of veneering, which is adhering it to the cabinet. The second hardest part, getting a good edge, can be alleviated by purchasing a flush-trim spriral-DOWNcut bit for your router. I use a razor blade to trim the edges, but it took me quite a few cabinets to get good at it. The downcut bits a supposed to be quite easy to use.

Whoa... Hank whipped out the boldface italic text. He must REALLY be disappointed.

Mental torture? I actually enjoy the crossover design part. It's really the reason why I did this project in the first place. Otherwise, I would have just gone with someone elses proven design. Torture is trying to make nice boxes with a circular saw and a straightedge. Don't get me wrong... I've enjoyed the woodworking aspect of this project, but it's obvious that my skills fall very short of yours. I doubt you use nearly as much bondo and wood filler, for example.

I imagine that my first attempt at laying down vinyl will look at least as good as my first attempt at veneering+staining.

Besides, I thought Audiophiles are supposed to like vinyl.

Don't worry, I'll veneer something someday. I'll either peel this stuff off sometime, or I'll build more speakers someday.

I couldn't detect the telegraphing until I ran my fingers across it. As long as it didn't block the absorption of the stain and make a blotchy visual effect, it'd be fine.

My overall observation is that people talk about veneering as an art. I believe that. I believe that my first attempt will leave much to be desired. Putting a paintbrush in my hand won't turn me into Claude Monet, and putting a can of 3M 90 in my hand won't turn me into Hank Frankenberg.

In other words, I think I would be doing an injustice to the art of cabinetry finishing by doing a lousy veneering job. Too much work went into these enclosures for me to use them as veneering practice. I'll practice on something else first.

My first veneer job was NOT artistic. As long as you buy 3M spray contact cement, I'll be a happy guy. Veneer doesn't have to be smooth as laminate. After all, it's wood, and wood isn't perfect. And you *will* build more speakers. After all, you're...HOOKED! BWHAHAHAHAHAHAHA!!

Hank: You're right... I'm hooked! I'll definitely build some nicely veneered speakers someday.

I have another hobby that is VERY 3M-intensive. It requires enough 3M Super77 adhesive spray and 3M fiber-reinforced strapping tape that you should be a happy guy even if I never veneer anything (but I will!).

I'm finding it very difficult to stop my phase line from dipping below -60 in my input impedance window. Grrr.

Edit: nevermind... Zobel helped this.

My network for the MTM is pretty freakin' expensive at 11 components, but it looks pretty flat. Tweeter looks a bit recessed, but I'd rather have that than 'bright'. I can adjust this to taste later by changing the two resistor values (one in front and one behind).

The phase null is off the charts, the phase line in the input impedance window pretty much stays withing +/- 30, and I "did it right" because it's X'ed over at 2.5k on the nose.

Any suggestions, comments, or observations? I'll be pulling the trigger at 2:30 PM.




Net 1 is the woofer (with Zobel), Net 2 is the tweeter.




Now onto the MT...

I don't know if the standard version of LspCAD can do this or not, but in Pro you can save a transfer function and then optimize to that curve using a new set of components. In other words, you can reduce the number of components in your crossover and get the same results. Sometimes it is successful, sometimes not.

Your results look good, if the xo isn't too expensive then I wouldn't hesitate to pull the trigger on that design.

Here is my MT crossover. My only concern is that to keep the phase line within +/- 45 degrees, I had to use a lot of series resistance which brings the tweeter impedance way up. I think this is okay, though. Please correct me if I'm wrong!





Look good? I'm going to order these parts any moment now.

Looks like my MTM crossover is $37.93 and my MT crossover is $28.26 at PE retail prices. Not sure how this compares with other DIY designs. I was expecting to pay about $30 per speaker for XO parts, so the MTMs are a bit spendy I guess.

My standard LspCAD doesn't seem to have the functionality you're talking about (try different components to end up at the same result), unless I'm just not looking in the right place.

The use of resistors both before and after your other components is something I don't normally do, but I don't see a problem with doing it that way.