The Center of the Sun - photos (1 Viewer)

[Hank Frankenberg]

Patrick, another fine, well-documented job! You'll enjoy the sound; W - T/M - W is the way to go for a center channel. Will you veneer or paint the cabinet?

 

[Greg Monfort]

>I can't really explain the dip near 100Hz.
====
My guess is a 1/4WL suckout due to the proximity of the hall opening.
GM
------------------
Loud is beautiful, if it's clean

 

[Patrick Sun]

Hank, well, I'm going to look for rosewood veneer (to match the main speakers), but I will experiment on pieces of MDF to make sure I know what I'm doing. I'd hate to ruin the construction late in the game. Plus, I don't look forward to cutting in all the driver holes, and the screw holes. I'm sure I won't paint it.
Greg, Hmm...that's a good possibility. Maybe a folding door in the hall? I'll do a 1m measurement to see if it's still present. What do you think of its frequency response as it is now? Too much on the high end?
------------------
PatCave ; HT Pix ; Gear ; Sunosub I + III ; DVDs ; Link Removed

 

[Greg Monfort]

Forget 1m measurements to find driver FR anomolies in-room. Stick the mike

 

[Andres Munoz]

Now that's what I call intimidating!
Came out very nice Pat. You da man!
How heavy is it again?
------------------
Andres
http://home.earthlink.net/~coolvirus/

 

[Patrick Sun]

Andres, it's one heavy mo-fo!
Greg, interesting, if I have some free time, I try those 2 measurements. I plan to do some work on flattening out the FR:
Just to clarify the graphs, here's the graph with measurements taken at 1m .
Here's the previous night's graph with measurements taken at 2m .
They basically show the same thing.
Believe it or not, but that tweeter is really attenuated already about -4dB (down 60% from non-L-pad value), so what I'm going to do is take out midrange L-pad to boost the midrange by around 3dB. In the original crossover plans from 10 years ago, there was no L-Pad for the midrange crossover. So I'm going backwards here.
I did a quick measure of one of my main speaker, and here's the graph of the main speaker at 1m . Sonuvagun, it looks sort of like the midrange trough from the center channel speaker measurements. The woofer is definitely on par with the tweeter output though. I had always thought the male vocals were a little thin on my speakers, now I know why. Looks like time to overhaul the main crossovers as well and rip out the L-pad on the midrange crossover. And since I have the new caps from Madisound, I can put those into the networks.
I'll do some more soldering tomorrow, and connect the woofers in parallel, put in the other woofer crossover components (I'll be using a 2mH inductor and a 80uF cap, so I can get rid of the DCR from the 8mH induction from the series woofer crossover). I'm hoping the parallelled woofers will boost that low end enough to be on the same sensitivity as the tweeter/midrange. We'll see. I may be brave and take out the midrange l-pad at the same time, then take my measurements and report back.
------------------
PatCave ; HT Pix ; Gear ; Sunosub I + III ; DVDs ; Link Removed

 

[TerryC]

Pat cool stuff. Gotta love those Digital Camera's!
On a side not, I notice you have your own webpage! How does one go about doing that and does it cost much per month/year? Any suggestions for someone that knows nothing about it? if the response is too long feel free to email me if you prefer.
Thanks
Terry

 

[Jay W]

Excellent work P|@, I finally went through the photos today (phew).

 

[Patrick Sun]

TerryC, well, I did register a Patcave domain name last year, but I still haven't decided on a webhost who'll give me a ton of webspace, so I'm "parking" my Patcave website on some place called DomainValet temporarily, and they give me 20MB of webspace which I'm using for the CC construction photos since Geocities has taken a dump. I need lots of webspace for my DIY photos, and my convention photos. I'm trying to find a webhost that'll give me 400-500MB of webspace for $10-$15/month if possible.
Anyhow, if you have an idea for a website URL name, register it for around $10-$35/year depending on who you use (like Link Removed
------------------
PatCave ; HT Pix ; Gear ; Sunosub I + III ; DVDs ; Link Removed

 

[Patrick Sun]

Okay, I made some changes to the L-Pads of my midrange and tweeter crossovers, but something happened to my low end response. If anyone can offer some ideas, I'd be grateful. You can refer to an earlier graphs which shows the low end response of the 2 woofer hooked up in series with one another.
Now, I've hooked up the woofers in parallel, which should give my a 6dB bump, but instead, it depressed the hell out of the low end response under 112Hz. Check out these 2 graphs:
Graph A : Woofers in parallel, Bypass Cap paralleled with the tweeter series L-Pad resistor.
I included this 2uF Bypass Cap to smooth out a 12KHz hump, but it just wrecked my tweeter response, especially with that dip around 4KHz, so I took it out. Here's the result:
Graph B : Woofers in parallel, NO Bypass Cap paralleled with the tweeter series L-Pad resistor.
To me, it looks better on the mid-to-top end (anything after 12KHz is pretty hurting), staying relatively flat witin +/- 3dB.
So my lowering of the L-Pad for the mirange raised its output by 1.5dB, and my raising of the tweeter L-Pad dropped it by 2dB for a more smoother response from 112Hz-12Khz.
But what has happened to my low end response with the woofer parallelled? My new woofer crossover is a 2mH inductor and a 80uF capacitor, crossover frequency is about 400Hz using a Butterworth. I used 3.5 ohms for the nominal impedance of the parallelled woofers. Should I have used 4 ohms? Either way, it doesn't really explain what happened. I'm temped to go back to my woofers in series setup.
Any ideas?
------------------
PatCave ; HT Pix ; Gear ; Sunosub I + III ; DVDs ; Link Removed

 

[Hank Frankenberg]

Patrick, I'm no crossover designer (for that, you need the best programs and a lot of experience), but I have never seen a single speaker crossover requiring an 80mF capacitor. Or did you mean to type 8.0? If it is 80, I'm surprised. Just an observation; I dont have an answer. Hang in there, you'll persist and figure it out.

 

[Patrick Sun]

Hank, this is what happens when you parallel two 8 ohm woofer drivers (to get an impedance near 4 ohms) and design the break frequency at 400Hz. Yes, 80uF is a huge capacitor!
I did check to see if I had wired one of the woofer drivers out of phase, so I reversed the leads on one of the woofers, but I got a steady SPL of 52-57dB for most of the low end from 45-100Hz, so I know I didn't have the woofers connected out of phase the 1st time. So I was able to rule that out.
------------------
PatCave ; HT Pix ; Gear ; Sunosub I + III ; DVDs ; Link Removed

 

[Eric M Jones]

TerryC,
If all you're looking for is the capability of having a unique URL to get to your site, then check out www.directnic.com.
They charge $15 a year for a URL, say (www.terryc.com) Since you already have a webpage you'll be good to go because all you have to do is go to directnic and put that address (http://terryctheater.tripod.com/shiv...bum/index.html) in as a redirect and everytime someone hits (www.terryc.com) it will seemlessly redirect them to your site.
I've used directnic for over a year with no problems. If you like to see how it works, my page address is www.jones4.com
-EJ

 

[Patrick Sun]

Okay, y'all can call me a doofus head. Last night's graph with all the bass missing was due to one thing: the Center Channel speaker was set to SMALL! That's why there was such lower output in the bass region.
Here's 2 graphs:
Graph C : Woofers are connected in series, and I took measurements for both amp sections (the HK PA5800, and my Sony DB930 receiver). The HK seems to be a little stouter and smoother. I took the opportunity to set the woofer crossover frequency at 500Hz by lowering the inductance from 8mH to 5mH. That got rid of a big dip around 400Hz on previous woofer crossovers attempts.
Graph D : Woofers are connected in parallel. Again, I'm showing you the frequency response with the 2 amp sections. The bass sort of overwhelms the midrange/tweeter due to the gain in effiency of the 2 parallelled woofers. Again, you can see that the HK can provide more juice to the speaker, almost a 3dB difference across the board compared to the Sony, which is probably only putting out enough to handle an 8ohm load, though the parallelled woofers present a 3.5-4 ohm load to the amp.
I'm not too crazy about the dip/hump/dip around 2KHz to 4KHz. Weird. I was wondering if I mess with the midrange's low pass 1st order inductor crossover (3600Hz) and extending it a little bit more, to around 4200Hz, using a 0.58mH inductor (current value is 0.68mH).
------------------
PatCave ; HT Pix ; Gear ; Sunosub I + III ; DVDs ; Link Removed

 

[Hank Frankenberg]

DoofusPat, it's okay, in the world of D-I-Y, we make embarrasing mistakes, like a few days ago when I installed a Shiva in my buddy's cabinet. He had just completed the cabinet and was in a state of wide-eyed anticipation with a room full of observers and there was almost no output. Guess who had put the amp wires on the wrong terminals?

 

[Patrick Sun]

Just be glad I don't document all my doofus moments when I take my photos.
Which graph's response would you prefer? Most bass (paralleled woofers) or a tad more on the high end (woofer in series)? Once I run the center as a small, the bass doesn't really matter to me as much, but I just wanted to make sure my woofers were putting out some semblance of bass.
------------------
PatCave ; HT Pix ; Gear ; Sunosub I + III ; DVDs ; Link Removed

 

[Patrick Sun]

I have decided to dub this center channel speaker project: The Money Pit!
I got my bonus yesterday, so I have some spare cash to throw at this project. So I decided to upgrade that current tweeter because it really does have a poor response up high, so I decided to go with a Vifa D27TG45-06.
I did a few more calculations for the initial new crossover, go the component values (plus some extras to move the break frequencies around). A quick visit to the audio shop, and I have more pieces for more audio engineering fun.
------------------
PatCave ; HT Pix ; Gear ; Sunosub I + III ; DVDs ; Link Removed

 

[Greg Monfort]

>Okay, y'all can call me a doofus head. Last night's graph with all the bass missing was due to one thing: the Center Channel speaker was set to SMALL! That's why there was such lower output in the bass region.
====
It never occurred to me that this would be changed once set. I have seen amps that have severe current limiting ckts if a low impedance is presented to them though.
I assume you're going to stick with the series wired drivers?
I'm surprised you hadn't measured your mains before trying to duplicate them as a CC. Looks like you need new tweeters and a modded mid section on all three.
BTW, are all these measurements at the same voltage level?
GM
------------------
Loud is beautiful, if it's clean

 

[Patrick Sun]

Greg, I'm wondering is my RS SPL meter just isn't up to snuff in the 13KHz and up region. I've tried 2 different tweeters now (Philips and Vifa), and they both show the same dropoff curve for those 3 datapoints (12.9Khz, 16.25KHz, 20.48KHz)
I will do a better job of selecting on volume level and sticking to it (that's 2 on my volume dial). This past week, I admit to being a little haphazard in my volume level (it's been close to the same, but not always the same, maybe +/- 3dB).
I have this hump in the 5Khz to 10KHz region, so I'm going to modify the midrange's 1st order lowpass filter into a 2nd order lowpass filter to get more break in it at a slightly higher break frequency (1st order break was around 3500Hz, but I think I'll break the 2nd order at 4300, this should smooth out a 6dB hump in the 5KHz-10KHz range. This just means putting a 2uF cap in parallel with my inductor. I may have to change the polarity of my midrange when I do this change.
------------------
PatCave ; HT Pix ; Gear ; Sunosub I + III ; DVDs ; Link Removed

 

[Greg Monfort]

Pat, I assumed (my bad) that your graphs were done with a computer program. If you're using an RS SPL meter, then you need to factor in the error corrections. They have a rising response in the 3-8khz BW, dropping like a stone from there. I have a first generation unit, and it is -20dB/20kHz/"C" weighting. As supplied, the RS meter is designed for critical voicing measurements, so is only ~flat in a narrow BW. I use a B&K for wide BW measuring.
This explains my head scratching (that I've been keeping to myself) over the FR, considering the drivers involved and their published specs.
Before you waste any more time with XO's, mod your meter, as your speakers aren't as bad as you think. Link Removed
GM
------------------
Loud is beautiful, if it's clean