SEOS-12 Fusion Max Build (Delta-10A, DNA-350) - Page 2

Bondo Stinko

Even though it's not terribly cold out right now (for this time of year), I thought I would Bondo in the storage room / cat box room in the basement instead of trying to warm the garage up. 

Bad idea. Wow, that stuff stinks! 

It's when the hardener is mixed in and the reaction takes place that there's some strong off-gassing. 

So, I still have to open up windows and the doorway to the garage to exchange air. And I'm banned from using chemicals in the house. 

Too bad it's an old, leaky garage, and not a nice new one with in-floor heating like my coworker just built!

Hey, that gives me an idea that maybe I can rent it for a couple nights for the painting? I don't think I'll be able to get mine up to 70 degrees. What do you guys think about 50 degrees?

Paint Delay

I think I'll have to wait until Spring to paint.

Bah, my neighbor's garage in-floor heating is actually not installed yet, so that's a no-go. 

While the outside will warm to the 30s this Sunday or Monday, and thus I think I can get the garage to the low 60s -- I don't think I can keep it that warm long enough to do several coats of paint I'm reading I'll need to apply (with sandings in between). Sounds like enamel takes a long time to dry between applications to cure, and while I read some people move their speakers indoors to cure, my wife will not accept the off-gassing for several days.

So I think I'm going to have to stop at the Bondo (or maybe the primer) phase and wait until it warms up in April. 

Quote:

Originally Posted by btinindy 


Well they will still sound fine naked. Also, if you use Duratex you can paint it indoors and the odor is non-existent. You also do not need to prime it. Just a thought.

Hmm, I liked the smoother finish that the flat black paint gives, but Duratex is looking more attractive. I didn't know it was non-odor.

But, for now, I chose to go naked. I worked on crossovers and finishing them up this past weekend.

Crossover

Wow, these are are uglier than Java's, let alone MTG's. I'm beginning to wish I had someone do them for me. 

I'd have to be really efficient with use of space to fit the LF and HF on one board, so I copied Java's crossovers.

I didn't have terminal barriers, so I bought some at Radio Shack. Where, naturally, someone had JUST bought up most of the two-position ones. I had to buy a 4 position and 8 position. 

Internal wiring was expensive at RS, and I couldn't find my roll of Romex, so I just used speaker wire. Later, I regret this, because I fear I reversed polarity somewhere on one of the speakers. 

I also burned a pinky, because I'm clumsy, and I cheaped-out and didn't get a soldering iron with a nice holder/guard dealio.

I did the HF on Friday and the LF on Saturday. 

Foam, Polyfill, and ErichFoam

My grandmother passed away on Friday, and I think I was driven to push through all Saturday and finish the speakers enough to listen to them. 

I live a couple blocks from JoAnne's, and WalMart is hellish to go to lately (very crowded since the oil boom). So I hoped I could find some mattress-topper type foam there and some polyfill.

There was a 50% sale on some foams and fills. I ended up with a 22" square of 2" foam and a 20oz bag of polyfill.

I used the thick foam on the back panels, and friction fit it elsewhere where I could. 

I had joked earlier about the great half-inch thick gray foam pieces that Erich packed the boxes with, how I hoped that wasn't a part I should use, because the cats loved scratching on it. Well, I actually ended up using it inside the speaker for certain exposed panels. I used spray adhesive.

Since I had a bunch of foam inside, I figured that was enough filling that I didn't need another 16oz of polyfill for each speaker, so I used about 10oz for each cabinet.

Driver Wiring and Install

Some simple things stymied me. Like how to attach the driver to the wire. The DNA-350 was simple, but I'd never wired a woofer before. This had little tabby-clips. I wasn't sure if Java had soldered or what connector he used.

After some research I saw the type of connector that slides over it, and the connector is crimped to the end of wire. "I think I might have those!" and I did. Five of them, and I needed 4. I crimped one, but in the wrong location, and it fell apart. I watched a video, and crimped another. But the presenter said to crimp the metal a little tighter so it doesn't fit too loosely. I did so, but it made it too tight to go on. Had to do lots of work on one little end for an hour. Here's a picture from elsewhere of the connectors:

I didn't use T-nuts or hurricanes...just used the wood screws Erich provided. Which was probably a mistake, because I think I need to open one up and perform surgery. How well will they hold in the MDF when I screw them in a second time? With the woofer, I can rotate and drill new ones if need be.

Connecting the Crossovers to Inputs

Oh god, the horror of my wiring. I soldered my leads from the crossovers to the binding posts. I unscrewed the posts and took the piece you connect the wire to and soldered that to the wires I made up that send the + and - each to the LF and HF crossover boards. (here is maybe where I screwed one up?)

It took forever to solder these: the binding post metal took longer to heat up and the wire would get too hot to hold so I kept moving my hand father away. Pliers and my wife helped get it done, but it was ugly. One solder joint I tried to take apart, and failed, so I gooped more solder on and called it good.

Listening...aw crap

I finished around 3am Saturday. Too late for a movie or measurements, but we watched one show on the DVR. That, and my playing some movie tracks on Sunday, it sounded kind of "phasey," like when the DVE disk or THX Optimizer plays out of phase signals and it's harder to locate sounds. 

So, imaging went downhill. However, just like everyone says, I found the volume going up way higher than before. Like -5dBFS. 

To fix/diagnose this, I thought I may have reversed polarity somewhere. Wiring up so late Saturday night / Sunday morning probably made me more error-prone than usual!

To test/fix this, during the SuperBowl, I swapped the right speaker's inputs from receiver. This seemed to clean it up. Sounds seem to lock on the screen better.

They sound pretty good, but I think I'm missing mid-bass.

Measurements with REW make me think that maybe it's not as simple as that. I suspect just the tweeter or woofer is wired wrong on one of the interconnects? 

I'm thinking on testing this evening by sending 300hz to both speakers, measure dB, then reverse polarity on inputs. Then also do it with a 3kHz signal.

Any other thoughts as to diagnosing this kind of screwup (which speaker, LF or HF?) before I open them up? Of course I feel so stupid and sick, wanting to bask  in the SEOS greatness but having mucked it up. 

Quote:

Originally Posted by Java 


It's worth repeating for Eyleron that you need to reverse polarity of the tweeter for this build. Not all are like that, but for this design you need to hook them up reversed. Woofer is a straight connection.

Eyleron,

Shoot a picture of the board you put together. You never know... I messed up my crossover build for my mains and it was an easy fix.

Ha, yes, definitely worth repeating. Tux did write that in the schematic. I saw that note, and thought it was a note to himself or something. 

So, I definitely reversed the compression driver connections.

Sorry, I thought I'd added the photos of the LF and HF crossovers of one speaker. Here goes...

Quote:

Originally Posted by tuxedocivic 

Do this. Find a batter and clip positive to positive and negative to negative on the speaker terminals. The woofer should move out. Flip the polarity and the woofer should move in. If that happens in the opposite direction, your woofer is out of polarity.

Battery Woofer Polarity Test

I did the battery test. The woofer moved out when connected positive-to-positive on both speakers.

Quote:

Originally Posted by tuxedocivic 
This is a reverse tweeter polarity speaker. Use a SPL meter or microphone and REW or some signal generator. Put the mic about 1m from the tweeter on tweeter axis. Turn on the signal generator to about 1200hz. Take a reading. Now try 5000hz and 800hz. All three SPL readings should be within, oh lets say 5db. That's a wide range, but you're in a room so who knows. If the 1200hz reading is something like 10db less than the other 2 readings, then your tweeter is out of phase with the woofer. As in, it's wired straight polarity and should be reverse polarity. Confusing, I know.

If you do those two things, we'll figure it out. The battery test is the first and easier thing to try. I'm going guess though that its the tweeter. If you have to start pulling things, take pictures of the boards so I can make sure there's no wiring mistakes. It might not be polarity. Could be something else.

Tweeter Out of Phase with Woofer Test

I tested, and although it wasn't down 10dB at crossover from 800hz, it was down 10db from 500hz.

  right speaker

left speaker

1. Didn't reverse the tweeter polarity as I should have.
2. Accidentally reversed polarity on only one speaker (which is good, in that I accidentally followed Tux's instructions. Bad in that I wasn't following my own.)

See, what bothers me, is that if I failed to reverse the tweeters, why did the imaging sound better when I swapped polarity at the speaker level yesterday? Placebo? 

I'll check that both are reversed as per design.

Quote:

Originally Posted by Java 

Last, in addition to the REW checks Tux suggests, did you check the impedance? It should be around 6.4 ohms. I'm a noob, but maybe the would point to some bad connection. On that, I had a bad joint in my surround build. It surprised me since I wrapped the wires before soldering (solder was more glue than connector), but there it was.

I tried to check impedance between components and different joints like you do in your video. I would get weird readings most times, like it was too low to report. I got discouraged. If I want to measure the whole speaker, can I do this now at the binding posts?

Quote:

Originally Posted by antisuck 


That actually makes sense if in fact one of your tweeters is wired backwards, as a lot of imaging cues reside in the upper frequencies. When you reversed one set of speaker leads you brought the tweeters back into correct phase - but knocked the woofers out of phase at the same time, hence the big drop in midbass you noticed.

Spot on! I opened both of them up by removing the horn.

The left speaker: wired backwards, as directed, but not as I intended in my fatigued state. 

Right speaker: wired positive to positive, not reversed, not as directed.

Quote:

Originally Posted by tuxedocivic 

Well, your cross over looks wired right. Your woofers are in the correct polarity. But the SPL test isn't as conclusive as I hoped. I would think they're both wired the same: wrong. But I'm not 100% sure.

I should start by asking, when you did the REW thing, was audessay and all that jazz turned off? I helped a guy recently who was getting a weird measurement, and that was the problem in the end.

And one last thing, what was the distance from the mic to the tweeter/baffle. If to close, it'll cause a null at 1200hz.

Oh, to be so triumphant at first. The theory was proven out: only one tweeter was correct. I fixed the "wrong" tweeter in right speaker. Sure that the numbers will come out closer to the left speaker. ....  wah?

It's worse?!   Despondant.

Quote:

Originally Posted by Eyleron 

That's a good point about the Room EQ thing. I'll turn MCACC off and run again. 

Crossover Test: MCACC Off

Well, the numbers changed a bit, but not so much at crossover.

Can anyone recommend good music and films that don't have any content between 800hz and 2,000hz?  I'll need to avoid an octave or so.  

Does this mean that if the woofers are correct polarity, and the tweeters are now correct polarity, and I have now an even more marked dip at crossover, it's next the wiring or the crossovers themselves of the right speaker?

Any other low-hanging-fruit to start with?

Quote:

Originally Posted by tuxedocivic 

The woofers are right we know that from the battery test.

The lead coming off the last capacitor on the HF board goes into the negative terminal of the tweeter? And the positive tweeter terminal goes into the ground/negative of the board?

Have you tried listening since the switch? If the phasy mess is gone but there's a hole in the response around the cross over, they maybe both out of phase. If that makes sense.

I knew reversing the tweet polarity was going to be a pain. Sorry it's such a pain.

I haven't checked the wiring inside the right speaker aside from the tweeter polarity. I guess that's next.

No no no, it's not a pain, don't be sorry. Reversing the polarity wasn't a pain...that was easy! In diagnostics we tend to look/hope for one cause.  It's finding that there's more than one thing wrong with the patient that's the pain-- but that's not your fault.

I did an impedance check on the speakers' binding posts. 5.8ohm on right speaker, 8.3ohm on left. Not sure what one should expect measuring that way?

Tomorrow I'll tear apart the right speaker and see what I can find.

I sure wish I used red and black wiring like Java did, instead of speaker wire where the negative has a faint white stripe running down it and silver metal. It's hard to discern in low light and probably too easy for me to mix up.

Thanks for all your help guys!