or Connect
AVS › AVS Forum › Audio › DIY Speakers and Subs › Testing Cheap Thrills with REW
New Posts  All Forums:Forum Nav:

Testing Cheap Thrills with REW - Page 5

post #121 of 163
Thread Starter 
Quote:
Originally Posted by tuxedocivic View Post

Could it be that Bill modeled and tuned the speaker with the DNA 360 but spec'ed the 350? They're similar, but not identical.

Yes, that is what the thread at diysoundgroup is titled, "Celestion 15" buyout and SEOS12/DNA360". When I look at Bill's response, his doesn't roll off anywhere near as fast as mine. He is only down 2.5dB at 10k from 1k. I think that the graphs I posted last may have had "Music enhance" turned on in my Onkyo 818. I've measured it with and without that setting and that really brings the upper range up, it also brings the lower notes up a tad, too.

Gear mentioned in this thread:

post #122 of 163
Thread Starter 
Quote:
Originally Posted by Willie View Post

I am not at all unhappy with the sound of the crossover design. The sound of my CTs can be characterized as 'laid back'. I love the sound. Of the speakers I have owned these replicate the sound of an acoustic guitar or violin in a live performance the best yet; Les Pauls and Strats sound pretty nice also. Underscores that not every voicing is for everyone.
According to this picture, the main frequencies of an acoustic guitar and violin fall below the area that I am boosting with the polarity change and the resistor mod. The harmonics of those instruments extend higher, though.


If you look at Bill's measurement, he is only down 2-3dB from 1k to 10k, like I mentioned above. Look at my measurements of my speaker, the mic and speaker were not moved. This graph is fully smoothed, allowing me to see the overall trend because I measured indoors. I bypassed the 6.2 ohm resistor. With the resistor bypassed, I am down around 4dB, without the modification I am down 7dB. Maybe my speakers are funny, but I don't think so.
post #123 of 163
Could be a mic calibration thing too. Dayton mics tend to measure more high end where as Cross Spectrum labs measure less, in my experience.
post #124 of 163
Thread Starter 
Good point. I am using the UMIK-1 but I also have a Cross Spectrum calibrated mic. I will make it a point to compare this weekend.
post #125 of 163
Quote:
Originally Posted by NathanJ View Post

According to this picture, the main frequencies of an acoustic guitar and violin fall below the area that I am boosting with the polarity change and the resistor mod. The harmonics of those instruments extend higher, though.


Great chart. Thanks for posting it.

I took delivery on a UMIK-1 earlier this week, however I am traveling for the better part of the next three weeks. As soon as my schedule allows I will run some tests and post the results. I am also considering picking up a DNA-360 and doing a comparison.
post #126 of 163
Thread Starter 
I finally got around to testing out my Cross Spectrum Labs EMM6 on a Tascam 122 MKII against my MiniDSP UMIK-1. I performed the loop back calibration on the Tascam and then used my CSL 0 degree calibration.

The UMIK-1 shows about 3dB higher at 10k. This means that using the EMM6 the CT drops a little harder in the upper octaves.

The graph below shows 2 mics, the very top line is the UMIK-1 with the resistor at 0 ohms
The 2nd line is the EMM6 at 0 ohm and can be directly compared to line 1.
3rd and 4th line show 2 ohm and 6 ohm.

These tests are inside my basement, not outdoor.

post #127 of 163
There ya go. The cross spectrum measures lower. That's been my experience as well. Whether is cross spectrum, or everyone else, I dunno. But I'm not sure how Dayton and Minidsp get their calibrations, but I do know how Cross Spectrum does. So...
post #128 of 163
Erich and I pulled out one of these today and tested it using one of the crossovers I had built before the woofers were gone. Here is that measurement with approximately the same scaling and smoothing applied for comparison purposes.

post #129 of 163
What polarity and vertical angle is that Matt?
post #130 of 163
Polarity is correct on the Cheap Thrills speaker. The design Bwaslo posted for this speaker is right.

Changes to the frequency response would be if you have them higher up on the wall, angled down or up too much, or possibly different mic distances during tests.

Also, I don't think you need to change the resistors for the high frequencies. I'd check the crossover assembly first.
Edited by Erich H - 10/27/13 at 5:36pm
post #131 of 163
That was how Bill specified it should be wired and it was slightly above axis. If you go below axis and flip the polarity it will measure nearly flat again. On axis both polarities measure quite similar.
Edited by mtg90 - 10/27/13 at 5:56pm
post #132 of 163
Ya I think Nathan is listening below axis.
post #133 of 163
Quote:
Originally Posted by Erich H View Post

Polarity is correct on the Cheap Thrills speaker. The design Bwaslo posted for this speaker is right.

Changes to the frequency response would be if you have them higher up on the wall, angled down or up too much, or possibly different mic distances during tests.

Also, I don't think you need to change the resistors for the high frequencies. I'd check the crossover assembly first.

Thanks go to Erich and Matt for the efforts of re-testing to confirm this, and explaining the above/below axis characteristics. Since most of my listening is positioned slightly below axis, I may opt for reverse polarity when I build my pair of CT's.

Are the Karma 12's going to be tested as well?

Howard
post #134 of 163
Thread Starter 
Thanks for measuring that speaker Matt. It is obvious that I have something wrong, I'm just not sure what it is. I noticed when looking mhutchins crossover that one or 2 of the coils looked different than mine, I figured that was standand practice with Parts Express and that they may have a few components that are functionally identical, but physically a little different. If that isn't the case, maybe that is my issue.

mhutchins crossover


my crossover
post #135 of 163
It's rare for someone to listen below axis. I favor above axis.
post #136 of 163
Quote:
Originally Posted by tuxedocivic View Post

It's rare for someone to listen below axis. I favor above axis.

Duly noted. I am probably one of the rarities. 99% of the time I listen near-field, with my ears even with the bottom of the woofers (Karma 12 cabinets about 43" above the floor on pro stands). I do occasionally set them up elsewhere, where other variables come into play, but always on pro stands.
Edited by zhillsguy - 10/27/13 at 8:20pm
post #137 of 163
In my shop I don't really have much choice as if they aren't up high then they'll constantly be blocked which is why I also like the resistor jumped as I think I'm never anywhere close to the sweet spot.(80'x45') building.

Would all the different waveguide speakers show this same characteristic with polarity?
post #138 of 163
Not just the waveguide speakers, but pretty much all multi way speakers in the world. The only ones that wouldn't are coaxials, and very low cross overs. Maybe some other things I'm not thinking of, like synergy horns or something. The lower the cross over and the closer the drivers are spaced together, the less flopping the nulls will do. I've got a little TM that doesn't null until 50 degrees below axis. But the large speakers, like this, will flip in about 20 or 30 degrees easily.
post #139 of 163
Quote:
Originally Posted by Erich H View Post

Polarity is correct on the Cheap Thrills speaker. The design Bwaslo posted for this speaker is right.

Changes to the frequency response would be if you have them higher up on the wall, angled down or up too much, or possibly different mic distances during tests.

Also, I don't think you need to change the resistors for the high frequencies. I'd check the crossover assembly first.

I agree with Erich, from my perspective as a novice DIY builder of the Cheap Thrill.

I don't have a box yet, so I have done extensive listening with multiple variables with drivers and XO's, just because I am curious about how things work. I have listened to the components without the crossover, and with crossovers (I split my crossover into HF and LF). I also screwed up building my XO's so now I know what it sounds like when it is done wrong! Mine sounded like a cheap AM radio before Tux put me on the right path. I have also compared the components by mixing and matching with the CD between the 360 and another pair of 205's.

I am telling you, that 360 is PLENTY bright with the standard XO (after I wired it correctly rolleyes.gif. Trumpets SCREAM on the high end, cymbals and drums CUT and are CLEAR, and there is no roll off on the top end that I can hear.

The way you are describing what you are hearing, sounds to me like the XO. I suggest you have someone other than you put some eyeballs on it, because you can't see your own mistakes. Or, speaking for myself, I couldn't. rolleyes.gif
Edited by wvu80 - 10/27/13 at 10:02pm
post #140 of 163
Quote:
Originally Posted by NathanJ View Post

I noticed when looking mhutchins crossover that one or 2 of the coils looked different than mine, I figured that was standand practice with Parts Express and that they may have a few components that are functionally identical, but physically a little different. If that isn't the case, maybe that is my issue.

Hi Nathan,

When I bought my crossover components, PE was backordered on some of the inductors specified, so I bought a different brand with the same values. Also, I am the first to admit my wiring design is not straight forward. I was trying to make a design where none of the wires crossed...I was just pretending I was making a circuit board design... Nerds, what are you gonna do? eek.gif

Mike
post #141 of 163
Thread Starter 
I think I will probably try remaking the crossover from the schematic, not copying anyone else's layout. Is there consensus that inductors don't have polarity?
post #142 of 163
Inductors don't have polarity, but they are sensitive to orientation and proximity when two or more are nearby. The farther you space them apart, the better. You also want to orient them 90 degrees to each other to minimize field interactions.
post #143 of 163
Thread Starter 
Thank you. I did already know that, but I just wanted to make sure I don't do something stupid before I remake the crossover. I have traced through the crossover again and I do believe that it is correct, but I will still make a new one just to rule it out. I am probably going to go so far as to make the crossover physically laid out like the schematic to make diagnostic easier.

Erich mentioned that measuring the speaker at a different position can yield different measurements. That is obviously true, however I am not overly concerned about how the speaker measures at 30" out and 6" up. I care how the speaker measures at distances that more closely approximate my final listening position.
post #144 of 163
^ ^ ^
Agreed, but you may need to measure outdoors to eliminate all the room effects.
post #145 of 163
Thread Starter 
I finally figured out the problem, I had the 2 similar inductors switched. I also decided to layout my own crossover and I am happy with how it turned out, here is a picture.



I haven't taken the time to really listen to the speakers yet, but the treble is staying up higher like it should. I performed some tests and I still stand by reversing the tweeter. I tested the speaker from about 30-36" out, and at 3 heights: -6, 0, +6 from the waveguide center. 6 down looks the best to me. Inline with the waveguide height and reverse polarity looks like it tests flatter than 6" above with the correct polarity.

I eventually plan on running these speakers behind an AT screen with the tweeter at about speaker level, isn't that the normal position?

Here is the 6" below waveguide graph:


At the center of the waveguide:


6" above the waveguide:


This one is comparing -6 down and 0 with reverse polarity against 6" up with normal polarity:
post #146 of 163
Thread Starter 
The speakers seem to sound much better now. I decided to measure the response from my LP at about 14 feet back and the treble dropped off faster than the previous graphs. Because of that I decided to try the resistor mod with 6 ohm (normal), 2 ohm, and 0 ohm. Here are the results.

post #147 of 163
Well that's a lot better smile.gif Glad you found the issue.
post #148 of 163
You'd be surprised at the number of emails and PM's I got from people worried if they should be flipping the polarity in their SEOS speakers after this thread. smile.gif Even different models. I do believe you freaked out a lot of people sir. biggrin.gif

But it's good to know you found the problem with your crossover assembly and they're now put together correctly. I'm considering just making all the kits come with preassembled crossovers, but not sure just yet.
post #149 of 163
Quote:
Originally Posted by NathanJ View Post

I finally figured out the problem, I had the 2 similar inductors switched. I also decided to layout my own crossover and I am happy with how it turned out, here is a picture.

Hi Nathan,

Awesome job sorting this out! Your testing, measurements and persistence are the backbone of this DIY group! wink.gif You did a great job with your inductor layout, and I appreciate all of your polarity testing and resistor measurements. What a great education you've gained through your own efforts, and we all benefit by your posts.

As you have discovered, testing speakers is tricky business. Small changes in the position of the measurement mic can show big changes in the frequency response. And then there are all the room effects that insert themselves into the measurements...For now, I am trusting that the designers have crafted the best speaker possible and designed the optimum crossover with the fewest compromises.

Now that I have confirmed that my speakers are all working in a similar fashion, I plan to focus my energies less on optimizing my speakers, which has been done by the designer, and more on optimizing my room and the rest of my system, which is distinctly below par.

Great work, man! biggrin.gif

Mike
post #150 of 163
Thread Starter 
I realized that I had forgot to save my measurement from yesterday when I had the coils swapped, I only saved the measurements from the crossovers that were corrected. I decided Ito unsolder the 2 tweeter coils and then using alligator test leads I was able to hook it up backwards like I had originally done.

Here is what it looks like when you have the tweeter reversed and coils backwards vs coils correct tested on axis at 30 inches:
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: DIY Speakers and Subs

Gear mentioned in this thread:

AVS › AVS Forum › Audio › DIY Speakers and Subs › Testing Cheap Thrills with REW