[HT cOz]

Yes it is, but there are some other issues with this speaker that are pretty significant that is masking a lot of the speakers true performance. I think I can correct this and then the differences in the crossovers will be more easily discerned. You'll see what I mean shortly.

I might also be able to put both networks inside the speaker and use just one switch between them, but it will require one more set of binding posts to be added to the back of the speaker. What do you guys think of that?

Danny,

I think that sounds great, but the purists would probably rather see it done like this:
1). Speaker with a binding post for tweeter and woofer:
2). Mystery box with 3 sets of input binding posts and 3 sets of 2 output binding posts.  They would likely want the external crossover box to contain three networks that are the original, cap upgraded and the new Maestro network. 

They would probably also prefer that they are not labeled and are truly mysterious

Maybe someone could volunteer such a box?

[johnzm]

i am sitting here, patiently waiting results. if Danny says these are in fact correctable issues i im probably going to buy a pair ofthese ASAP!

[DFaulds]

Does that mean that there will a difference in the cicuit of the two crossovers, or merely just a difference of part sources?  If not the later, I just don't see the point.

I think the switch with additional binding posts ensures that the differences are isolated well enough.  That is if the crossover itself isn't different between the two.

Yes it is, but there are some other issues with this speaker that are pretty significant that is masking a lot of the speakers true performance. I think I can correct this and then the differences in the crossovers will be more easily discerned. You'll see what I mean shortly.

I might also be able to put both networks inside the speaker and use just one switch between them, but it will require one more set of binding posts to be added to the back of the speaker. What do you guys think of that?

[jtwrace]

i am sitting here, patiently waiting results. if Danny says these are in fact correctable issues i im probably going to buy a pair of these ASAP!

Yeah, watch for the GR crossover correction kit.   

[werd]

Yeah, watch for the GR crossover correction kit.   

I need one of those for my kef's. I would like to try on my AZ's too but first things first.

[johnzm]

Yeah, watch for the GR crossover correction kit.   

hey it woudnlt be the first time



i actually have a pair of those too haha.  so which speakers sound better danny the insignias or the b2031p

[TRADERXFAN]

Keep in mind the original contention went something, loosely, like this...
__________________________________
"Hey look at this off axis graph of freq response and impulse, these speakers are amazing and only 152 bucks"

>"but how do they sound?"

"Just look at the graph!  what else is there to say?"

>"Well, in my experience, that isn't the whole story to be told on how good they sound. You can't just look at the graph and make a decision"

_______________________________________

So showing how they can be made to sound better without changing the way they look on the graph, that would be equally demonstrative of the basic point "the freq graph isn't enough to know how they will sound"

Somewhere it came down to a more distilled version of the question "would or would not component substitutions in the crossover be significant improvement, as long as they were exact same value?"

-Tony

[dwk]

IMHO it's a bit unfortunate that all this activity sprang up around the 2031s instead of the 2030's (the 6" version) In all the early buzz over at audioholics and hometheatershack there seemed to be a pretty strong consensus that the 2030 was a better overall speaker particularly through the midrange - maybe not surprising given the 8" vs 6" driver difference. It gives up even more of the already scant bass response though, so might not really be suitable for standalone use without a sub. 
 The speakers appear to be fairly similar and so broad observations will probably apply to the 2030 as well, although specific component values etc will obviously differ.  Just throwing this out since anyone thinking of actually buying a pair may want to do a bit of research on the 2030 vs 2031 before plunking down the bucks.

I'm currently using the 2030p as a computer monitor and bone stock they aren't too bad once I stuck some foam in the ports - despite the 'waveguide' (which is much shallower on the 2030 than the 2031), there seems to be some diffraction off the slot ports flanking the tweeter. I suspect that driver quality will limit how good these can ultimately be and may undermine the value proposition of an xover upgrade, but it's still an interesting process to watch.

[Danny Richie]

Okay, I have a lot of information to post about this thing.

I have measurements, problems, fixes of the problems, listening impressions, solutions for simple comparisons, etc.

So I may post a little at a time as my time permits.

Keep in mind that I am doing all of this for free...

Danny,

I think that sounds great, but the purists would probably rather see it done like this:
1). Speaker with a binding post for tweeter and woofer:
2). Mystery box with 3 sets of input binding posts and 3 sets of 2 output binding posts.  They would likely want the external crossover box to contain three networks that are the original, cap upgraded and the new Maestro network. 

If I did all of that then I'd have twice as much in binding posts as the whole speaker cost.

I want to make this as easy as possible for the guys comparing it. So here is what I am looking at in that regard.

It was very easy to replicate the network. Using the higher quality parts, the measured response for the woofer and tweeter are identical to the stock crossover.

Now, the tweeter network has one 4.3uF cap in line with it and a 12 ohm resistor to ground. That's it.

The woofer network has a 2.0mH inductor on it and a 10uF cap to ground. That's it.

The easiest and least expensive thing to do is add one set of binding posts to the back of the speaker. I then add only one double pole double throw switch to the back of the speaker also. So there is an "on" position in one direction and an "on" position in the other direction.

Se each cap and inductor in line with the drivers will all be connected to the drivers all the time but each to its own set of binding post. So they are only in the signal path if connected to their binding post. The two components in shunt from each circuit can then run to one switch to connect or disconnect them to ground.

So plug into the stock binding posts and flick the switch towards its binding posts and all of the stock crossover is used.

Plug into the second set of binding posts and flick the switch towards its binding posts and all of the new crossover is in the signal path.

That is as easy as I can make it, and it doesn't cost a lot due to less complexity.

If that is okay with everybody then we can move forward with this thing.

[HAL]

That sounds good to me! 

[Danny Richie]

Okay, I need two good quality double pole, double throw switches.

Anyone know of a good source?

[wushuliu]

Okay, I need two good quality double pole, double through switches.

Anyone know of a good source?

I Like NKK. Well-made.

http://www.alliedelec.com/search/productdetail.aspx?SKU=8700409

datasheet:

http://www.alliedelec.com/Images/Products/Datasheets/BM/NKK_SWITCHES_OF_AMERICA/870-0443.PDF

[Danny Richie]

Actually a single pole double throw switch will work fine.

[HAL]

Sorry, I cannot resist.  Must have power! 




Unfortunately it is probably bigger than the speaker.

[werd]

Frickn ha!!!! ^^^^^^^^

[Danny Richie]

Okay here are the measurements.

First off is the on axis frequency response.



Not bad actually. There is some cancellation going on in the throat of the wave guide that is causing some cancellation in the 15kHz range, but I don't think anyone will notice too much.

The real issue though was the peak at 750Hz.

Checking the spectral decay we can see that there is some excess energy there from something.



That is going to show up in the upper vocal ranges for sure.

I have seen this type of peak before and was pretty sure this was going to be a fixable issue. This is the sort of peak that you get if you don't radius the back side of the woofers mounting hole. This was later confirmed when I removed a woofer.

It is funny that as well engineered as this speaker is in some areas it really fell short in others. Behringer engineers that read this should be sending me a thank you note soon for that one. It cost nothing to fix in production.

Vertical off axis looks pretty good.



The horizontal off axis looks good too.



The impedance has a pretty heavy rise to it but not bad. It shows a pretty low tuning in the 37Hz range and a little bump at 50Hz that is likely a standing wave issue in the box. This is typical of a square shaped box with not enough internal dampening material. For those of you that are doing a DIY design please remember to only allow the width of the box to be about 1/2 of the depth of the box if possible and no greater than about 2/3rds of the depth. This will help the standing wave issues inside.



I also noted that the ports are pretty small in size (not too long either) for a tuning that low. This could mean some port noise. Typically to tune a box that low with an 8" woofer being involved then the ports would have to be bigger in diameter and much longer to tune it that low and not have port noise.

We can also see that the impedance drops to 4.1 ohms in the 147Hz range. So if using a tube amp, make sure it has a 4 ohm tap.

Next is the listening impressions.

[Danny Richie]

Okay here is the summary of my listening impressions.

I am REALLY familiar with my system and how it sounds with various speakers. And most of my test music is stuff that I have used may times for listening for specific things. So when something stands out or is absent then it is quickly noted.

I would also note that I don't usually give impressions of a speaker that really doesn't have enough time on it. However, I have gotten a good feel for what I can expect to change over many hours of play and what is a true character of the speaker. Plus I know all of you guys are anxious.

All in all this speaker has a lot of promise but needs some work. I think it can really be made into a nice speaker. However most of what stood out to me were negative issues. An attention getter is usually a bad thing. If nothing draws attention to itself and I hear only the music, then that is a good thing.

I played several things to let them warm up a little and get a little feel for them, then started to listen critically.

Vienna Teng's "Waking Hour" CD has been one of those make of break CD's that can quickly bring light to the good or the bad. Initially the piano sounded a little flat and lacked some familiar resolution. The speaker also added a little bit of excessive ringing as well. Vocals were excessively heavy and had a glaring edge. Vocals also had a lot of coloration and sounded smeared and a little muddy. I also noted that the side panels were really singing. That added a lot of distraction. I also noted some kind of buzzing that sounded like it was coming from the port on the left speaker.

So I decided that some of this heaviness could have to do with the low impedance load. So I switched to the 4 ohm output taps on my amps.

That helped some. Most of the heaviness in the vocal and bottom end went away and now I could really hear the glare in the vocals.

I then decided I needed a fresh reference.

So I dropped my pair of N1's in their place (back to 8 ohm tapes on the amp...). Suddenly all was right in the world again. A veil was lifted in the vocals and it all became more transparent again. Imaging was much better. The piano had real attack and decay that was the instrument only and not something added by the speakers. I felt like I could touch Vienna's breath as her lips separated....

Okay, let's try this again. Back to 4 ohm taps and the Behringer's, and new music.

Listening to some Keb Mo the vocals lacked some resolution. The guitar was not bad. These speaker can hit the notes, but it was the space between them that was missing. On bass notes the ports sounded like they were resonating and so did the box. It was really adding a lot of coloration.

Track 4 (I don't know what the song is) from the NY Audio Rave CD (famous CD) is a make or break track for most speakers. It has a bass heavy section that can destroy a lot of speakers and this was on of them. The coloration masked everything. I also noticed that the drums that usually move across the top of my room going from one upper corner of the room to the other upper corner of the room were not doing so. Drums were now at speaker level and stayed within the two speakers. Also, complex stuff ran together.

Further investigation revealed that some of the screws that held the plastic front face plates on were not tightened up very well and was allowing the face plate to buss around the port area. That fixed the buzzing sounds that they occasional made.

Bass notes were still a real issue though. In the song "When I Fall" by the Barenaked Ladies, that bottom end was real resonate and a bid distraction.

I would expect to get some smoothing out with more burn in time, but that missing space between notes isn't going to magically reappear. The bass issues won't go away either. The cabinet needs help and the ports may need some stuffing as well.

Next is some coverage of the fixes.

[Danny Richie]

So I took one speaker and stripped it of drivers and the felt dampening material.

I then took a die grinder to the woofer hole and knocked out the sharp edge. I just carefully milled it off using a egg shaped cutter. It was not as big or as nice of a job as if it had been cut from the beginning with a router, but enough to make a dent in the response.

I then put the stuffing back in and reassembled the speaker.

Here is the new measured response.



That looks better.

And the spectral decay:



That looks a lot better.

Behringer please send a royalty to GR Research for next generation version with this correction. 

Now I have to do that to them both.

Both gutted now, you can see the difference in the two of them before making them both look the same.



I then cut into them both a little heavier and made them the same. There are also some small blocks that they must be using during assembly that stuck up right behind the woofer. Those will need to be cut down as well.

Lining them with a little No Rez and figuring out what to do about the port noise and I think we will have a speaker that is free of enough problems to really hear what they can sound like. Those issues would have masked over some of the differences in components. It will be much easier to hear those differences now.

[jtwrace]

That's amazing! 

[werd]

Now you da man!!!!!!