MBOWI 3-Way Phase and Measurement Question

[Alex2507]

My version of these speakers has the Peerless 10" woofer used in this design:

http://murphyblaster.com/content.php?f=MBOW1_3-WAY_previous.html

The 3/4" round over evening out diffraction artifacts around 4 KHz that Dennis talks about has me wondering if my odd placement and addition of non rounded over angles could produce enough of an acoustic effect to make a tweeter with reversed polarity from the original crossover (xo) design produce a graph showing a surprisingly smooth frequency response (FR) curve at the listening position (LP) about 5' away.



Mic placement is responsible for the wacky LF stuff going on there. I should have saved the measurements from having the tweets in proper phase but you can trust that the nulls and peaks were considerably larger than the ones you see here:




I double and triple checked my wiring on this.

Right at the speaker, the - on my woofer is the same as the + on my mid. So that checks out. But that graph is generated by a reverse polarity on the tweet. I'd like to understand how reverse polarity is generating a smoother FR and I'd like a few tips on mic height and distance from speaker for when I'm not interested in the LP FR.

I put the measurement gear away for now. Trying to incorporate YPOA into this is complicating the issue as well. I suppose I could run YPAO in phase and reverse phase to be stored in the Yammy 2600's ample memory. Right now with the tweet out of phase YPAO sees correct polarity but YPAO admits that their ruling on polarity might be meaningless.

The ? is whether my bookshelf placement might be the cause for this polarity oddity and how should I be measuring FR.

TIA

 

[Swerd]

There are 3 traces in your graph. What are they? And which one(s) are you talking about?

The MBOW1 3-way is a 2nd order design. That means the drivers at each of the two crossover points should be in opposite polarity with each other. The schematic diagram shows that. The mid is wired with opposite polarity to the woofer, and to tweeter.

When you say:

"Right at the speaker, the - on my woofer is the same as the + on my mid."​

Does that mean the woofer and mid have reversed polarity with respect to each other? If so, you have it right.

But I don't understand what you're saying about the tweeter. According to the schematic, it should also be in opposite polarity to the midrange.

 

[everettT]

Wouldn't it create a null or suck out at crossover region if wired in that manor?

 

[Alex2507]

There are 3 traces in your graph. What are they? And which one(s) are you talking about?

The top is a bit of a mystery. The middle one is the mic calibration file I think and the bottom gold graph is the FR at the LP.

The MBOW1 3-way is a 2nd order design. That means the drivers at each of the two crossover points should be in opposite polarity with each other. The schematic diagram shows that. The mid is wired with opposite polarity to the woofer, and to tweeter.

When you say:

"Right at the speaker, the - on my woofer is the same as the + on my mid."​

Does that mean the woofer and mid have reversed polarity with respect to each other? If so, you have it right.

Yes, I do have that right. Look at the schematic and you'll see that electrically those points are the same with nothing other than copper in between, showing 0 Ohms.

But I don't understand what you're saying about the tweeter. According to the schematic, it should also be in opposite polarity to the midrange.

Right, what I'm saying is that I have the tweeter wired backwards from the schematic and it produces a smoother graph at the LP than being wired as per the schematic. Actually I have taken measurements from about 18" away from the speaker sort of on axis and both results support my findings that the graph is better with the backwards hook up.

I don't know enough about how angling the bookshelf and changing the plain of drivers changes the phase as recorded at the LP nor do I really know a lot about properly measuring the FR. I'm using REW and a calibrated ECM8000 mic.

 

[Swerd]

The top is a bit of a mystery. The middle one is the mic calibration file I think and the bottom gold graph is the FR at the LP.

Got it. I'll pay attention to the gold trace.

Right, what I'm saying is that I have the tweeter wired backwards from the schematic and it produces a smoother graph at the LP than being wired as per the schematic.

So, if I understand correctly, you have the midrange wired in standard polarity, and the woofer & tweeter in reversed polarity. Right? If so, that should work no different from what is shown in the schematic. As long as the mid is wired in opposite polarity to both the woofer & tweeter, you have it right.

Somehow, I get the vibe that I am being optimistic, or you wouldn't be asking.

I don't know enough about how angling the bookshelf and changing the plain of drivers changes the phase as recorded at the LP nor do I really know a lot about properly measuring the FR. I'm using REW and a calibrated ECM8000 mic.

I doubt if angling the bookshelf would do that.

Edit: But exactly where you put the microphone (how close, how high, and at what angle) can make a very large difference.

I can see Dennis doing a prolonged face palm as he reads that you are "using REW". He has privately complained how he wishes people never tried to use it to make speaker frequency response measurements. There are many reasons why their measurements can differ from his, and he honestly doesn't know why.

 

[j_garcia]

I doubt there is a large difference in the chamfer vs radius would make a difference in the listening area (maybe off axis a little?) but not that you'd notice I'd guess.

My mains have a ~5 degree tilt in the cabinet, but I think that is more because the ribbon throws such a HUGE stage that if you aim it straight out it sort of goes beyond the listening area. So I think it is more that than any time alignment.

For the Tweeter, Dennis may have had his reasons for that config, and maybe he felt the response difference wasn't enough to matter audibly? The tweeters themselves can have different response also. When you order tweeters from GR, he sends the graphs of each driver and hand selects a pair that match very closely.

I still love my CA0W1s

You can see the angle of the Carnegies

 

[Alex2507]

Wouldn't it create a null or suck out at crossover region if wired in that manor?

Dennis talks about the midrange being wired out of phase and the need for it about half way down the page of the link you provided a few days ago.


http://murphyblaster.com/content.php?f=MBOW1_3-WAY_previous.html

Got it. I'll pay attention to the gold trace.
So, if I understand correctly, you have the midrange wired in standard polarity, and the woofer & tweeter in reversed polarity. Right? If so, that should work no different from what is shown in the schematic. As long as the mid is wired in opposite polarity to both the woofer & tweeter, you have it right.

Somehow, I get the vibe that I am being optimistic, or you wouldn't be asking.
I doubt if angling the bookshelf would do that.

I can see Dennis doing a prolonged face palm as he reads that you are "using REW". He has privately complained how he wishes people never tried to use it to make speaker frequency response measurement. There are many reasons why their measurements can differ from his, and he honestly doesn't know why.

Way back when you said that this was the best thing ever in terms of DIY lessons you were correct. On top of using REW I also took an Ecectricity/Electronics class in high school so ...

I understand the diagrams regarding polarity and see that you do as well. The mid is the only thing that is suppose to be out of phase. As of right now, my speakers are as per the schematic and YPAO is disabled and SRV is playing Life by the Drop on repeat. I can't switch the polarity quick enough to tell with ears alone on this one tune even though it is a very simple song.

REW is put away for now. I got my ears on.

 

[Swerd]

I understand the diagrams regarding polarity and see that you do as well. The mid is the only thing that is suppose to be out of phase. As of right now, my speakers are as per the schematic and YPAO is disabled and SRV is playing Life by the Drop on repeat. I can't switch the polarity quick enough to tell with ears alone on this one tune even though it is a very simple song.

REW is put away for now. I got my ears on.

Excellent! If it ain't broke, don't fix it anymore!

I always get confused by the words and phrases we use for talking about in-phase, out-of-phase, and speaker polarity. So I try to be literally precise when I talk about it.

 

[Alex2507]

Excellent! If it ain't broke, don't fix it anymore!

So you're saying NOT to yank the drivers in the ST's for a look under the hood ?!?

I doubt there is a large difference in the chamfer vs radius would make a difference in the listening area (maybe off axis a little?) but not that you'd notice I'd guess.

My mains have a ~5 degree tilt in the cabinet, but I think that is more because the ribbon throws such a HUGE stage that if you aim it straight out it sort of goes beyond the listening area. So I think it is more that than any time alignment.

For the Tweeter, Dennis may have had his reasons for that config, and maybe he felt the response difference wasn't enough to matter audibly? The tweeters themselves can have different response also. When you order tweeters from GR, he sends the graphs of each driver and hand selects a pair that match very closely.

I still love my CA0W1s

You can see the angle of the Carnegies

I'm not convinced that time alignment isn't a consideration or even 'the' consideration that led to that 5 degree angle. I think I know that it's a consideration with the way the GR mid surface mounts to a baffle but in post 1 you can see how far off axis my bookshelf is from the bass cabinet. It's got to be ~ 22.5 degrees but none of that should affect the upper xo point.

 

[Swerd]

So you're saying NOT to yank the drivers in the ST's for a look under the hood ?!?

You mean you don't rotate the drivers every 250 hours ?

There are raging debates among owners & audiophiles over which rotation pattern should be used. I favor switching the upper left midwoofer with the lower right, the upper right with the lower left, and a simple switch with the tweeters.

Then there are those who actually rotate the drivers, some with and some without, switching upper and lower midwoofers.

The owners seeking the ultimate improvement reverse the internal wires. This can help even out asymmetric copper crystal distortion. While doing this, some actually keep the wires overnight in the freezer, just in case they need cryo-treatment renewal . I think this last one is a bit extreme, as the original cryo-treatment is probably sufficient. And, it is questionable if a typical home freezer has a low enough temperature.

 

[Alex2507]

I doubt there is a large difference in the chamfer vs radius

Actually Dennis talks the 3/4" round -over being an improvement over a square baffle and has the graphs showing that difference in the link in post 1. To me it looks minor on the graph (~4khz) and I don't think I could pick that out reliably. Chamfers weren't considered.

EDIT: I have the roundovers on the bookshelf portion but I exposed squared off parts of the bass cab when I effectively pushed the bookshelf back to toe it in. Plus the desk is there ... I'm still wanting to know how seriously I can take REW.

 

[j_garcia]

I mean if you have chamfers vs rounds that are about the same size, results should be closer than a square corner.

 

[Alex2507]

Another question: would switching the placement of L3021 and C3021 have any bearing on what the tweeter sees? I suspect it wouldn't because my 7uF cap is in between the 9 Ohm resistor and 1 mH coil.

Another question: If I was going to make a center channel with 2 GR Research mids ... front ported? This I only ask because I happen to have a single OW1 tweeter not doing anything.

 

[Swerd]

Another question: would switching the placement of L3021 and C3021 have any bearing on what the tweeter sees? I suspect it wouldn't because my 7uF cap is in between the 9 Ohm resistor and 1 mH coil.

The order of those components, when they are in series with each other, in a parallel part of the network, does not matter. It functions as a notch filter also known as a trap filter or LCR trap.

Another question: If I was going to make a center channel with 2 GR Research mids ... front ported? This I only ask because I happen to have a single OW1 tweeter not doing anything.

Yes, I would do that front ported. Use two of these: 1¾" inner diameter × 4" long tubes. Use them as they are without trimming the length. The cabinet should be double the MB27 volume of 0.38 ft³ – or roughly 0.75 ft³.

It may be that the OW1 tweeter is about 3 dB less sensitive than what you need with two GR Research M-130 midwoofers. In the standard ST, Dennis uses OW2 tweeters because the are about 3 dB more sensitive than the OW1. But there may be a way to get away with using the OW1 that you already have. Pad the M-130s down with a 1, 2, or 3 ohm resistor. Maybe the simplest way is to to use one M-130 instead of two and build a MBOW1.

 

[Alex2507]

Maybe the simplest way is to to use one M-130 instead of two and build a MBOW1.

... clearly.

It's likely to sound better than anything that I would cobble together but I do appreciate seeing how the doubling up of MB27 box volume and ports carries over from doubling the drivers. Makes perfect sense. Now if I bi-amp that tweeter ...

Thanks for the rundown on that notch LCR business too. I might already know that if I had read that book, right?

 

[JohnnyK]

Attached are photos of my MBOW1-3way. They are of the original design with the original woofer. I did not like the look of a small box sitting on a large box so the woofer cabinet and the Mbow1 cabinet share a common front baffle. I removed the drivers from the Mbow1 and mounted them onto the new front baffle. When viewed from head-on the Mbow1 cabinet is hidden behind the front baffle. I rounded over the outside edge of the front baffle. I have enjoyed these speakers for years and I still think they sound great.

 

[everettT]

Did you take measurements before and after? I like the look of the baffle.

Attached are photos of my MBOW1-3way. They are of the original design with the original woofer. I did not like the look of a small box sitting on a large box so the woofer cabinet and the Mbow1 cabinet share a common front baffle. I removed the drivers from the Mbow1 and mounted them onto the new front baffle. When viewed from head-on the Mbow1 cabinet is hidden behind the front baffle. I rounded over the outside edge of the front baffle. I have enjoyed these speakers for years and I still think they sound great.

 

[BoredSysAdmin]

I also took an Eclecticity class in high school so ...

First honest thing you said in a while ;-)

 

[JohnnyK]

I did not take any measurements. I just did what Dennis suggested. Also, I used to run the woofer cabinet with a port. I have since plugged the port because I found that the bass was too boomy. Now get this---I use two AR-3a speakers as subwoofers. They are behind the Mbow1-3ways. They produce a nice, solid bass. Not boomy at all.

I also use a MBOW1 as a center speaker. It is located in the center cabinet. I installed bass traps on the walls.

 

[Alex2507]

Attached are photos of my MBOW1-3way. They are of the original design with the original woofer. I did not like the look of a small box sitting on a large box so the woofer cabinet and the Mbow1 cabinet share a common front baffle. I removed the drivers from the Mbow1 and mounted them onto the new front baffle. When viewed from head-on the Mbow1 cabinet is hidden behind the front baffle. I rounded over the outside edge of the front baffle.

I just saw this post last night. Good going on that look. On the off chance you still come around, I am curious to know if you have coupled the two cabinets with fasteners.

My MBOW1 3-ways are going to dlr's house this Saturday for a New England DIY GTG. PETT forum has the particulars.

EDIT: BTW, I am stealing that design concept. Thanks.

Edit II: I invited SWMBO to the festivities this Saturday. She said, "I don't think so."
I think she may only be interested in me for my looks.

Edit III: I can't believe I saw those pic's 3 years ago in Kurt's thread and forgot seeing them. I clicked 'like' and it has a date attached.
Obviously I have too much time on my hands.