ND105-4+ND20FA-6 in 4" pipe... lots of questions

[mintos]

I'm trying to make a near field PC monitor with the ND105-4 and ND20FA-6. The woofer will be mounted in a 4"PC pipe facing upwards for an omni affect.

I've never tried an omni project before, so I'm wondering how it would sound compared with a normal layout of the speaker facing foward. Also, how do I use PCD to model a speaker facing up? I will also have the tweeter behind the woofer so should I put in a - for the z axis?

The ND105 also looks ragged around 900hz. Is there a way to smooth it?

[jbruner]

I'm working on a similar project. I need to get busy and post an update.

Some things I've learned-

It seemed to work better with the tweeter positioned roughly above the center of the woofer. (For PCD 6.20 and older, z-offset is a positive number for acoustic centers behind the baffle plane.)

A generous band of felt just inside the mouth of the pipe really helped the sound. You can't really get much stuffing that close to the woofer, and even though it looked like it took up much of the air pathway immediately around the woofer basket, it didn't hurt the bass response at all.

I'm using the factory response curves too, but I take them with a grain of salt. The response is probably not that bad, especially off-axis. It looks like they referenced 1 watt instead of 2.83v, so I increased the level by 3dB in Response Modeler.

It's hard not to wind up with full BSC with this driver. These might be best out in the open. As an omni, I would try to leave at least 2 ft. from surrounding objects.

I modeled baffle step and diffraction in RM, toggling to 80 deg or so off axis, which almost makes the diffraction signature disappear. Then I used PCD on just the woofer. (Put high resistance and/or small capacitance on the tweeter side to get it out of the picture.) I toggle it to between 80-90 deg. off-axis, (don't forget to enter piston dia.) and save it out as a .frd file. Now I can start a new PCD session and use that file for the woofer.



This sounds very good. I prefer it with the large cap, which eliminates the unwanted boominess. These tweeters won't play as low as would be ideal in this type of design, but I don't mind the dip around 3k.

The tweeter is about 3/8" behind the center of the woofer. Your tweeter should work just as well with a different crossover.

I decided to use a small baffle for the tweeter so I could reduce on-axis diffraction and get more low end from it.



I have listened to this a lot, but I don't have this together right now. I still don't have all the pipe I need and I've been busy, but I plan on completing them in the next couple months.

[mintos]

hmm... this thread doesn't seem to have a lot of love.

So from what you said ... to model 90 degree omni speaker, I have to first model the woofer with close to 90 axis. Then I have to extract the frd and use it in a new PCD session with regular angle and the tweeter. Correct?

[jbruner]

Well, that's how I did it. I don't know if there is a "proper" way to model an omni without measurements.

[mintos]

fair enough, I'll give it a try. So far the XO I came up with won't cost me more than 20 bucks as I have 18g .58 mh inductors lying around.

By the way, what's the internal dimesion of your pipe? I'm going for 4" dia internal and about 8" long.

[jbruner]

It's currently a 38" long MLTL 4" internal diameter. Will probably cut back to 34-36" for the final version. You will probably need to stuff your short pipe with fiberglass to get enough damping in that small space. Even then you may wish to experiment with large caps on the woofer.

[Paul Carmody]

Well, to do such a project the proper way to do it would be to measure the woofer 90 degrees off-axis, and use that FR. The tweeter would be normal on-axis.

You might be able to model the woofer's Frequency Response 90 degrees off-axis, then use that. Not sure, though. It'd take some prep work.

[Wolf]

Well, that's how I did it. I don't know if there is a "proper" way to model an omni without measurements.

That's how I would go about it.
Wolf

[mintos]

In PCD it doesn't allow you to directly change the vertical off axis angle directly. So do I set the horizontal distance to minimum and then adjust the z only?

[Wolf]

In PCD it doesn't allow you to directly change the vertical off axis angle directly. So do I set the horizontal distance to minimum and then adjust the z only?

No. Leave the Z alone, and use the 'vertical panning' to get what you want. Take it to about 90 degrees, and then you have your plot. Leave the X offset as set for the system you are planning.
Later,
Wolf

[mintos]

I must be doing something wrong. When I tried to start a new PCD session with just the woofer and used the vertical panning to get to 88 degrees, my graph looks very bad.

Looks like there's no mid, a straight line going down.

[jbruner]

I must be doing something wrong. When I tried to start a new PCD session with just the woofer and used the vertical panning to get to 88 degrees, my graph looks very bad.

Looks like there's no mid, a straight line going down.

It looks like maybe your piston diameter is off by a factor of 10. Should be about 75 mm.

[fbov]

Well, to do such a project the proper way to do it would be to measure the woofer 90 degrees off-axis, and use that FR. The tweeter would be normal on-axis.

+1

My Pluto clones were based on measurements made in the intended application, both drivers measured on the tweeter axis so I didn't have to play with offsets. In two incarnations, PCD was dead-on with the bad tweeters and close but not quite with the good. I haven't investigated why.

The key is to use a tweeter that can reach the midwoof's max 4pi frequency, the highest frequency at which 90* response is equal to its 0* (on-axis) response. I used a 140mm driver and crossed at ~1500Hz. With a smaller driver, you can cross a bit higher.

HAve fun,
Frnak

[mintos]

looks like I'll have to find a different tweeter. Need it to cross lower. How's the OX20SC00-04?

[mintos]

I finally got around to finish up my XO. The blue line is the modified FRD for the ND105-4 at 88 degrees. I'm going to order the parts today, but I want to see if anyone has some suggestions. tnx

[mintos]

btw... I have lots of those .58 mh inductors around from the sale of those speaker terminals from last year. Thus you see that value a lot

These are the only parts I need. Not to expensive. But those PVC pipes costs a lot @ homedepot.

Item Qty. Price Total Remove
6.8uF 100V Non-Polarized Capacitor
Part Number: 027-336
In Stock
$0.42 $6.72
Dayton DMPC-6.8 6.8uF 250V Polypropylene Capacitor
Part Number: 027-424
In Stock
$2.79 $16.74
Pressfit Speaker Terminal 3-1/8" Satin Nickel Binding Post
Part Number: 260-299
In Stock
$2.25 $4.50
Jantzen 0.15mH 20 AWG Air Core Inductor
Part Number: 255-022
In Stock
$3.27 $6.54