AVS Forum banner

Building from scratch. NO KIT

8K views 151 replies 20 participants last post by  dashpuppy 
#1 ·
Well, adventuring off of kits and building my self a pair of 2 way speakers with my own choice of drivers ( with in my specs ).

I went with a PP103 10" Efficient driver & Fountek NeoCD3.5 Ribbon Tweeter. Both drivers are 95.6DB efficient.

Drivers, 10"





Ribbon,




WinISD box,



The goal is to build a pair of speakers that are efficient, and clean sound with 4 class a amps.

I will be using a MiniDSP 2x4 for the crossover.

I don't have the wood working tool's, so im leaving the box building to a cabinet making friend of mine :)
 
See less See more
4
#6 ·
Neo 3.5 has a recommended cut off of 2k5/3rd order. This isn't likely to end happily.
 
  • Like
Reactions: notnyt
#12 ·
Hey Dashpuppy. I see you're in Nanaimo. I'm in Ladysmith, about 30mins from you, and I work in Nanaimo Mon-Fri. If you run into troubles (which I agree with the above, you probably will) contact me. First time speaker design is often alluring, but the results are often bad. You may have a honeymoon phase with your new creation, but it'll subside. I usually warn newbs that there's nothing about using a minidsp that changes the science of XO design. The only difference is you can disregard impedance. The drivers still need to match directivity, SPL capability, avoid tweeter strain, avoid woofer breakup, blend through the XO, keep distortion low, etc.

Where did you get the TS parameters for that woofer? I've never heard of it.
 
#16 ·
Dash, we don't design speakers based on spec sheets. Spec sheets lie.

A woofer's operating range is usually limited by break-up, not roll-off at the high end. A tweeter's by roll-off at the low end, and distortion. When you measure your drivers for FR, you'll find out the woofer's actual operating range. Distortion curves will also tell you where the tweeter's happiest. Of course, that's only on-axis.

That's in addition to all the stuff tux and Mfusick mention. Take Tux up on his offer!

HAve fun,
 
#29 ·
#31 ·
I wish you were over here teaching me !! LOL playing with this software

Not sure i'm on the right track, but this was my first measurement, then using the 2x4 went to this.

remember i'm learning and playing, i'm using a pair of my 100$ 2 way speakers i bought 10 years go just before A&B sound closed the tweeter has a small cap on it and the woofer has nothing it's wide open.

Number 1




Number 2

 
#37 ·
Top right, button called limits. You want to adjust the top/bottom so that you have 5dB per division. I usually go with 40-100dB or 50-110dB but it depends on your screen resolution and whether or not you run REW fullscreen or have it windowed. You may also want to close in on the frequency limits, 20hz-20khz is a more standard range there. Finally click the frequency axis button to the left of the limits button, I see that you have it in linear scale and that button will switch it to logarithmic scale which is more useful for looking at frequency response data.
 
#50 ·
Yes much better.

Now under the graph dropdown on the top toolbar you can apply smoothing to the response which will help cut down on the noise you are seeing. Try starting out at 1/12th octave but be sure to play around with the other settings to see what more or less smoothing does.

Finally there is the IR Windows button, this allows you to gate or window the measurement, basically choose the period of time from which REW uses to calculate the frequency response. By properly gating a measurement you can close in on the time of flight of the direct sound from the speaker and eliminate some reflections from within your room which take longer to get to the microphone and this can get you a cleaner measurement. However there are limits it it's use. Lower frequencies take longer to complete a cycle, often longer then the time it takes some of the reflections to make it to your mic so a short window to eliminate reflections often comes with a loss of low frequency resolution. Exactly how low you can get useful resolution is directly related to how far you can place the speaker from nearby objects. Indoors it is unlikely you will get useful direct sound data below 500hz or so unless you are taking nearfield measurements which have their own set of downsides.
 
#52 ·
Dash Dog, you are frustrating me, man! :D

I'm going say what everyone is probably thinking, and suggest you get EXACTLY what you want and order a couple of these:
http://www.diysoundgroup.com/waveguide-speaker-kits/fusion-10/fusion10-pure-kit.html

If you have any questions about it technically, you can ask the designer of the speaker and cross over, your new-found best friend, Ryan!

You still have to help him move speakers, though. :p
 
#53 ·
I'm going to help Ryan move them for sure !! lol !!

For now i'm not buying any kits, i won't learn designing if i buy someones kit they already did :) sorry dude, I'm very stubborn :)

Ryan did say he has a few waveguides i can try and play with..


I also have 2 other speaker projects I'm working on too :)
 
#54 · (Edited)
You should be measuring at 3ft, and using variable smoothing.
You are -20db @ 100Hz vs 2KHz.

Your Y axis shouldn't be 180db and -180db.
http://www.guinnessworldrecords.com/world-records/quietest-place

^^^
180db is the SPL that a space shuttle launch at 3ft produces, and -20db is beyond what the Guinness world record of quietest-rooms can achieve (6ft solid cement with 6ft of 100% absorption).
That would be like measuring your cars horsepower in contrast to the Hadron Collider's 26 TeV's (you can do it, but the scale is absurd...)

110db and 50db would be more appropriate. Which is 5db hotter than THX Ultra2 Reference Level.
 

Attachments

#55 ·
You should be measuring at 3ft, and using variable smoothing.
You are -20db @ 100Hz vs 2KHz.

Your Y axis shouldn't be 180db and -180db.
http://www.guinnessworldrecords.com/world-records/quietest-place

^^^
180db is the SPL that a space shuttle launch at 3ft produces, and -20db is beyond what the Guinness world record of quietest-rooms can achieve (6ft solid cement with 6ft of 100% absorption).
That would be like measuring your cars horsepower in contrast to the Hadron Collider's 26 TeV's (you can do it, but the scale is absurd...)

110db and 50db would be more appropriate.
Oh you mean i shouldn't have the microphone 1" from the speaker :)

I moved it back later on 1m, when i was playing etc etc, Im taking Ryan's advice and going to visit him for a bit and take some lessons :p
 
#56 · (Edited)
1inch is fine if you want to measure the drivers individually for XO'ing.

If doing full bandwidth sweeps like this, it is best to be 1.5-3ft from the speaker, otherwise the mid will be too far off axis vs the tweeter (depending on the height of the mic and directivity of the drivers of course...)
and further than 3ft and you'll probably get too many room influences. (Measuring outside well away from objects is the ideal/more-preferred way, but few do it, and that is fine I guess, it's kind of inconvenient.)

I can see 1099's in the OP's future very soon.

Many people have complained about the miniDSP's noisefloor going active with it (but some of those folks are using 107db+ efficient tweeters so...)
With subs and woofers you'd never hear it; just mostly HE tweeters in silent rooms.
 
#57 ·
1inch is fine if you want to measure the drivers individually for XO'ing.

If doing full bandwidth sweeps like this, it is best to be 1.5-3ft from the speaker, otherwise the mid will be too far off axis vs the tweeter (depending on the height of the mic and directivity of the drivers of course...)

I can see 1099's in the OP's future very soon.

Many people have complained about the miniDSP's noisefloor (but some of those folks are using 107db+ efficient tweeters so...)
I was actually joking :) i know it has to be 1m away my first screen shots were just to show that I'm playing learning and having fun with this new software. Never used it before or a microphone. Just used my ears and used a lot of full range setups..
 
#66 ·
The original H290 horn is cheaper than the newer model that you're showing .

( original ) Eminence H290 from "Q" Components for $36.cdn


There's little to no functional difference between these two models ( that's particularly relevant when you're going to be using custom filters within a miniDSP ) .

- I thought you said you weren't against saving a bit of money / $30. saved is money best left in your pocket ( or applied to other costs of your project ) .

:)
 
#69 ·
Was up late lastnight listening to some WG's as ugly as this round one is, it sounds pretty decent.. Will not hate it, but looks fugly LOL !!



2x4 using it for 2 drivers one cd & 10" one per channel on my class a's.



Playing with cross over points..






Different unit, ( same CD )



Totally not using one of these :( was shocked !!

 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top