3-Way Crossover Assistance Request.

[Istoc]

I have currently a very rudimentary 2nd order LW-R 3 way crossover design in my speakers. It has no attenuation to account for variables in the sensitivity, nor do I believe the crossover points to be optimally placed.

My goal:
I want to redefine the crossover starting from scratch to make my mid-bass and mid range more prominent, and tone down my Woofer and Tweeter. I also plan on increasing the crossover points by a fair margin since my cabinet design shifted significantly since I originally created the crossovers and it allows me more wiggle room with the implementation.

Current Data:
Tweeter Data Page: ScanSpeak Discovery D2608/9130 1" Textile Dome HDS Tweeter: Madisound Speaker Store

Midrange Data Page: Scanspeak Discovery 10F/8424G 4" Midrange 8 ohm: Madisound Speaker Store

Woofer Data Page: Scanspeak 22W/8534G Discovery 8" Woofer: Madisound Speaker Store

I am thinking of lowering the overall sensitivity to around 85db and leaving the midrange without any attenuation to make it more prominent.
I was also thinking that according to the spec sheets I might set my crossover points at 5khz and 900-1khz. Any information would be greatly appreciated, as well as input into design and development.

For those that might need the information about the cabinetry here is the link for the building photobook: http://photobucket.com/IstocsCustomCabinets
The password is: DIYAudio

As it can be seen I deep set the drivers.

[wintermute]

Hi Istoc, It's very hard (for me) to say without seeing measurements on your actual baffles. You really don't want to attenuate the bass driver if you can avoid it.

Have you considered going active for the bass crossover and passive on the MT crossover? This will allow line level adjustment. That's my plan, though I haven't implemented it yet so can't vouch from experience.

Looking at the FR curve of the mid part of the problem may be the rising response after a bit less than 1k. If you are using plain old 2nd order electrical your acoustic slope will probably not be what you want/expect and this may be compounding problems with the higher frequencies.

It's quite easy to add an Lpad on the tweeter, this should probably be your first step. Due to baffle step the bass may be less of an issue. If you can take measurements though that will give you the best chance of working out what the best option is.

Tony.

[jerome69]

Introduction to designing crossovers without measurement

Could help

[Istoc]

I already read through that posting, actually I contacted Allen directly to see if I could get his assistance. I would do measurements if I had equipment to do so, also I don't run any external amplification so until/if I do Active Crossover networks are kind of out of my area. Honestly the bass drivers presence is fine, it very rarely makes itself more noticeable in the speaker layout blending fairly well. Its more the fact I want to see if its possible to make the mid range more present because at this time its seriously lacking. I don't believe its a driver issue or I would probably just modify the opening and use a different similar driver(perhaps larger). I think it has to do with it being quite a substantial bit lower in sensitivity in comparison to the other two drivers in such close proximity. Sorry about the wall of text I'm using quick reply.

[picowallspeaker]

Originally Posted by Istoc

I
I am thinking of lowering the overall sensitivity to around 85db and leaving the midrange without any attenuation to make it more prominent.
I was also thinking that according to the spec sheets I might set my crossover points at 5khz and 900-1khz. Any information would be greatly appreciated, as well as input into design and development.

(senior noobie inside talkin' )
I always see Dave-Planet 10 that dislikes the use of domes , especially in the critical area around 1 KHz ; Also, that frequency was used in the past ( and still nowadays ) when crossing from a 15" to a 1"or 2" compression driver , but that's another story . So I don't see a good use for a big cone ( a cone midrange compared to a dome or planar ) starting from there .
Often , the mid needs to be attenuated in a 3 way .
What Tony said , to go active , might be a kick start for expanding dynamics and also playing with different points / slopes , I think that's fun !

[wintermute]

Hi Istoc, I think we need to establish what you currently have crossover wise. When you say rudimentary 2nd order L/R do you mean as per a stock crossover calculated for a particular frequency and impedance?

Do you have any impedance compensation on the drivers?

I suspect that there is a bit more going on than level adjustments alone will fix, but I am assuming that you only have theoretical 2nd order L/R electrical filters. How that electrical filter interacts with the driver and resulting acoustic slope of the crossover is what actually matters.

You may also want to check out Daves guide for designing without measurements software right at the bottom of the page). His guide steps you through using various free software to get spl and impedance curves and the use of Passive crossover designer for simulating the crossovers.

It would be worth while doing this and putting your current crossover into the sim just to see what the sim thinks will be the result. It might surprise you! I am of course assuming that you haven't done so already....

One other question as well. Have you tried inverting the mid drivers polarity?

Tony.

[Istoc]

I have no outside Impedence correction, other than what the Inductor and Capacitor in line would equate to which when I checked was almost a nil value. I'll most definitely read his information as well, I can easily swap the polarity and see if it becomes more clear. I will have to follow up more on this later today as its time for me to head to work. I'll check in every couple hours at work to see if there are updates here though.

The crossover design was purely speculatory according to online calculators to be at certain frequencies. I spoke with Wolf while I was making them and followed some of his instruction to verify the values were accurate. At this time the values I used for crossing over are escaping me although I'll try my best to find the current values relative to hz.

Also anything to do with excell I'll have to use Google Docs I don't own anything microsoft, I'm using a student copy of Windows 7 I got when I was taking IT classes. Will most of the stuff respond in google docs do you think?

[picowallspeaker]

(hi !!)
The fact is that our assumptions are made by seeing the photos ,and I see two 7 mH inductors and about 50/100 µF caps . Working with the values you have on hand -just for fun...- you can try to skip the inductance on the midrange after the cap ,maybe place it on the woofer to obtain 14 mH , and double the cap in the HP section . The mid needs to be in anti-phase . And you may try to lift the upper cabinet about 15-20 cm and use 1st order filters between wf and mid . Also a reflex port would change the presence of the mid .

[AllenB]

Without being able to measure, you will face limitations. Still, when you post your crossover you ought to get some suggestions for changing it. As for Wintermute, my first assumption might be that impedance compensation would give you more control and probably improve things.

Originally Posted by Istoc

to verify the values were accurate.

Accuracy of component values is not going to be a significant issue just yet.

I'll try my best to find the current values relative to hz.

Just the component values (as a schematic) will be fine.

[sreten]

Hi,

I'd say both your suggested c/o points are too high for the best performance.
As the Scanspeak graphs (into 2pi space, large open baffle) can be trusted,
The links blow contain the process for designing a correct crossover.

Its a fairly protracted process, as it should be for expensive drivers.
RJB has a tutorial on designing a 3-way c/o :
FRD tools example 2

2nd order L/R acoustic for bass/mid and the
mid/treble might be possible with those drivers.

rgds, sreten.

undefinition (see if nothing else, the excellent FAQs)
The Speaker Building Bible - Techtalk Speaker Building, Audio, Video, and Electronics Customer Discussion Forum From Parts-Express.com
Zaph|Audio
Zaph|Audio - ZA5 Speaker Designs with ZA14W08 woofer and Vifa DQ25SC16-04 tweeter
FRD Consortium tools guide
http://web.archive.org/web/200909021...esigningXO.htm
RJB Audio Projects
http://web.archive.org/web/200909022...ve99/Spkrbldg/
Speaker Design Works
HTGuide Forum - A Guide to HTguide.com Completed Speaker Designs.
A Speaker project
DIY Loudspeaker Projects Troels Gravesen
Humble Homemade Hifi
Quarter Wavelength Loudspeaker Design
The Frugal-Horns Site -- High Performance, Low Cost DIY Horn Designs
Linkwitz Lab - Loudspeaker Design
Music and Design

Great free SPICE Emulator : SPICE-Based Analog Simulation Program - TINA-TI - TI Tool Folder