diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Multi-Way (http://www.diyaudio.com/forums/multi-way/)
-   -   New 3-way project - "The French Touch" (http://www.diyaudio.com/forums/multi-way/229495-new-3-way-project-french-touch.html)

yov 6th February 2013 09:24 PM

New 3-way project - "The French Touch"
 
5 Attachment(s)
Hi All,


I'm new to this forum and I have a new speaker project !

A few words about myself. I live in Paris, France. I'm 30, married and a child.
I'm an engineer and I like: creating stuff and listening to music...
This is what brought me to DIY audio.


Two years ago, I completed my first speaker project, with which I'm not really satisfied today... I want to start something better. It is a two-way speaker based on 2 Monacor SPH-176 woofers and a 1 Monacor DT-254 tweeter. I have a Sherwood PM-9805 Vinyle, Yamaha A-S700 and Cambridge Audio A500.

There are several problems with my current setup:
- Sound is too directional and doesn't fill the room, specially the high frequencies which are too close to the ground level. I need to sit down in the direction of the speaker to hear them loud.
- It sounds like the sound is lacking mid-range. The bass are blurry and the tweeter is ok but the mid-range is absent. The spectrum doesn't sound linear.
- Not enough power
- Don't look good

So, now I would like to have a 3-way speaker and to make it more fun, I decided to use only French loudspeakers. This is why I named my project "The French Touch":

- 1 tweeter: Audax TW025A28, 8 Ohms, 94dB, 70 WRMS
- 1 midrange: Davis Acoustics 16GKLV6M, 6.5", 8 Ohms, 95dB, 130 WRMS
- 1 sub: Atohm D300P04, 12", 4 Ohms, 91dB, 200W RMS

I have attached the plan of the speaker. Sorry for the metric system...

I would like to add two shelfs at the bottom of the speaker, in order for the tweeter to be higher to the ceiling. I don't know if it's a good idea. It doesn't look good... In any case I need an idea to put my speakers higher than just putting them on the floor.

I intend to use a first order passive filter with the following layout:

- Tweeter : 5kHz - 20kHz
- Midrange : 300Hz - 5kHz
- Sub : 20Hz - 300Hz

I have attached the values of the components for the filter.

I intend to post my progress on this new project, here.
Because I have very limited budget and time, I will have to distribute the building phase over at least 6 months...

PeteMcK 6th February 2013 10:20 PM

re: "Sorry for the metric system..." - don't be, only colonials who don't don't know how to have a proper revolution use anything else...

Re: your crossover, it's too simple, a kevlar cone on the mid & the tweeter are going to demand something more than a simple 1st order xover to sound good,
& you need to think about the relative sensitivities of the drivers - the mid & tweeter will need some padding: L pad calculator - attenuation dB damping impedance decibel loudspeaker speaker voltage divider - sengpielaudio Sengpiel Berlin

tweeter height - should usually be at ear level

tvrgeek 6th February 2013 10:21 PM

Usually, you try and get the tweeter at ear level when you are sitting down listening. So, stands. I would not integrate them. Just my choice. A 12 inch three way should be easy to make as a floor stander.

Being an engineer, I am sure you know textbook crossovers are darn near useless. Crossovers need to deal with the phase, position offset, comb effects from not being coaxial, cabinet edge diffraction, baffle step, efficiency differences, their inductance, mechanical rolloff, breakup modes, and the list goes on.

What background reading have you done before approaching this project?

Tid-bits: With a 300 crossover, you may be able to deal with baffle step with just level adjustments to the mid-range and tweeter. No network. But with first order crossovers, maybe not. Sometimes just shifting the woofer LP frequency down a little can pull it off.

I am really skeptical about the 5K first order on the mid. The tweeter will LOVE it. I know nothing about the Davis drivers, so I am just guessing. They may need come notch filters, and maybe a Zobel. Don't know until you see. I have not found any mid larger than 3 inches that can do 5K first order cleanly.

Why do you intend to use a tuned chamber for the mid? At 300 crossover, you are a couple octaves away from the resonance, so a sealed chamber that is "obviously" large enough would be fine.

It is current wisdom to put the ports on the rear of the chamber to minimize mid-range bleed ( reflections out the port. )

The rigidity of the box is important. One trick I liked was to make the mid-range chamber tie the front and back together. Brace and chamber all in one.

That should give you enough food for thought for a day.

yov 6th February 2013 11:04 PM

Quote:

Originally Posted by PeteMcK (Post 3358990)
Re: your crossover, it's too simple, a kevlar cone on the mid & the tweeter are going to demand something more than a simple 1st order xover to sound good,
& you need to think about the relative sensitivities of the drivers - the mid & tweeter will need some padding: L pad calculator - attenuation dB damping impedance decibel loudspeaker speaker voltage divider - sengpielaudio Sengpiel Berlin

Thanks !
In fact, I was hesitating about using this filter instead:
Visaton HW 3/130 NG - 8 Ohm
It includes this Lpad attenuation that you talked about for the mid and the tweeter.
But it is built for a 8 Ohm Sub. Mine is a 4 Ohms.
Perhaps I can add a 4 Ohms resistance in cascade with the sub ?
Is this filter powerful enough (180W) ?

If I switch to a second or third order, I would prefer to buy it rather than build it because the price difference increases with the component quantity...

Quote:

Originally Posted by PeteMcK (Post 3358990)
tweeter height - should usually be at ear level

I usually listen to music just standing or moving around in the room. I like changing all the time the track, listening to particular parts, etc. etc. This is why I would like to have the tweeter quite high.

yov 6th February 2013 11:42 PM

4 Attachment(s)
Quote:

Originally Posted by tvrgeek (Post 3358994)
Being an engineer, I am sure you know textbook crossovers are darn near useless. Crossovers need to deal with the phase, position offset, comb effects from not being coaxial, cabinet edge diffraction, baffle step, efficiency differences, their inductance, mechanical rolloff, breakup modes, and the list goes on.

What background reading have you done before approaching this project?

Let's say that I have a generic education in electrical engineering.
I know about filters but mainly from a mathematical point of view.
I have very limited experience in applying filters theory to practical cases.
In particular, I don't know how far I need to care for each of those criteria that you mentioned.
I have read a few articles/DIYtutorials on the internet. What I have seen very often in those sites is that finding a mathematical answer to the perfect filter for a given set of loudspeakers is almost impossible. Most of the people say that they make some tests using different filters...

Quote:

Originally Posted by tvrgeek (Post 3358994)

Tid-bits: With a 300 crossover, you may be able to deal with baffle step with just level adjustments to the mid-range and tweeter. No network. But with first order crossovers, maybe not. Sometimes just shifting the woofer LP frequency down a little can pull it off.

Ok, I didn't know about baffle diffraction step.
Will read that and come back later :

http://www.diyaudio.com/forums/multi...pensation.html
Baffle Step Compensation
Baffle Diffraction Step
bafflestep

The Sub can hardly go above 300Hz because at this point the impedance is going above 10 Ohms...
I could choose 200Hz and leave the mid in the tuned chamber...

Quote:

Originally Posted by tvrgeek (Post 3358994)
I am really skeptical about the 5K first order on the mid. The tweeter will LOVE it. I know nothing about the Davis drivers, so I am just guessing. They may need come notch filters, and maybe a Zobel. Don't know until you see. I have not found any mid larger than 3 inches that can do 5K first order cleanly.

I have put in attachment the parameters of the loudspeakers.


Quote:

Originally Posted by tvrgeek (Post 3358994)
Why do you intend to use a tuned chamber for the mid? At 300 crossover, you are a couple octaves away from the resonance, so a sealed chamber that is "obviously" large enough would be fine.

I tried to simulate the closed chamber for the mid in WinISD. The -3db frequency is 150Hz. Isn't it too close to the first order xover cut-off ?

Quote:

Originally Posted by tvrgeek (Post 3358994)
It is current wisdom to put the ports on the rear of the chamber to minimize mid-range bleed ( reflections out the port. )

Ok, didn't know that. I will make the change on my plan. Thanks.
Is it possible to put the ports on the left/right of the chamber ?

Quote:

Originally Posted by tvrgeek (Post 3358994)
The rigidity of the box is important. One trick I liked was to make the mid-range chamber tie the front and back together. Brace and chamber all in one.

Thanks ! :)
I wanted to keep my speaker as a single box.
It took me a while to find out a solution for two tuned chambers with the same width and completely different loudspeakers.

PeteMcK 7th February 2013 01:18 AM

a pre-built crossover won't match the characteristics of your drivers, for best sound, a crossover needs to be design specifically for the drivers you're using.
This is because the impedance values "8" &"4" ohms are the nominal impedance of the drivers - the actual impedance varies with frequency, so the xover needs to be designed for the actual impedance at the frequency the crossover is working at...

Re: the tweeter problem, another way to approach this is to use a tweeter with good wide dispersion characteristics e.g. SEAS 22TFF

Inductor 7th February 2013 02:49 AM

Quote:

Originally Posted by yov (Post 3359045)
Perhaps I can add a 4 Ohms resistance in cascade with the sub ?
Is this filter powerful enough (180W) ?

Quote:

Originally Posted by yov (Post 3359081)
Let's say that I have a generic education in electrical engineering.

You will need a ~180 W power resistor.
There are many engineers committing mistakes... ;)

yov 7th February 2013 09:50 AM

Quote:

Originally Posted by PeteMcK (Post 3359189)
A pre-built crossover won't match the characteristics of your drivers, for best sound, a crossover needs to be design specifically for the drivers you're using.
This is because the impedance values "8" &"4" ohms are the nominal impedance of the drivers - the actual impedance varies with frequency, so the xover needs to be designed for the actual impedance at the frequency the crossover is working at...

I agree. A filter will not behave the same on each speaker setup. It needs to be designed specifically.

In theory, we should integrate the actual impedance curve in the filter computation. But I think this is a very complex computation only possible to make through simulation software and difficult to double-check in theory... The phase also, I am not sure what is the perceptible limit in the temporal shift between the speakers. Basically this is why I don't want to use complex filters. Simple is always better, right ? :)

I think using the average impedance value of the speaker on the working frequency range should be a good start. If the cutoff is correctly chosen, on the working frequency range the impedance of the speaker should be stable enough and have a trustable average value :).


Quote:

Originally Posted by PeteMcK (Post 3359189)
Re: the tweeter problem, another way to approach this is to use a tweeter with good wide dispersion characteristics e.g. SEAS 22TFF

I would like to keep it with "made in France" loudspeakers :)
It is the spirit of my projet :)
How can I check the dispersion characteristics of my audax tweeter ? (except listening to it ?)

yov 7th February 2013 09:54 AM

Quote:

Originally Posted by Inductor (Post 3359272)
You will need a ~180 W power resistor.

Ok, but if I put a 4 Ohm resitor in cascade with my 4 Ohm loudspeaker, it will behave like an 8 Ohm loudspeaker from a filter point of view. But I will loose half the power of my amplifier in the resistor !


Quote:

Originally Posted by Inductor (Post 3359272)
There are many engineers committing mistakes... ;)

Yes, but very few accept it ;)

yov 7th February 2013 01:01 PM

Quote:

Originally Posted by tvrgeek (Post 3358994)
Tid-bits: With a 300 crossover, you may be able to deal with baffle step with just level adjustments to the mid-range and tweeter. No network. But with first order crossovers, maybe not. Sometimes just shifting the woofer LP frequency down a little can pull it off.

Ok, I read about baffle step and given my baffle width, mine should occur at 115/Wb = 115/0.373 = 308 Hz with 3 to 6 dB loss in the sub.

My sub is already 3dB less sensitive than the mid and the tweeter. In the end, taking into account the bafle step I have a difference of at least 6dB between my Sub and my mid/tweeter. Happily this bafle step falls exactly at the cutoff frequency of my xover. If I want to attenuate my mid and tweeter I can add a resistor in parallel or in cascade. For example an 8 Ohm resistor in cascade to attenuate a 6dB.

But it's frustrating to add a pure resistance to my mid and tweeter !! They


All times are GMT. The time now is 08:08 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2