First DIY !

gareth

Member
2008-01-31 3:38 pm
I am building my first speaker. It will be a two-way bass-reflex design using an Audax AP210Z0 8inch woofer and an Audax TW025M0 tweeter.
I want to build a simple first-order Butterworth crossover for these drivers. I was thinking of crossing over the drivers at 2000Hz. Looking at the manufacturers data for the tweeter it says that the resonant frequency is 976Hz but looking at the supplied graph there is a resonant peak at around 2700Hz of about 25-27ohms.
Does anyone know if this peak will cause problems and if my design is viable? If not are there any suggestions that can be given?

Thankyou
Gareth
 
Hi gareth,

Welcome and congratulations on your project. There are several issues here, and please take all comments you get as help and not criticism.

First, 2-way designs with an 8" woofer are rare and difficult to pull off - they're out there, but again they are few and far between. They usually require high order (steep) crossovers that can be quite complex, with many components, and thus expensive. This is because the operable frequency ranges of the drivers don't have a good overlap - either the tweeter has to operate at very low frequencies, or the woofer has to operate at very high, or both.

This brings me to my second observation about your design - a first order crossover in this case, especially for the tweeter, seems to be quite inadequate, so I must ask, how did you arrive at this design? Is it something you actually researched or modelled with software, or simply what you want to do? I'm afraid that if you build it as stated you will find the drivers distort and the sound quality will not be anywhere near its potential.

Now, I haven't looked up any data on these drivers - I'm opining on the fly here - so perhaps you could post frequency response curves, etc. for a more scientific assessment, but in lieu of that I have to agree with MaVo that you should consider building an established design for your first project - this is the most commonly given advice here for beginners, and most others too!

If you've already bought the drivers then you should, or have someone, model an appropriate crossover for them, if it's possible.
 
hermanv said:
Madisound will model your design and design a crossover (for a fee unfortunately). Then they will sell you a PCB and parts or even build the crossover for you.

As these things go, the fee is reasonable.


I've seen some of Madisound designed xovers. Those were not very careful designs that didn't even consider baffle loss/diffraction and phase tracking between drivers. You get what you pay for. I wouldn't use Madisound's xover design service.
 

vdi_nenna

Member
Paid Member
2000-10-10 7:27 pm
PA, USA
Try to use drivers that over lap in frequency by at least 1000hz.
If you intend to use a 2-way with an 8" driver, try to use a good 1.5 inch tweeter. I think 2000hz is a low xover point, but it all depends. It's not impossible. You will definiately need a high order slope crossover. Try for 18db in the tweeter section. You won't have to worry about phase issues. If crossingover low, you'll need to protect that tweeter. 18db will help protect from low freqs.

Driver selection and crossovers are everything. A good crossover is the difference between a speaker that makes sound and one that will be very pleasing to your ears. This takes a lot of tinkering and listening to get it right.

The low freq impedance peaks can be smoothed out by a zobel network. It's a capacitor and resistor (usually the impedance of the woofer) in parallel with the + and - connections of the woofer. I could be wrong, but you shouldn't worry as much with the impedance at the higher freq end of the woofer as you do at the lower freq end. Of couse, ideally you'd want it to be as smooth as possible.

Hope this helps some.

Vince
 

gareth

Member
2008-01-31 3:38 pm
Thanks for you reply. I have chosen a first order crossover simply because I felt that it would be easier to design for someone of limited knowledge. Maybe now I will think again.

Do you know of any (cheap/free) software that could help me?

I have the PDF's for the drivers I wish to use if they are of any use to you.

Thankyou
Gareth
 

gareth

Member
2008-01-31 3:38 pm
Thanks, I was beginning to think that someone might say that a three way would be the way forward. I appreciate your advice.

I will make a .jpg available containing the data for the TW)25M0, the same is available for the AP210Z0 if needed. I hope the quality of the .jpg is OK

Will the three way be much harder to design? I have a pair of Wharfdale 4" mid-bass units which I could incorporate or perhaps I should look for other drivers.

Thanks
Gareth
 

Attachments

  • audax-tw025m0-jpg.jpg
    audax-tw025m0-jpg.jpg
    64.1 KB · Views: 637
vdi_nenna said:
You won't have to worry about phase issues.

Yes, you'll have to. Acoustic phase and rolloffs are closely related to speakers' polar and power responses around the xover point.

Driver selection and crossovers are everything. A good crossover is the difference between a speaker that makes sound and one that will be very pleasing to your ears. This takes a lot of tinkering and listening to get it right.

True.

The low freq impedance peaks can be smoothed out by a zobel network.

A Zobel is used to flatten a driver's rising impedance as frequency increases, not to suppress an impedance peak in the middle though it can help a bit.
 
gareth said:
Do you know of any (cheap/free) software that could help me?

Read these:

http://www.rjbaudio.com/Audiofiles/FRDtools.html
http://www.geocities.com/woove99/Spkrbldg/DesigningXO.htm
http://audio.claub.net/software/jbagby.html


The tweeter spec sheet you posted looks strange. Near 3 kHz resonance peak for a 1" dome tweeter w/ a rear chamber? Something must be wrong. I think the 977 Hz Fs on the spec list should be the correct number. The impedance plot seems not correct.

P.S. you don't need to put multiple posts. Use "quote" instead.
 

vdi_nenna

Member
Paid Member
2000-10-10 7:27 pm
PA, USA
Jay, you didn't read my post correcty, but that's ok.

18 db as opposed 12 db, so he has better tweeter protection at a low crossover point and doesn't have to reverse the terminal conections.

I didn't say a Zobel will fix middle freq, I said low freq. The imped. at or near resonant freq of the driver.

See example charts-
http://www.partsexpress.com/pdf/295-366z.txt

http://www.partsexpress.com/pdf/296-420s.pdf
See the peak in the Aurum Cantus? That's what I'm talking about, not the rising at the higher freqs.

Vince
 
vdi_nenna said:
Jay, you didn't read my post correcty, but that's ok.

18 db as opposed 12 db, so he has better tweeter protection at a low crossover point and doesn't have to reverse the terminal conections.

It seems that you talked about simple binary phase (+/-). I talked about acoustic phase tracking in a real design. Without using CAD and careful in-box FR modeling, you can't pull exact 2nd or 3rd order slopes out of simple electrical order. Drivers' acoustic center offset also causes phase misalignment, too. That's what I mean.

I didn't say a Zobel will fix middle freq, I said low freq.

I don't understand what you mean here. A Zobel isn't for flattening peaks at low freqs, either.
 

vdi_nenna

Member
Paid Member
2000-10-10 7:27 pm
PA, USA

gareth

Member
2008-01-31 3:38 pm
Jay_WJ said:


Read these:

http://www.rjbaudio.com/Audiofiles/FRDtools.html
http://www.geocities.com/woove99/Spkrbldg/DesigningXO.htm
http://audio.claub.net/software/jbagby.html


The tweeter spec sheet you posted looks strange. Near 3 kHz resonance peak for a 1" dome tweeter w/ a rear chamber? Something must be wrong. I think the 977 Hz Fs on the spec list should be the correct number. The impedance plot seems not correct.

P.S. you don't need to put multiple posts. Use "quote" instead.

I was thinking the same thing in that the resonant peak near 3000Hz and that the quoted resonant frequency (fs) are not the same.

What would you suggest?

The bass driver, AP210Z0, seems to have a fairly flat response up to about 2000Hz and a sensitivity of 89dB. Maybe I will use a 2.5 - 3dB attenuator with the tweeter.
 
vdi_nenna said:
You expect a newbie to do this?

If not, he should learn how to. Relying simply on textbook ideas can fail miserably. I learned how to do it in a few months.


I don't see any lowered impedance at low frequencies, only flattened impedance curves that had a rising profile before a Zobel was applied. Perhaps, our languages are different.