To crossover or not or ?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am looking at a woofer that is fairly flat and has a natural 12 db roll off at
5000 Hz (Silver Flute W20RC38-04)
I am looking at a ribbon tweeter that needs to be crossed over at at least4500 hz (according to manufacturer RT1CA HiVi)

If I run the woofer without a lowpass how then does one determine the highpass frequency and slope for the tweeter?
 
I believe you could try putting a "mock" 2nd order (12dB) LP filter at 5,000Hz in Passive Crossover Designer (hopefully it should do nothing to the FR response) and then put in the values for the tweeter. This way you should be able to mimick the LP filter.

Alternatively, you could try putting the woofer in PCD by itself with no crossover and just have one for the tweeter. I think this works, though I've never tried it. :)

Also, don't forget that you may need an L-Pad if the woofer is too efficient due to the lack of an LP filter.

You will need SPL Trace if you don't have any .frd or .zma files for Passive Crossover Designer.
 
Wouldn't taking 12dB/octave from a 12dB/octave slope mean it's 24dB/octave?

Unless, of course, the Passive Crossover Designer doesn't account for the rolloff of the woofer/tweeter.

Personally, I'd low pass the woofer - at the frequencies you're talking of, it would be starting to beam.
 
Administrator
Joined 2004
Paid Member
Second what Andy said, but if you want to try it, here is how.

Since your woofer has a 2nd order roll off at 5K, you need to be near that with the tweeter. A simple 2nd order high pass at ~5K is a start, but will it actually give you a 5K 2nd order acoustic roll off? To simulate you'll need to get the tweeter FR into software and throw filters at it until you find one that gives you a curve to match the top of the woofer. That should not be too hard.

But then the real world kicks in and you'll find that you probably need to spread the crossover frequencies a bit to make up for the driver offsets.

Do you have the software to do the sims?

EDIT: I just took a look at the HiVi driver and it's going to be very hard to get a 2nd order acoustical roll off at 5K. 3rd order is easy. 3.2uf and 0.65 mH or there abouts. Nice 3rd order roll off at 5K.
 
Last edited:
Second what Andy said, but if you want to try it, here is how.

Since your woofer has a 2nd order roll off at 5K, you need to be near that with the tweeter. A simple 2nd order high pass at ~5K is a start, but will it actually give you a 5K 2nd order acoustic roll off? To simulate you'll need to get the tweeter FR into software and throw filters at it until you find one that gives you a curve to match the top of the woofer. That should not be too hard.

But then the real world kicks in and you'll find that you probably need to spread the crossover frequencies a bit to make up for the driver offsets.

Do you have the software to do the sims?

EDIT: I just took a look at the HiVi driver and it's going to be very hard to get a 2nd order acoustical roll off at 5K. 3rd order is easy. 3.2uf and 0.65 mH or there abouts. Nice 3rd order roll off at 5K.


My son was doing some google searching and found this interesting speaker
DECWARE / Small High Efficiency MTM Towers MG944

What might the thoughts be on this??

I was thinking crossover at 5000 and it looks like this one is at 8000? :eek:
 
Administrator
Joined 2004
Paid Member
Interesting find. So someone is using these drivers and making it work.

Maybe the electrical high pass on the tweeter is 8K, but I don't see how the acoustical could be - too much gap with the woofer.
My 1st guess for a x-over for this combo would be the afore mentioned 3.2uF+0.65mH on the woofer along with about ~0.3mH on the woofer.
Not quite sure about the inductor on the woofer because I can't tell what the impedance is. But I think you'll need something on the woofer to get it to blend right with the tweet.
.
 
I wouln't run this woofer without a xover. It starts to break up at about 1KHz (you can see the bump in the SPL trace) from here up its off axis response will be getting worse and worse. By the time you get to 5KHz it will be a tight beam. Also the peak at approx 4KHz of about 5dB will make this speaker sound quite unpleasant on axis, this is because 4KHz is the point of maximum sensitivity for the human ear so you really don't want a peak here. The top end of woofers output also tend to have a large amounts of distortion so whilst the sweep shows a 5dB peak here there may also be harmonics from lower frequencies adding to this which will make it sound even brighter and harsher. I would cross this speaker over with a tweeter at no higher than 2.5KHz (lower if you can find a tweeter that extends down further) and try to suppress the peak at 4KHz as much as possible.

Some people like the forward sound you get from wide open woofers describing it a live sounding. If you like this then ignore what I have said above. I tend to like acurate, easy to listen to speakers and find any distortion gives me listening fatigue very quickly.

A good way to get an idea of the effects I am describing is to run a graphic equaliser on a song you know well and turn up the 4KHz region. Since not many people have graphics these days this used to be quite hard to experiment with however now all you need to do now is download iTunes and use the graphic equaliser on that (It's mysteriously under the view menu rather than tools).

If you want to simulate the response of your xover system on axis, you can do it by downloading speaker workshop and SPLTOOLS SPLTRACE. Use spltrace to copy the response from the data sheet into a file that speaker workshop can use. Then do the same for the tweeter you have chosen. Then you can design and simulate a xover. This is quite a long winded process, there are several tutorials on the web about it but you will have to search for them as I can't remember the sites.

Regards,
Andrew
 
Some people like the forward sound you get from wide open woofers describing it a live sounding. If you like this then ignore what I have said above.

This is probably the reason why i am happy with the earlier mentioned Morel, Audax Combo, infact when i was at a local dealer who sells Dali Speakers i found the speakers demoed there sounding dull and lifeless compared with what i am used too ;)

Actually my speakers have tricked me more than once, several times sounds in music or movies have made me checking the door for visitors, the phone for calls and surrounding rooms for intruders.
 
The crossover Steve uses in these is just a single cap with the tweeter run reverse polarity.

What would be the purpose of reversing polarity of tweeter?

Just another thought: what if the hump at 5000 ( I do not know how long the wavelength would be-distance from cone to basket???)
Might damping the basket with something tame the hump (and higher frequency too) extending the usable range of this woofer?/

The issue I have is no testing equipment, but I did find this-only a different tweeter is used.
ChadGray.info

But I guess which baffle would make a difference too....
 
Speaker design is always a compromise. Without a mid range, you are going to struggle to cover the whole range without one of the drivers breaking up. If you find a tweeter that is possible to Xover at 1KHz ( you would probably need very steep filters) I suspect it will break up at about 16-18KHz at the very highest. So I would accept a bit of break up in the response as the off axis will not have collapsed very much by this point. Pushing the tweeter to low can sound much worse than pushing the bass drive a bit to high, also pushing the tweeter too low risks destroying the tweeter either with too much power (Heat) or excessive excursion.

If you put a mid range in you have to accept much more complexity and cost as well as the problems with intergrating the phase responses from two Xover regions rather than one.

On complete aside as this thread is on the silver flute driver. You could cover much more range with an electrostatic, but that is really complicated, I believethe QUAD one uses a fresnel lense type solution, to try to control the polar response. Even then you have to xover into a sub woofer as electrostatics struggle to move large amounts of air. (I really don't have much experience with these other than an interesting conversation with Peter Walker at one AES conference, that I struggled to follow)

I think its all the possibilities that make speaker design so facinating. Unfortunately since the rescession forced me to give up the Audio Industry, I only get to tinker.

My experience of Morel Drive units is they are usually well controlled in frequency response and quite well damped as well. This makes them ideal for running with minimal filtering. With alot of filtering it would be easy to describe them as a bit dull but very accurate. Unfortunately I suspect the driver suggested here will not be nearly as good, however as someone has shown with the link above, some people clearly like them enough to buy a design based on running them full bandwidth. Its the peak at 4KHz-5Khz that would worry me but perhaps they have managed to minimise it using out of phase signal from the tweeter, (the phase cancelation tends to give the sound a slightly airy presentation almost like it has a bit of reverb in it) or perhaps some people just like that really forward sound.

Regards,
Andrew
 
In order to make a decision, you need to look at the distortion of each driver. I recon it's not a great idea to cross an 8 inch woofer with a planar tweeter to begin with. Your crossover is going to fall in the range where neither driver performs it's best. Judging form FR of Silver Flute it has a large brake up at around 5k that may need to be notched even with a second order filter around 2500hz. Planar tweeter will not perform well that low also. Flute also has some wigle woogling going through out 1 to 2k region which may indicate either motor shortages or nonlinear artifacts. Even if a on axis based FR crossover is designed, 8 inch woofer will beam starting around 1500hz and cause havoc in of axis response. Planar aren't exactly known for a great off axis response either. Not exactly match made in heaven.
My suggestion would be to switch to a 7 inch woofer and possibly a good silk dome with a low fs. I've worked with a smaller Silver Flute woofers and they are quite nice for the price.
 
Last edited:
On the other hand --- I've heard one of the Hi-Vi planars on top of the Beta-8 and it's a wonderful combo. Who'd have thunk it? Crossed at about 6K, IIRC. Just a small coil on the Beta. (HiVi RT8II, I think)

Mike,

I think you may be onto something here. I really believe that a crossover point should match the off-axis performance of both mid/woofer and tweeter.
"If" the Planar's off axis FR is attenuated at the approximate frequency that the mid/woofer starts beaming, it may not create the weird phase-iness that in many situations degrades the sound. In fact, there are many instances that a limited directivity speaker may be preferred, like a narrow room or one with a lot of hard surfaces. In this regard they wouldn't be the best for an auditorium, nightclub or even at a party in a large room, but as a personal speaker for one or two people seated together or for critical listening it may be just the ticket.
:cloud9:
Just an opinion, of course.

Best Regards,
TerryO
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.