12" woofer with ff85k

[wicked1]

Hi,
I'm finally getting back to this. I want to build a proper xover. (Right now I'm using a 500hz commercial xover.. 2nd order on top, 3rd on bottom. It's not quite great....)

so, I've got the ff85k on top, in a 4" PVC tube that extends out the back of my speaker cabinet, the tube is heavily stuffed. 12" woofer on bottom.

My xover ideas are leaving the 85k w/ no xover, and then put a 1st order on the woofer.
My main concern here is I'm using this on an 80ish watt amplifier, and the 85k's can't handle all that.. So maybe a filter on the ff85k's is a good idea.

And then the next idea is to put a 1st order on the 85k, and 2nd order on the woofer. My idea there being the 85k is naturally dropping off at the xover point, so a 1st order on it would about equal a 2nd order drop off in frequency, matching the 2nd order on the woofer. (is that correct?)

Then I'm also not sure exactly where to xover. I want to as low as possible. Does anyone have a general idea of where the frequency might be naturally dropping off on my 85k's in their stuffed PVC tubes?

Thanks for any advice.

 

[xrk971]

Hi,
I'm finally getting back to this. I want to build a proper xover. (Right now I'm using a 500hz commercial xover.. 2nd order on top, 3rd on bottom. It's not quite great....)

so, I've got the ff85k on top, in a 4" PVC tube that extends out the back of my speaker cabinet, the tube is heavily stuffed. 12" woofer on bottom.

My xover ideas are leaving the 85k w/ no xover, and then put a 1st order on the woofer.
My main concern here is I'm using this on an 80ish watt amplifier, and the 85k's can't handle all that.. So maybe a filter on the ff85k's is a good idea.

And then the next idea is to put a 1st order on the 85k, and 2nd order on the woofer. My idea there being the 85k is naturally dropping off at the xover point, so a 1st order on it would about equal a 2nd order drop off in frequency, matching the 2nd order on the woofer. (is that correct?)

Then I'm also not sure exactly where to xover. I want to as low as possible. Does anyone have a general idea of where the frequency might be naturally dropping off on my 85k's in their stuffed PVC tubes?

Thanks for any advice.

I don't think you can let the driver run full range to naturally fall-off - it will blow the driver as xmax is easily reached at relatively low voltages.

This is for a FF85WK - hope it is close enough...

If you put a heavily stuffed 4 in dia x 6 in long PVC pipe to form a sealed chamber (1.25 liters) behind the FF85wk, it has a slow fall-off of about -9dB/decade below about 200 Hz:

However, the cone displacement will exceed xmax of 0.35 mm below 267 Hz at 2.83V:

If you use a 300 Hz -12dB/oct high pass filter, here is the resulting response:

And the corresponding cone displacement is now controlled:

However, you will only achieve about 90 dB SPL before hitting xmax with a 300 Hz -12dB/oct HPF cross over.

EDIT: For the FF85K, the shapes are similar but you will reach xmax of 0.55mm at 6.2V rms with a 300 Hz HPF and this gives about 94.5 dB - much better.

 

[Greg B]

Hmm, there are lots of different ways you could do this. They all have pros and cons.

If it was my project, I'd go for a first order crossover at about 1kHz. Since the FF85 is in a stuffed pipe open to the back, the impedance is likely to be pretty flat. A textbook highpass should do a reasonable job. A 20uF cap should be about right for a highpass. For the woofer - perhaps a 2.5mH laminate core inductor if no zobel is used. If a zobel is used, about a 1- 1.5mH aircore would be a good first try.

It probably sounds like crossing the FF85 at 1k is wasting it, but it's not really. It's using its wide BW to your advantage, and is contributing to the sound well below the xover point.

 

[wicked1]

If you use a 300 Hz -12dB/oct high pass filter, here is the resulting response:

And the corresponding cone displacement is now controlled:

[/IMG]

Thanks a lot!
And I guess using a -6db high pass @ 300hz would not be enough to control the cone displacement?

 

[xrk971]

Probably not - depends how loud you want to listen.

 

[wicked1]

Thank you.

To GregB, my woofers don't quite reach to 1k well, so that won't work.

Then, if I'm going to stick to a 2nd order filter (which is what I've got on the 85k now) do you think it's worth moving the xover down from 500 to 300hz? 500 is just directly in the middle of the frequency range and seems like a bad place to 'mess w/' things. But... I'll save $100 or so if I leave it as it is .

 

[Greg B]

Thank you.

To GregB, my woofers don't quite reach to 1k well, so that won't work.

Hmm, most ostensibly pro woofers are designed to be used to 2k or so. Does it peter out before 1K, or just not sound good? Or do you just not like the idea of a 'high' crossover?

Which MCM is it? This one? 55-2962 or 55-2982?

Then, if I'm going to stick to a 2nd order filter (which is what I've got on the 85k now) do you think it's worth moving the xover down from 500 to 300hz? 500 is just directly in the middle of the frequency range and seems like a bad place to 'mess w/' things. But... I'll save $100 or so if I leave it as it is .

500Hz isn't in the "middle of things" anymore than 300hz is (middle C is 261.6Hz), but a lower frequency does mean a longer wavelength, which makes integration easier.

3rd order highpass could be done pretty low, if you are set on a low crossover. Worth a try. In fact, why not just try a bunch of different things out, and find out what sounds best?

You will need some sort of measurement device.

 

[murphythecat8]

Hmm, most ostensibly pro woofers are designed to be used to 2k or so. Does it peter out before 1K, or just not sound good? Or do you just not like the idea of a 'high' crossover?

Which MCM is it? This one? 55-2962 or 55-2982?



500Hz isn't in the "middle of things" anymore than 300hz is (middle C is 261.6Hz), but a lower frequency does mean a longer wavelength, which makes integration easier.

3rd order highpass could be done pretty low, if you are set on a low crossover. Worth a try. In fact, why not just try a bunch of different things out, and find out what sounds best?

You will need some sort of measurement device.

indeed. I would try the lowest xo point possible for better integration, the lower the better really.

 

[xrk971]

Note that 200 Hz 2nd order has been used by many folks with the FF85Wk and they don't seem to have any problems with xmax. I think 0.35 mm (WK) or 0.55mm (K) may be overly conservative.

 

[Scottmoose]

It's not overly conservative, it is what it is. Which is the point. There is no one definition of Xmax, manufacturers rarely state which they used, and the multiple different methods of putting a number to it almost invariably give different results if applied to the same drive unit. It isn't a measure of where the driver flies to bits, or the mechanical limit of travel (Xmech). Fostex (probably) are simply applying a physical measure to VC gap & winding length. YMMV. Historically, their units tend to have fairly benign characteristics beyond the stated figure, steadily compressing with distortion levels rising reasonably progressively, rather than anything more dramatic happening. I wouldn't beat the daylights out of them, but that's more a case of preserving longevity -no mechanical appliance likes being flogged within an inch of its life on a regular basis.

Since working on orchestral power-distribution most of the musical energy is < 500Hz, that used to be a favoured XO frequency, at least 2nd order & preferably higher to ensure excursion doesn't increase excessively < the XO point.

 

[planet10]

On our FF85k/FF85wk projects, XOs have been (except for Tysen) 1st order series or PLLXO between 240-350 Hz. Tysen had a 2nd order HP PLLXO at 333. By design we have tried 160 Hz, but midbass improved moving the XO up. At one point a 2nd order active at 100 Hz was tried, as well as just rolling off the woofer below 100 Hz (FF85 sats reached about 100 Hz on their own). These last 2 improved things but not near as well as other solutions,

The rated xMax is small, but the actual physical excursion capability is much higher.

Currently (happily) listening to uFonkenSET with FF85wKeN in my big system with no help. At some point soon i will drag the uFonkenSET with woofT (2 x Peerless 830870 per) into the system (PLLXO 1st order at 240 Hz)

dave

 

[planet10]

It depends on the type of music, but studies have show that typically half the energy in the music is <300 Hz ie (in the ~ borrom 4 octaves)

dave

 

[wicked1]

Which MCM is it? This one? 55-2962 or 55-2982?

It's the 2982 version.. But, a few days ago I ordered the sb acoustics someone recommended when I first started looking in to things last winter.. So, was planning on using them, and they only go to 1k.
The Madisound Speaker Store
But.. earlier today when I started getting in to the details of things, I noticed they're 6 ohms! the ff85k's are 8. My amp is a tube amplifier, so this could be an issue... The reflected load on my tubes will be changing at the xover frequency.

I always jump in to this speaker thing as an afterthought to building an amplifier, but I apparently need to spend more time thinking things through . Speakers are complex!

 

[planet10]

But.. earlier today when I started getting in to the details of things, I noticed they're 6 ohms! the ff85k's are 8. My amp is a tube amplifier, so this could be an issue... The reflected load on my tubes will be changing at the xover frequency.

Not worth worrying about. Just use the tap that sounds best. If you look at the impedance of any driver you will see that the nominal impedance is just that, nominal.

dave

 

[grindstone]

Speakers are complex!

Maybe everything is if we dive deep-enough?

Missed this the first time. Wicked, you have the right spirit and your curiosity has served-well. You are in a great spot and truly have Many directions that can turn-out as workable. So many, in-fact, that, I'm shutting my piehole except to offer encouragement. Your combo of empirical and research methods would indicate that the "Keep trying stuff" method (a compliment, incidentally) is nearly guaranteed to work.

"Blending" is a hairy bit, and there's some good advice here. That SB appears to be a fine part, it just may need some more zoot to make it filterable (dep on baff) or could be Fine (tm) dep on how you choose to use it and in what space. Anyone bold enough to go after analog synths will find this a short (if not imperceptible) detour.

Reading with interest. Keep it coming as time permits and thanks for the reports. Parting words: We really never do know until we know so continue to "just try stuff" as it may be possible. Fine "hearing education" comes from places that don't often simulate well-enough to build (let-alone that which is tough to sim at all even if we're Really smart).

 

[wicked1]

Your combo of empirical and research methods would indicate that the "Keep trying stuff" method (a compliment, incidentally) is nearly guaranteed to work.
Anyone bold enough to go after analog synths will find this a short (if not imperceptible) detour.

Thanks, I think I'm done w/ the research and modeling and I'm unfortunately to the experimenting part.. (unfortunately because I ordered more expensive inductors than I'll end up using . I should get a DSP to figure it out, and switch to passive later if I want.... live and learn.

And did I mention my synth in this thread, or are you stalking me? I did used to be really in to building synths and drum machines and things. Even just using (nevermind building) an analog synth is a great introduction to electronics and physics.

 

[murphythecat8]

It depends on the type of music, but studies have show that typically half the energy in the music is <300 Hz ie (in the ~ borrom 4 octaves)

dave

what is energy in music?

 

[wicked1]

Update:
SB Acoustics drivers aren't working.. They have beautiful deep bass, but are missing low/mid frequencies. Higher bass guitar notes and some male singing voices are not there. (xovered at 275hz. I tried various xover points, up to 500 and even no xover, and still, they are missing some midbass frequencies).
They'd be great for 3ways, or for subwoofers, though. Really super clear deep bass. (Probably going to put them up for sale, if anyone wants some sbacoustics drivers)

So, I played w/ the xover and the MCM drivers I've got... Moving the xover down to about 300hz and using good components improved things drastically!! They don't go as low as I'd like (which is why I tried the sbacoustics) but they do sound very good now. This setup w/ a subwoofer should keep me happy for a while.

I struggled w/ the xover a little. I've read to design them for the actual impedance at the xover frequency... (which is above their stated 8 or 6 ohms) But in practice I found designing them around the Re of the driver (which is lower than their stated 8 or 6 ohms) is giving me the expected results! I'm just using the manufacturers impedance curves for my calcs, so that could be the issue.
But whatever... That's why I got extra components to play with. It sounds good now .

 

[Greg B]

I had a feeling that might happen. IME, it's best to use helper woofers that are of the wideband mid/bass persuasion, as these are going to cover the meat of the lower midrange. Glad to hear you got good results with the MCM. I've had cheap MCM drivers outperform 'fancy' drivers more than once.


Crossovers - yeah, try different values until it gives the desired results.