Tannoy 611 Crossover Conundrum

[Sonusthree]

Hello,
These speakers (dual concentric 3 way) seem to have a pronounced response around the crossover point at ~ 2.5kHz which I would like to address somehow. I have added extra damping material to the inside of the cabinet which has helped greatly with the upper mid range "hash" but I now need to sort out the tweeters filter.

So....my next move, I think, is to raise the tweeters crossover point slightly to equalise this peak and reduce this irritating sibilance. The user manual states that the crossover is entirely first order at 400Hz and 2.5kHz. Is this the right approach? Am I missing something here?

In the Tweeter's filter what is the function of the inductor?

I have (hopefully) posted a crossover schematic which I have labelled with the information that I have so far. I don't have any values for the inductors and only the measured DC resistance for he individual drivers.

Could I just change the 3u3 cap for a 2u2 or would this be a drastic change?

Overall the speakers are rated at 6 Ohms.
Any help would be very much appreciated.

[sreten]

.

[Sonusthree]

Yes. I see your point.

[Sonusthree]

Does anyone know what the inductor L1 is for?
I'm desperately searching for an answer myself but getting nowhere.
I'm also trying to simulate similar circuits using TinaTi and Isis with no luck yet.
Any clues would be greatly appreciated.
S3.

[augerpro]

Total newb here but I think that L1 is there because it's 2nd order crossover.

[Sonusthree]

Thanks Augerpro. All advice is very welcome!
The manual says "First order low pass, First order high pass". I suppose that this is innacurate.

I'm still unsure wether or not to raise the high pass frequency, convert to a 1st order or add a notch filter??

Any ideas?

[PeteMcK]

It's only 1st order on the mid,

As a 1st step I'd try a Zobel on the mid.

Check out the on-line calculators here:
http://www.apicsllc.com/apics/Misc/filter2.html

That page will help you recalc xover values if you want to change them too.

Do you mave the model numbers of the drivers?

[Sonusthree]

\Cheers Pete.

I'm pretty sure that the harshness is coming from the 1" HF horn. The internal foam damping has helped greatly with the mid but obviously the HF compression driver is unaffected by this.
How would a zobel on the mid help in this case?

The drivers are Tannnoys own :
Model: 611
Year: 1991-3
Dual Concentric: 2033
Dual Concentric Size: 8"
High Frequency: 1662/2033
High Frequency Size: 25mm
Low Frequency: 2035
Low Frequency Size: 8"
Second Low Frequency:
Crossover Code: 1127

I'm going to investigate further tommorrow but the weakness appears to be either an unwanted resonance in the horn or it is crossed too low to cope adequately. My thinking here is that if I cross it a bit higher then it may be under less strain and therefore sound better.

I'm having great difficulty with the online crossover simulators because I don't know the values of the inductor L1 ("66" is handwritten on it.) and also the true impedance of the HF driver (4.4 Ohms D.C. resistance).
I'm not sure how to fill in the gaps?

[PeteMcK]

If you're sure it's the tweeter, a zobel on the mid won't help.

Using the calculator 'backwards' I get a value of 1.33mH for the inductor. (and a Z of around 10-11.5ohms).

Simplest thing to do is what you initially suggested, use a smaller cap, and see what happens (2u2 will get you up around 3KHz). Ideally you would need a change in inductor too, but only a small change downwards is indicated by the calculator.

[sreten]

Hi,

Sorry for deleting my post but it was incorrect.

I had assumed what you label as the mid was the 0.5 way
bass and what you label as bass, is in fact the bass mid unit.
In this case the c/o is 2nd order electrical, though its possible
c/o slopes approximate 1st order acoustic is the c/o region.

I think your diagram is possibly wrong, but could be right,
overdamped 2nd order on the 0.5 way unit, 1st order on
the bass/mid unit, 2nd order on the tweeter.

Still it appears not to be a true 3-way.

Adding Zobels is NOT the way to go, c/o values would then
be completely wrong, including phase at the c/o point.

Reducing the tweeters capacitor is worth trying.

/sreten.