Need Help on a x-over Cap question

[jnash]

Hello,

I am trying to upgrade some vintage Falcan Acoustic cross-overs for my T27 tweeter. The current circuit is an inductor across the input with a cap eitherside of the inductor, one in front and one behind on the positive rail.

+ -----||-------||------- +
|
e
I/p e (inductor) Kef T27
e
|
- ------------------------ -

The Falcan mod says to add a ".47mF p/e" cap in parallel across the two capacitors from the + i/p to the + of the tweeter and to reverse the T27 polarity connection.
I have tried this using a KIMBER KAP .47uF-450V CAPACITOR that has a red and a black lead (connected red to the input) but it made my T27 out of phase with the other drivers in the speaker.
red black lead
|-------||-------| <---- new KIMBER cap
| |
+ --|--||-------||---|---- -
|
e
i/p e (inductor) Kef T27 (reversed polarity)
e
|
- ------------------------ +

Questions: What is a "p/e" type capacitor? , and
What capacitor type should a try this again with?
What’s going on with the phase?

[sreten]

Originally posted by jnash
Hello,
I have tried this using a KIMBER KAP .47uF-450V CAPACITOR that has a red and a black lead (connected red to the input) but it made my T27 out of phase with the other drivers in the speaker.

So you reverse the tweeter connections ? and thats it ?

sreten.

[jnash]

Sorry the graphics didn’t turn out correctly.
No, what I did was cut the cap out and revert the T27 tweeter back the way it was to check that things were back to normal.
I choose not to just swap the leads on the tweeter since I assumed it would still be out of phase due a wrong cap type choice. Digging into the speaker cabinets every time I make a change is tricky due to their age.

Attached is the corrected circuit in a text file

[Jay]

I guess p/e stands for polyester (or mylar), a dielectric non-metalized version of MKT. In my experience, mylar is more suitable than MKP for very high frequency (in crossover or zobel).

I don't think (I don't know) that type of capacitors have such a (big) efffect on phase reversing (Cannot be 180 degree), so I guess it must be okay if you choose "better" capacitor such as paper or paper in oil. But that will be too expensive for the job.

I think it has been a standard that black lead is the output lead of hi-end capacitors. I don't think that reversing will reverse the signal phase to 180 degree either.

Your crossover is third order. Even without the 470n the phase behaviour is "unpredictable" or cannot be said totally in phase. 470n is very small but because the manual says you have to reverse the tweeter, I think you better use your ear to decide whether you want to reverse or not. Still, you won't have a perfect phase response. So keeping the 470n out is also an option

Just a thought.

[Bill Fitzpatrick]

I'm sure that the .47 cap addition and the reversing of phase are suggested for two different reasons. The .47 would be to bypass the questionable? existing caps.

The reversing of phase suggestion is not made because the .47 is installed, it is made because either someone thought it sounded better or because it measured better.

[sreten]

Originally posted by jnash
I choose not to just swap the leads on the tweeter since I assumed it would still be out of phase due a wrong cap type choice.

this doesn't make any sense to me.

The instructions are to fit the capacitor and reverse the leads.

The wrong type of capacitor won't change reversing the leads.

The wrong capacitor value is a wrong value.

The kimber capacitor if correct value should be fine.
The fact that the unit is out of phase shows it works,
as you also have to reverse the tweeter leads.

sreten.

[Hobbit13]

Don't know if this is a usefull post, but I had two speakers wich used the t27, a kef104 and a imf compact 2. Both had a fairly simple xover circuit. In the both of the circuit the plus of the the bass was linked to the min of the tweeter.
Phaseproblems did occur when you changed that.