Soup Ceramique MK-ll

[Cousin Billy]

Hello all

I have a few questions about Tony G's new crossover.

He uses two inductors in the midrange, both 680uH, seperated by a cap of 6.8uF. Is there a benefit to seperating the inductor values?, or is this a cost saving exercise?.

There are 100K ohm resistors in use across the caps for the tweeter. What is the benefit of values this high?.

There is a 1000 ohm resistor in parallel to the tweeter. Again, why a value so high?.

Cousin

[AllenB]

Originally Posted by Cousin Billy

He uses two inductors in the midrange, both 680uH, seperated by a cap of 6.8uF. Is there a benefit to seperating the inductor values?

No, I'd expect the same result with them together.

There are 100K ohm resistors in use across the caps for the tweeter. What is the benefit of values this high?.

There is a 1000 ohm resistor in parallel to the tweeter. Again, why a value so high?.

Cousin

It appears as if he might have used these components in crossover optimiser software, and the optimiser rejected them (didn't need them) by setting them to an out of range value. If this were true then I'd have deleted them.

[crazyhub]

I guess there is a purpose behind the use of components which do not modify the filter function; probably something like listening enhancement...

[crazyhub]

I asked Tony; here is what he said me (with his permission):

<<<<< R4 and R5 work as bypass resistors like the MKP1837 works as a bypass capacitor.
R6 lowers the tweeter level a fraction.
The two coils were a tryout due to this: http://www.isophon.de/pages/pdf/stp_12_10_151.pdf
(see Symmetrische Frequenzweiche)
I did a comparison between symmetrial and un-symmetrical. The symmetrical version seemed fraction more spatious, but I can't be sure.
I need to do some more experiments to make sure it's not my brain expecting to hear an improvement!>>>>>

[Cousin Billy]

Thank you Crazyhub, that is exactly the type of answer I was looking for. I went to the link, and will try to get the information translated from German.

At this point I would like to thank Tony, and to a slightly lesser extent, Troels. Had it not been for these two fine men, I would never have tried to build my own speakers. The speaker that impressed me the most was Tony's Humble Statement. If my pictures up load properly, you will see a resemblance.

Here is where I am at. Accuton BD25-6-034 Diamond tweeter.
Accuton C173-11-095e midrange (there is no cone breakup at higher frequencies).
Dual SEAS woofers.

I could not get a series crossover to sound fabulous. I will admit that there was that 'oneness' to the sound, but too much other stuff happening. I had a LEAP program output done, and am now just tweaking the crossover to my wifes liking.

First question; Is it simply a matter of breaking up one large midrange inductor into two smaller one's?. Again, I couldn't read the article in German.

Second question; Smoothing the impedance rise of the dual woofers in series at Fs requires a very large inductor. How do I calculate a single resistor value that will do almost the same thing?. And where exactly should this resistor go?.

I am sure I will have a lot more questions in the near future.

[AllenB]

Originally Posted by Cousin Billy

Smoothing the impedance rise of the dual woofers in series at Fs requires a very large inductor. How do I calculate a single resistor value that will do almost the same thing?. And where exactly should this resistor go?.

It may not be necessary. Do you have a particular reason for it as there may be other ways to do it.

[Cousin Billy]

Allen, I think you may be right, but I want to try something regardless.

I am using the Atma-Sphere MA-1 amps. They are OTL (output transformerless). They are a current source amp so do not mind high impedances. It has been said that a smoother impedance, especially at Fs, will yeild a tigher base. Since the speakers are closed, the base may be tight enough. The fact they are still on wheels means there will be some movement. This has left me with a slight feeling of woolyness during loud passages. Once the spikes are in, there may be no need to tighten anything.

The woofer crossover is made up of 18mH + 35uF. Testing shows very little cone breakup, so no other parts will be added outside of this possible impedance filter or resistor.

[AllenB]

Aside from the benefit of bass enhancement...

A current source emulates a high impedance. In effect, the frequency response (which may normally be flat) will begin to look like the impedance plot. As there is no logical connection between these plots, you will want to flatten the impedance (bass issues aside)

Your main issue should be the bump at the crossovers, otherwise you may end up with some harshness. After this, you can play with the bass.

Chose a value of resistor (say 20-50 ohms) where if you put it in parallel with the lowest impedance the woofers show (upper bass region) doesn't get too low for the amp. Then try other values.

I would do all my testing, measuring and crossover tweaking on a non current source amp. Then when happy, equalise all the impedance peaks and move on to your current sourced amp.

[crazyhub]

Originally Posted by Cousin Billy

First question; Is it simply a matter of breaking up one large midrange inductor into two smaller one's?. Yes Again, I couldn't read the article in German. There is not much to find out about the article, at least not in a scientific way.