I did changed the TL072 (wich has a bandwidth of 4 Mhz)
to a LF357 ( wich has a bandwidth of 20 Mhz) I changed all of them, but nithing happend the result is the same !!

 

Hello,

When you look closely at the curve of the filter, the lowpass does not come from the 0 dB.
The software will calculate a higher frequency as expected due to this offset.

Greetings,
Bertus

 

so do you have any suggestion what to change I am using simetrix software as you know

I zoomed the output for you

 

Hello,

I think the two frequencies of 23 kHz and 53 Khz are to close together to get a 0 dB point in between.
You could try to change it to chebychev chacteristics in stead of butterworth.
I do not know the software you are using.

Greetings,
Bertus

 

Ok thx a lot I'll try the chebychev

and a bout the software its free and nice easy software

you can downlade it if you like from this site

http://www.simetrix.co.uk/site/downloads/download.htm

 

Your filters are only second-order so their slopes are not sharp. The 53kHz lowpass also cuts 23khz a little. The 23kHz highpass also cuts 53kHz a little.
Their frequencies are too close together and with the not-sharp slopes then they affect each other.

 

what a bout if we use state variable filter (beasd on three filter ?

 

Hello,

Do you mean like on this page of the wiki ?
http://en.wikipedia.org/wiki/State_variable_filter

Greetings,
Bertus

 

kind off but there is better inf see my attchment

 

Hello,

OK, you could try a Q of about 1.3

Q = fc/bandwidth

Greetings,
Bertus

 

To work RT what value do you sugest for CT ?

I have tryed 2n but it didnt worked

 

Hello,

I would try 1 nF for Ct.

Greetings,
Bertus