Hi guys,
I am trying to build a state-variable audio filter and it simply just doesnt work as it should, by changing R3 and R4 it suppose to change the cut off frequency but I've simulated and built it on the breadboard they both seems to get the same result.

Bandpass don't appears to work as bandpass and highpass don't work as highpass, well hipass actually have the same response as bandpass :?

Something is just not right and I do not know what it is, can somebody help me!!!

Thanks in advance!

Attached Images

	circuit_copy.jpg (70.6 KB, 327 views)
	sim_copy.jpg (74.1 KB, 327 views)

 

When you start with LP feeding into a HP will give BP. It appears that you have it the wrong way around?

 

Well, that's what it appears to be on the book Active-Filter Cookbook I have and also here on maxim-ic's
web site

I don't understand how IC1 can be a hi pass filter when it is just an non-inverting op amp configuration, I am currently very lost on this right now...... ops:

 

Oh, the answer is, the summing node. Sorry, I only looked at what stage was going into what without analyzing properly.

 

?

Sorry, please elaborate....

Do you mean I need to sum the HP, LP and BP together into the same output?

Also, by changing R3 and R4 am I simutaneously changing the LP and HP cut off point as well as the BP 3db cut off point?

Thanks!

 

Before you go any further, you have the polarities of the inputs +/- mixed up. That is your problem :!:

 

THANK YOU for pointing that out!!!!!

Will go back and do some more simulation.....

 

That is what had me fooled, I did not recognize it as the classical "universal" active filter.

Advantage of this is 2nd order response for LP and HP but difficult to obtain flat respone from BP with values selected for high Q

 

Hi,
I have corrected the circuit but still the simulation response just dont look right.........sigh........

 

This is how I remember this setup. The first stage is somewhat different to yours. (Ignore my values as I just used the program's default drawing values)

Attached Images

filter_141.gif (8.6 KB, 316 views)

 

It works in SwitcherCAD III. See below.
Edit:
Oops! I didn't notice that R12 (R5 in my schematic) is 20k. I tried that - it changes the response slightly, but still works. The One's configuration gives the same amplitude curves, but with different phase relationships.

Attached Images

	statevar_sch.gif (5.9 KB, 303 views)
	statevar_wav.gif (12.2 KB, 303 views)

 

Variation in R5 should alter the Q of the circuit

 

Hey TheOne,
Did your circuit simulate correctly on your PSpice?

 

I have simulated it in the past, but not recently. I know it works as also established by Ron H in his post.

 

Right, thanks.

I am just trying to find out if Orcad Pspice student version is not able to simulate correctly or it's something that I am not doing right .......... Can someone that uses Orcad student version confirm this?

Thanks in advance!!