Filter expert needed...

Luke123 · 13th April 2005, 04:44 PM

I need a filter expert on this.
This is a fourth order 10 KHz lowpass filter made of a pair of 2nd-order sections.
There is an upward departure from target at -80dB level.
Is this due to excessed phase rotations in the opamps?
Can this be cured with frequency compensation?
If so, how do you apply this with S/K and MFB type of lowpass filters?
If you can also help me with a link, it will be appreciated!

Luke

richie00boy · 13th April 2005, 04:54 PM

What filter topology are you using? My Sallen-Key filters work fine.

Luke123 · 13th April 2005, 05:13 PM

Sallen-key or multi-loop feedback design...they have the same problem . Same problem with OP275, OPa213x, and others. All departs from perfect LP4 transfer functions.

Is your a dual 2nd-order cascade?
Tuned somewhere between 1 and 10KHz?
Did you measure it down to -100 dB?
I read something long time ago about this...don't remember where.

Luk123

Mr Evil · 13th April 2005, 05:31 PM

It's due to finite op-amp bandwidth. Use an op-amp with a high GBP to minimize the problem.

Bricolo · 13th April 2005, 09:44 PM

There has juwt been a nice article on the bob pease show about that
available on www.national.com

it's due to the opamp bandwidth, but optimal component values can give better results

Luke123 · 13th April 2005, 10:29 PM

Hey Broccolo!

Do you remember the title of the article?
Any cue available?

Luke123

Bricolo · 13th April 2005, 10:41 PM

on national's website, do a searh on "bob pease show"
I think you have to register an account to have access to it

and it's not an article per say, it's a show

video+sound, you need real player or windows media player

amplifierguru · 14th April 2005, 12:23 AM

What chips are you using - as has been said, ones with higher GBW are the go!

OPA627 is best! But you can use lesser chips if the alignment is low Q i.e. Butterworth or Bessel, Chebyshev give steeper cutoff but require more GBW and give a distortion spike just below cutoff frequency.

If it's for a quality audio use MFP will give lower distortion and best to use more sections of low Q depending on the application.

Hope this helps.

amplifierguru · 14th April 2005, 05:38 AM

Some more...

A rule of thumb is for MFP sections GBW should be > 100xGAINxfn (natural for section)

For Sallen-Key GBW should be > 100xGAINxQxQxQxfn so if Q is high it's very high.

Luke123 · 14th April 2005, 01:27 PM

Thanks for collaboration

When I started this thread, I just wanted to expose my problem in a very simple way. Bricolo pointed me out at the Bob Pease show link. (Thanks by the way.) But I have already seen it and this is what inspired me to write this thread actually. The picture is a screen capture from a Webench simulation. I did this because I don't have the original measurements of the specific problem. Exposing the thread this way was much more simple for me.

Amplifier Guru : The two LPF2 in cascade had a F-Q values of 1300Hz/1.3 and 1500Hz/1.35. The tried Opamps were OP275 and OPA2134. I am aware of your law for needed bandwidth versus frequency/Q/gain and these opamps does not infringe this law.
Anyway I remember having tried it with AD745 with the same results.
I'm sure it have something to see with a custom frequency compensation for slowing down phase rotation in VHF.
But how?

Luke123

13th April 2005, 04:54 PM	#2
richie00boy Did it Himself diyAudio Member Join Date: Nov 2003 Location: Gloucestershire, England, UK	What filter topology are you using? My Sallen-Key filters work fine. __________________ www.readresearch.co.uk my website for UK diy audio people - designs, PCBs, kits and more

13th April 2005, 05:31 PM	#4
Mr Evil diyAudio Member Join Date: Aug 2004 Location: Behind you	It's due to finite op-amp bandwidth. Use an op-amp with a high GBP to minimize the problem. __________________ https://mrevil.asvachin.eu/

13th April 2005, 09:44 PM	#5
Bricolo diyAudio Member Join Date: Nov 2002 Location: Grenoble, FR	There has juwt been a nice article on the bob pease show about that available on www.national.com it's due to the opamp bandwidth, but optimal component values can give better results __________________ Just remember: in theory there's no difference between theory and practice. But in practice it usually is quite a bit difference... Bob Pease

13th April 2005, 10:41 PM	#7
Bricolo diyAudio Member Join Date: Nov 2002 Location: Grenoble, FR	on national's website, do a searh on "bob pease show" I think you have to register an account to have access to it and it's not an article per say, it's a show video+sound, you need real player or windows media player __________________ Just remember: in theory there's no difference between theory and practice. But in practice it usually is quite a bit difference... Bob Pease

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

13th April 2005, 05:13 PM	#3
Luke123 diyAudio Member Join Date: Jan 2004 Location: Quebec	Sallen-key or multi-loop feedback design...they have the same problem . Same problem with OP275, OPa213x, and others. All departs from perfect LP4 transfer functions. Is your a dual 2nd-order cascade? Tuned somewhere between 1 and 10KHz? Did you measure it down to -100 dB? I read something long time ago about this...don't remember where. Luk123

13th April 2005, 10:29 PM	#6
Luke123 diyAudio Member Join Date: Jan 2004 Location: Quebec	Hey Broccolo! Do you remember the title of the article? Any cue available? Luke123

14th April 2005, 12:23 AM	#8
amplifierguru Account Disabled Join Date: Aug 2004 Location: Earth	What chips are you using - as has been said, ones with higher GBW are the go! OPA627 is best! But you can use lesser chips if the alignment is low Q i.e. Butterworth or Bessel, Chebyshev give steeper cutoff but require more GBW and give a distortion spike just below cutoff frequency. If it's for a quality audio use MFP will give lower distortion and best to use more sections of low Q depending on the application. Hope this helps.

14th April 2005, 05:38 AM	#9
amplifierguru Account Disabled Join Date: Aug 2004 Location: Earth	Some more... A rule of thumb is for MFP sections GBW should be > 100xGAINxfn (natural for section) For Sallen-Key GBW should be > 100xGAINxQxQxQxfn so if Q is high it's very high.

14th April 2005, 01:27 PM	#10
Luke123 diyAudio Member Join Date: Jan 2004 Location: Quebec	Thanks for collaboration When I started this thread, I just wanted to expose my problem in a very simple way. Bricolo pointed me out at the Bob Pease show link. (Thanks by the way.) But I have already seen it and this is what inspired me to write this thread actually. The picture is a screen capture from a Webench simulation. I did this because I don't have the original measurements of the specific problem. Exposing the thread this way was much more simple for me. Amplifier Guru : The two LPF2 in cascade had a F-Q values of 1300Hz/1.3 and 1500Hz/1.35. The tried Opamps were OP275 and OPA2134. I am aware of your law for needed bandwidth versus frequency/Q/gain and these opamps does not infringe this law. Anyway I remember having tried it with AD745 with the same results. I'm sure it have something to see with a custom frequency compensation for slowing down phase rotation in VHF. But how? Luke123

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Transformer expert needed.	nairod156	Solid State	4	13th January 2009 11:35 PM
3-way crossover expert needed	andrew01	Multi-Way	28	9th November 2007 11:53 AM
Is Lowpass filter needed in 6ch amp?	Wynand	Chip Amps	4	24th November 2005 04:34 PM
Help needed for active filter	k1jroth	Chip Amps	0	23rd February 2005 03:14 PM
TL expert needed	MadMax	Multi-Way	1	12th June 2003 03:48 AM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read