Help with Sallen Key hpf

Started by ljudsystem, February 09, 2018, 10:32:56 AM

Previous topic - Next topic

ljudsystem

Hi guys

I'm trying to build a op amp based high pass filter fixed at around 70 hz to remove "mud" from my tone.

I've been reading up on op amp biasing but something is obviously wrong with my circuit since it distorts when i push it. The problem seems to be the 1M resistor and 4.7uf cap in the bias network. I'm not sure what they are supposed to do. When I removed them and just used a voltage divider I got no sound at all.

The first part is supposed to be a buffer (taken from the Klon buffer schematic) and the second part is a standard Sallen Key HPF.

Any help would be appreciated.


samhay

The 68k resistor connected to the (+) input of the second op-amp should go to your Vb point, rather than to ground.
Your Vb is the junction of the two 100k resistors make as 4.6V.
I'm a refugee of the great dropbox purge of '17.
Project details (schematics, layouts, etc) are slowly being added here: http://samdump.wordpress.com

ljudsystem


PRR

> second part is a standard Sallen Key HPF.

Audio swings both ways. "Standard" filters are often drawn using bipolar + and - power rails. When you have a single 9V rail, you must bias-up the opamps' DC point to about halfway, 4.5V of 9V.

As Samhay says, you got the first stage right, but didn't rig the second stage to idle at +4.5V. Since this is a popular question, I've taken the liberty of marking-up your plan; it may help others.



Guitar? Unless there is something very odd about your rig (or you record acoustic in a rumbly studio), 70Hz may have little effect. 150Hz 6dB/oct is very common in lead guitar rigs. Have a few 15nFd caps (or 33K resistors) on hand to find a happy place to shave bass.
  • SUPPORTER

ljudsystem

Thank, I get why it distorted. I connecting the 68K to Vb and now it doesn't distort.

The idea was to filter out the "sub harmonics", low E is 87 hz but i sometimes play drop D so i chose 70 hz. I usually do this on the computer with recorded guitars and bass. If I put a low pass filter on the guitar and listen to whats bellow 80-90 hz there's a bit of rumble and mud.

the 4.7uf keeps the 4.5 volts steady, right? but I'm not sure about the 1M resistor (the one with the questionmark) whats it's purpose? Should it be the same size as the 68k in the second stage?

Kipper4

Welcome ljudsystem.

Good question.

I'm hoping for a tractor analogy. It's been a while.

I'm still learning everyday.

Rich
Ma throats as dry as an overcooked kipper.


Smoke me a Kipper. I'll be back for breakfast.

Grey Paper.
http://www.aronnelson.com/DIYFiles/up/

PRR

The "1Meg" should be small compared to opamp input bias current, but large compared to guitar impedance.

68K there would dump all your guitar's top-ring (about 100K-250K impedance) into the 4.7uFd, sound would be dull. OK for a "Dull Box", but not if the label on the tin says "Bass Cut".

1meg is typically fine for guitar.

Opamps vary. '5532 has large input bias currents. A 1Meg here might cause a part-Volt offset. The intended 4.5V might go to 4.2V DC. That may be tolerable if output is DC-blocked and signal levels don't go toward 9Vpp. For other reasons the '5532 is not a best-choice to interface guitar. The popular and cheap TL072 is a good (not best, but not so it matters) hiss-match, and its input bias current is incredibly teeny. That "1Meg" could be 10Meg maybe 100Meg and not throw the bias far off. However there's no good reason to go way-high (and damp board would cause more trouble). 1Meg is a fine pick. If out, 680K or 2Meg, all the same.
  • SUPPORTER

Transmogrifox

When trying to mix guitar with other instruments, I consider 250 Hz to be the center of the "collision band" -- that is, where bass, keys and guitar all converge in a muddy swamp.

When mixing sound live, I often roll back the keys and guitar with a shelving filter starting at 250 Hz and try to let the bass carry the low end.

Either way a useful filter might be something like the Ibanez TS-9 style tone control tuned a little lower.  It's a 1rst-order filter (6 dB/Octave) and the cut-off frequency changes as you turn the knob.

In this case half the pot turn that does high-cut would not be useful, but grounding the one end of the pot so it goes from 0dB to something like +12 dB high boost could be useful.

I like knobs, but the truth of the matter is, once I find the sweet spot with the knob I stop messing with it.  The hard-wired Sallen-Key is probably great for what you're doing if you already know the selected cut-off happens where you need it.
trans·mog·ri·fy
tr.v. trans·mog·ri·fied, trans·mog·ri·fy·ing, trans·mog·ri·fies To change into a different shape or form, especially one that is fantastic or bizarre.

reddesert

Quote from: ljudsystem on February 09, 2018, 03:26:52 PM
the 4.7uf keeps the 4.5 volts steady, right? but I'm not sure about the 1M resistor (the one with the questionmark) whats it's purpose? Should it be the same size as the 68k in the second stage?

The 1 Meg resistor on the + input of the first op-amp connects the input to Vref (4.5 V), so that the op-amp is amplifying your AC signal oscillating about the 4.5 V reference voltage.  The 1 Meg has no effect on frequency response, because the R*C value is so large. It's just there to provide a bias voltage.

The 68k resistor is part of the filtering network. In this Sallen-Key filter there are essentially two RC high-pass filters each of 33nf and 68Kohm, referenced to Vref.