LPF question

classic Classic list List threaded Threaded
21 messages Options
12
Reply | Threaded
Open this post in threaded view
|

LPF question

Alamondm
I was wondering if someone could tell me what type of LPF this is. Sallen 2nd order, inverting, non-inverting etc.. I’m looking to change the freq sweep and would like to run it through a lpf calculator but I’m unsure of what type it is.

Reply | Threaded
Open this post in threaded view
|

Re: LPF question

induction
Do you have a bigger or easier to read image? I can't read this one very easily.
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

Alamondm
I will try this one
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

induction
Looks like Voltage-Controlled Voltage Source (VCVS) low-pass filter (2nd order, non-inverting), which is a variation on a Sallen-Key filter with gain.
In case it's not obvious:
- the top left op-amp is an input buffer
- the bottom left op-amp is a buffer for the 1/2 rail reference voltage
- the op-amp on the right is a makeup gain stage
- the stuff between C2 and C6 is the filter

Here's some links:
https://en.wikipedia.org/wiki/Sallen%E2%80%93Key_topology
http://nicadd.niu.edu/~fortner/course/phys475/lect/p575_07b.pdf
https://www.electronics-tutorials.ws/filter/second-order-filters.html
https://electronics.stackexchange.com/questions/74877/gain-in-sallen-key-filter

Plenty more with a google search if you're interested.
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

Alamondm
Thank you for the info. This is some good reading. I’ll take in as much as you want to share.
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

Alamondm
I’m not sure if I have this figured right. An I correct in thinking the C50k potentiometer figures in to the filter formula. Would the filter be comprised of 3k3 + 50k = R1 53.3k with C3 making the first filter @ 298.6 Hz. Seeing as though the info / links above don’t show a 1 to 1 diagram with the schematic I have , I’m having a little trouble putting it all together. I’m all about learning though,
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

induction
You've got it mostly right. The part of the circuit between c2 and c6 in your schematic is actually 1:1 with the relevant circuits in the links if you appropriately add the impedances in series and in parallel.

I'll use uppercase letters for the formulae in the links, and lowercase letters for components in your schematic:
For the corner frequency calculation, we have fc = 1/(2*pi*sqrt(R1*R2*C1*C2))
Comparing to your schematic, we have R1 = r4+rv1, R2 = r5+rv2, C1=c3, C2=c4.

rv1 and rv2 both range from 0 to 50k, so fc ranges from 298.6 Hz when rv1=rv2=50k, to 4822.9 Hz when rv1=rv2=0. These are pretty low frequencies for a low-pass filter, so I'm guessing this circuit is for bass guitar or a preamp for upright bass or something similar.

The gain is given by Av = 1+R3/R4. Without c5, the gain would be frequency-independent, but the impedance of a capacitor is given by Zc = 1/(2*pi*f*C), which is frequency-dependent. Another name for frequency dependent gain is EQ, so we really should account for c5 in the frequency calculations.

This is really only a slight complication though because c5 is tiny. The impedance of a cap and resistor in parallel is |Z| = 1/sqrt(R^-2+(2*pi*f*C)^2), which is gives R3=10k at 20 Hz and 9.9k at 20 kHz. This gives a gain of 1.45 at all frequencies that humans can hear, which means we can ignore it for the purpose of calculating the corner frequency of the filter. (At f=160 kHz, the gain is about 1.3, and as the frequency approaches infinity, the gain slowly drops to 1. So c5 is probably just there to prevent oscillation at ultrasonic frequencies.)

As a final note, I'll point out that the corner frequency calculations are symmetric with respect to R1, R2, C1, and C2. So most likely, rv1 and rv2 should a dual-gang pot rather than two separate pots, since the effect of both pots would be identical.
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

Alamondm
Thank you so much for taking the time to explain that. I pretty much understand, except for how to do the math formulas. I will research that more. RV1 & 2 are a dual gang C50k. I was close . I had the lower freq correct but was using the wrong C for C2 . With the aid of online calculators I may know just enough to get myself underway. The schematic is from a LPF for bass. Im not to keen on the frequencies they have it specd at thus the reason for this topic. Thanks again for your help, 52 and still learning, love it.
Matt
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

induction
My pleasure. If you need help customizing it to a specific frequency sweep, let me know.
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

Alamondm
Let’s see if I have figured this out correctly.
If I leave R4 =3k3 + 0k ( min pot value) and change C3 to C3=2n2n,  cutoff freq = 21,992
If I leave R5= 3k3+50k (max pot value) and change C4 to 15n , cutoff freq = 199.06
Does this sound correct.
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

induction
This post was updated on .
If you keep both pots at 50k and change both c3 and c4 to 2n2, the corner frequency sweeps from 1.3 kHz to 21.9 kHz.

If you keep both pots at 50k and change both c3 and c4 to 15n, the corner frequency sweeps from 199 Hz to 3.2 kHz.
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

Alamondm
So you need to keep both C3 and C4 at the same value?
 I thought I could make them different to change the sweep. I have more reading to do I think.
The 50k is the dual pot value .
I was thinking that RV1, R4 and C3 made one part of the filter and RV2, R5 and C4 made the other part of the filter , I’m missing something. I will dig deeper.
Ultimately I’m trying to get a 250hz to 22k sweep. I really need to learn this.
Thanks again,
Matt
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

induction
You don't have to keep c3 and c4 the same value. They can be different from each other if you like. My corrections are for your math results, not your approach to the modification. Your frequency results do not match what you should get by changing only 1 cap, but do match what you get by changing both caps.

Let me be a little more explicit:

Matches your approach: If you keep both pots at 50k, keep c4 at 10n, and change c3 to 2n2, the corner frequency sweeps from 636.6 Hz to 10.3 kHz.
Matches your results: If you keep both pots at 50k and change both c3 and c4 to 2n2, the corner frequency sweeps from 1.3 kHz to 21.9 kHz.

Matches your approach: If you keep both pots at 50k, keep c3 at 10n and change c4 to 15n, the corner frequency sweeps from 243.8 Hz to 3.9 kHz.
Matches your results: If you keep both pots at 50k and change both c3 and c4 to 15n, the corner frequency sweeps from 199 Hz to 3.2 kHz.

Or maybe you intended this approach: If you keep both pots at 50k, change c3 to 2n2 and change c4 to 15n, the corner frequency sweeps from 519.8 Hz to 8.4 kHz.

I can give you values that will result in a sweep from 250 Hz to  22 kHz if you like. But if you want to figure it out for yourself that's totally understandable. Let me know.

FWIW, 22 kHz is pretty high for a bass (or even a guitar for that matter). Amplified guitars have frequency ranges from about 80 Hz to about 4.5 kHz. Guitar speakers generally can't reproduce frequencies outside this range. I don't know what frequency range bass guitar speakers can reproduce, but a 4 string bass guitar has fundamentals in the range of 40 Hz to 400 Hz, while the overtones go up to about 4 kHz. 6-string basses have overtones up to about 5kHz. So I'm guessing you could set the top of your sweep at about 5 kHz without losing anything. This would spread the useful range of corner frequencies across the full range of the dual-gang pot. Personally, I'd test it on a breadboard before soldering anything.
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

Alamondm
I’m in over my head here. I do not know how to calculate the formulas. I didn’t take that kinda of math, more of a geometry guy. I can somewhat understand them but not enough to calculate.
I checked the specs on my bass cabinet and it has a range of 35Hz to 14Khz.
My intention was to keep both pots at 50k, change c3 to 2n2 and change c4 to 15n but that doesn’t give me the results expected. I have been using an online passive RC LowPass filter calculator. So I’m only calculating 1 filter at a time.
I assumed that the R/C and potentiometer worked independent of each other but now I see that’s not the case. I give up and humbly ask what are the values for the 250Hz 22Khz sweep. Maybe if I know the answer I can work backwards and figure it out for the future. I’m an old construction guy trying to learn something new. I did google how to calculate formulas but all that got me was a bunch of Excel spreadsheet stuff.
Thanks again
Matt
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

induction
If you keep both pots at 50k, make r4=470, r5=680, c3=10n, and c4=15n, the corner frequency sweeps from 257 Hz to 23 kHz.

There are other cap and resistor values that would work, but I chose these because they get you pretty close to the specified frequencies with standard components that should be easy to find.

I don't think that you will be able to hear any difference after the corner frequency goes above 5k, though.
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

Alamondm
Thank you very much. This has been very educational. Hopefully I will be able to figure this out with some more time. You have given me a lot of food for thought and for that I deeply thank you. Knowledge is key. This project started as a LPF to use as a high cut for bass, I am a bass player and an avid DIYer. In your opinion what do you think would be the best sweep for this. 500Hz to 5k maybe? That would also be useful for guitar as well?
I have seen another LPF diy kit project that I could just buy and build but there is no learning involved in that. This kit advertises the sweep being from 250Hz to 22K so that is where I came up with my sweep numbers.
Thanks again,
Matt
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

Alamondm
I haven’t given up yet, let’s see if I get this one right, lol.
Leave pots at 50k
Make R4=1k, R5=1K, C3=10n, C4=15n , corner frequency sweep = 254Hz to 13Khz.
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

induction
Alamondm wrote
I haven’t given up yet, let’s see if I get this one right, lol.
Leave pots at 50k
Make R4=1k, R5=1K, C3=10n, C4=15n , corner frequency sweep = 254Hz to 13Khz.
You got it. Good job.

In your opinion what do you think would be the best sweep for this. 500Hz to 5k maybe? That would also be useful for guitar as well?
When I'm building a circuit, the way I answer these types of questions is to build the circuit on a breadboard and experiment until I get what I want.

If I had to guess, I'd think the original sweep was probably good enough for both guitar and bass. So I'd approach it like this:
1. Build the original circuit on a breadboard. This would take maybe 5 minutes assuming I have all the components, which I do in this case, since there's nothing exotic in this circuit and I have a pretty big supply of components. (Breadboards are cheap and re-usable, so I'd advise buying one if you don't already have one and you plan on making more circuits in the future.)
2. Then I'd play through it for a while and use my ears to decide whether I wanted something from it that it wasn't giving me.
3. Then I'd start modifying the sweep and maybe the gain and/or pot tapers to explore what's possible with the circuit. I'd keep notes on what I find. I'd do this step whether I was happy with the circuit in step 2 or not, since this is how you grow your intuition on circuit design. Trust me, this part is 1000x easier on a breadboard than on a soldered pcb or vero. Along the way, you can explore whether having the extra range on the frequency sweep gives you extra options you actually want, or if it just makes the top half of the frequency pot sweep mostly useless, which makes it harder to fine-tune that parameter. Convenience vs. control is one of that balances that each designer has to figure out, and can be a highly personal decision. When you're building a commercial unit, you have to impose your choices on the user, which results in different decisions than if you're building for yourself. In the latter case, you can just make it exactly how you want it.
4. Once you decide exactly what you want to build, use the breadboard circuit in the context you plan to use the final circuit in for a little while to see if you change your mind about anything. What sounds powerful in isolation can often sound muddy in a live band, etc.
5. Once you make your final design choices, build it on vero or pcb or whatever, box it up and start using it.

This is the process I follow and I find it very useful. Others have different approaches that work well for them, so I don't want you to think that this is the one true path or anything. It's just what works for me.

Finally, about your specific question of the frequency sweep: EQ for guitar or bass usually has the frequency range limited to what's useful for that instrument. If you crank up the 15 kHz band on a distorted electric guitar and you'll mostly just hear an increase in hiss, so many guitar-specific EQs won't even give you the option. The same filter on a mix of several instruments including drums, keyboards, etc. where there is actually useful information in the 15 kHz band will have a different result. It's likely that the LPF kit you saw with a corner frequency sweep up to 22 kHz would be best applied to a full band mix or a synthesizer or something. Limiting that range would almost certainly make it more useful for guitar or bass or any other instrument that doesn't have useful fundamentals or overtones above 5 kHz. Different tools for different applications. But you can't be sure what you want until you try it out. Figuring this out is what step 3 above is for.
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

Alamondm
Great info. Thank you for taking the time to educate me. I have saved all the info to a folder for easy reference as needed. I have learned a lot and also learned I have a lot more to learn, lol.
All the Best,
Matt
Reply | Threaded
Open this post in threaded view
|

Re: LPF question

induction
Sorry if I made it sound more complicated than it is.

I always recommend working things out on the breadboard before building because it makes everything 10x easier, IMO. But a lot of people go straight from the schematic to the layout to building the pedal. If you want to go that route, you totally can. For me, breadboarding is as much fun as building, and I find it much easier to customize the circuit before soldering anything, so I encourage others to do it too. I wanted to offer suggestions on how to think about how to make design choices and decisions when you build a pedal. But they're just suggestions, and your process is totally up to you.

I just don't want you to think that there is some huge barrier between your current skill level and building this circuit. You can totally do this, and you'll learn plenty in the process. Let us know if you have any more questions.
12