Quick filter Q

[DIY Bass]

Hi all, I am working on a reverb pedal.  I play Bass and guitar, so I am thinking of a switchable high pass filter that I can engage when I am using bass, just to reduce some of the mushiness that you can get when you run bass through verb.  I am thinking of running the dry signal through unchanged, and just filtering the input to the wet signal.  I have (hopefully) attached a portion of the schematic just after the input buffer.  The input of this section is just as the dry signal is routed through to the dry/wet mixer at the output.  The output from this is part of the conditioned signal going into the A/D converter.  The only components here that are not part of the original schematic are the capacitor and switch.  Is this the right way to do it? It almost seems too simple.  I really just want for somebody who is a bit more knowledegable about filters than me to check that this really is a switchable HPF at about 80Hz.  Thank you

[PRR]

The filter is a mild 80Hz low-cut (hi-pass), yes.

There's no power or bias connections shown, so I don't think it is ready to build.

It has high-ish hiss and low input impedance, and it inverts. Consider a passive low-cut and a unity-gain follower. This can also lead to a larger R which leads to a smaller C which might save you a penny.

[DIY Bass]

There are power and bias connection, I just didn't show them to simplify this diagram.  I am using Alesis chips, and the A/D converter needs inverted and non-inverted inputs.  From memory this op amp stage is the first part of the datasheet circuit.  The output goes to the converter, and also to another inverting op amp which goes to the other A/D converter input.  Would I get less hiss from a low cut filter?  I don't like the sound of high-ish hiss

[DIY Bass]

OK, looked at passive filters and that looks just as easy to implement.  I have one (hopefully) final question to make sure I understand.  The filters circuits I have seen are referenced to ground.  In the case of a stompbox, is it still referenced to ground, or to the bias supply voltage?

[ElectricDruid]

The thing that strikes me about that circuit is that the switch switches between "DC coupled" and "AC coupled". This is actually something you sometimes see on modular synth circuits where you might want to process DC levels instead of audio, but here it just seems like it is likely to cause thumps since there is almost certainly some DC level difference. I'd be inclined to add a second cap to the other path that allowed *all* audio to pass (e.g. big enough to have no noticeable HP effect) and also to make sure that the caps had drain resistors to prevent voltage building up on them, like you see on the input of many pedals.

HTH,
Tom

[garcho]

n the case of a stompbox, is it still referenced to ground, or to the bias supply voltage?

If it's part of the op amp feedback circuitry you'll want "ground" to be at the bias voltage. Roughly half-supply (think audio AC signal "riding" on DC). Depending on the op amp you might not always choose exactly the halfway point between the rails but it's usually the right spot and/or good enough.

[garcho]

Would I get less hiss from a low cut filter?  I don't like the sound of high-ish hiss

Outside of resonance boosting sibilance, it won't be the type of filter itself that causes hiss.
What PRR was referring to with input impedance is something all pedal makers deal with. Op amps theoretically "take in" zero current. Imagine they have a voltage sensor and the only data they work with is voltage data. You want to provide that op amp sensor with ALL the voltage data that comes coursing from your pickups into your guitar pedal. Often times, due to complicated crap like your cable inadvertently becoming a BPF, you lose some of that signal to ground along the way. Your circuit has a BJT op amp and a 10kOhm resistor which do not provide enough protection against losing some of your voltage data. NE5532 work great for signals in mixing consoles (where they have been famously used by Neve et al.) but generally, for guitar pedals, using JFET op amps is a better idea to combat hiss and voltage data loss.

[DIY Bass]

Looks like I should provide more of the original circuit and that way I should be able to get better help I guess.  Here is the full input circuit up until the A/D converter.  The line disappearing to the bottom goes to the dry end of the wet/dry mix pot.  I was hoping to add in a switchable high pass filter just to cut bass from about 80 Hz.  I guess maybe it could be an option to send the whole lot into the reverb and cut the bass at the output too?  Anyway, what is the best approach, given that this circuit has been built by a few people and I know it works, so I don't want to alter it too much?  I can use different op amps if that would help.  I do have a small stock of OPA2134 in the shed.

[DIY Bass]

Bump, because I'm still about confused about what I need to do.

[PRR]

The double inversion is not needed. Save a half-chip.


Do the inputs of the AL1101 self-bias? I am more used to seeing the opamps direct coupled to the ADC to set a DC bias at the ADC input.

[DIY Bass]

According to the datasheet, the AL1101 does indeed have on-chip biasing set to half the power supply voltage.  As far as I know, this circuit is almost a direct copy of an Alesis Picoverb, and that is how they did the input (which I admit does not necessarily make it the best way).  The line that you marked with a ? goes to the "Dry" side of a Wet/Dry mix pot at the end of the circuit after all the digital processing is complete.  From my reading of the circuit there is an input buffer for impedance and buffering purposes, which also gives the buffered Dry output ready for mixing back in at the end, and then the double inversion to give the + and - inputs to the AL1101.  If I follow you, the 2 inversions are not strictly necessary.  That does make sense.  As there is already an unused half-chip on the whole circuit, that would eliminate an entire chip from the board, which could be useful. 

Where and how would be he best place then to add in a switchable HPF?

[PRR]

> switchable HPF?

Go back to the first 3-opamp plan, put the HPF between the buffer and the first inverter.

Is it essential? Unless you have subsonic rumble bigger than your signal, the processor can do a HPF digitally.

[DIY Bass]

I am intending to just use the effects built in to the processor so I have no way of altering the programming at all.  I am not up to that level of coding yet :-( and don't really want that level of complexity anyway.  The HPF is just to drop out low frequencies when plugging a bass into it, as I find full range reverb on bass can get muddy/mushy and I want the option to clean it up a bit in the low end.  Would I just then switch in a single capacitor as I first looked at, or would I be better to switch in an R/C network?

[garcho]

Maybe for the sake of simplicity, and for the fact that most electronic devices that break, most often break at the human user interface - switch , knob, whatever - just hardwire the HPF? I always cut those low frequencies for the wet signal too, I agree with you about the muddiness factor. But what instrument might you use with this pedal that you would ever WANT sub 80Hz in the wet reverb chain? Maybe I'm misunderstanding you again.

[ElectricDruid]

One further thing to mention is that it's quite likely that there is some highpass action in the digital reverb code already. This is because if you take an input which has even a little bit of DC bias and recirculate it, the output starts to drift off towards one rail or the other. The simplest way around this is to apply a highpass filter at the lowest frequency of interest to block the DC.
I haven't heard the Alesis chip you're playing with, so I don't know how it sounds on its own, but you never know, they might have thought of this already. Keith Barr was a smart guy, after all (understatement).

Tom


<edit>dumb spellchecker

[DIY Bass]

The main reason I was going to switch the filter was just in case it did introduce anything like noise etc to the signal, that way you could switch it out when there was no chance of there being sub-80Hz in the signal (such as with a guitar).  Electric Druid is right too, it may already be taken care of in the software.

[garcho]

what kind of board will this be? are you getting it fab'ed or perf or etch or something? if you're going the perf style route, you might as well breadboard the thing. there are fairly cheap SOIC to DIP converters or sockets. if you're gonna etch it or have someone fabricate it you could add a little blank "experimental" section tied in to where you would introduce the HPF if it turns out you need one. 

[DIY Bass]

That's what I decided to do.  I am intending to get a board fabbed, so I have made space on the board for a few components and a switch.  I will jumper over that part, but can always add the components in later if I need to, and I very much doubt that I'll ever use a switch.  If I decide I need a filter I'll just add it in and be done with it, just like you suggested.