Input Capacitor on First Pedal in Line - Needed?

[DIY Dood]

I need to note that I am "electronics lite" :-).

I understand that an input capacitor will block DC. But it also will block some bass frequencies.

SO... could it be eliminated on the input side of the first pedal coming off the guitar? Is there any DC to worry about that would cause hum or other issues? I'm thinking that it's all A/C off the guitar (assuming no internal boost circuits).

Enlighten me!

[Clint Eastwood]

You can use a jfet input stage where the jfet is biased with just a source resistor, then there is no dc at the input. search for  'Jfet guitar buffer schematic', or 'Tillman preamp'.
By the way, the input capacitor will not block bass frequencies of your guitar if it has a large enough value. And this value  can be determined for a given circuit.

[idy]

Are you concerned about an "always on" or buffered bypass pedal? Why not all the other pedals?
If the input cap is the usual value it will not affect guitar freqs, maybe not even bass. That is one of the differences between a "guitar" and "bass" pedals, lower cutoff.

The above reply tells it: there is one kind of preamp/buffer stage that requires no input cap, a JFET. (some tube amps are like that too...)

Other stages will not work correctly as your guitar will "load" the bias voltage for the active device (transistor or opamp) and throw things out of their ideal zone. Plus DC going up into your guitar will mean scratchy noise when you adjust volume pot.

That first cap's value, in order to allow useful guitar freqs through, varies according to impedance of the circuit. With high impedance inputs (like MOSFET) it can be less than 1nf. With Ge transistor it will need to be bigger, like 2.2uf (fuzz face) (more than a thousand times bigger). With an opamp, you can set the impedance yourself and choose the cap.

Buffers have high input impedance and can use smallish input caps.

And there are reasons for excluding subsonics. Acoustic guitars need that cutoff , in general unneeded bandwidth is the devil's playground.

[ElectricDruid]

I understand that an input capacitor will block DC. But it also will block some bass frequencies.

SO... could it be eliminated on the input side of the first pedal coming off the guitar? Is there any DC to worry about that would cause hum or other issues? I'm thinking that it's all A/C off the guitar (assuming no internal boost circuits).

You *could* do this, but it's based on a faulty assumption that an input capacitor "will block some bass frequencies".
Let's assume the input impedance of your first stage is 1M (typical op-amp buffer bias resistor value). An input capacitor of 22n will start to roll-off frequencies below 7Hz. Given that the lowest note on a guitar is about 80Hz, and the lowest note on Bass guitar is about 40Hz, you won't hear any loss of bass frequencies. And that's with a *22n* cap. Even a 2n2 gives a cutoff at 72Hz, below the lowest guitar note. Stick 100n if it makes you feel better but don't think that you'll be able to hear the difference.

Even if you drop the input impedance down by ten to 100K (pretty poor in today's world) you'd only have to increase the capacitor by the same margin - that's 220n. Still not a huge value. It's easy/simple/cheap to put 470n film caps on inputs if required.

There's really no need to *not* have a cap to block DC, and your circuit will thank you the time that you plug in something that *does* have a big DC offset on the output.

Basically,  I totally agree with what IDY said. You don't want to be boosting subsonics further down the line, so decide what the "lowest frequency of interest" is and make the cutoff there. Otherwise you're just *adding* rumble and noise.

HTH

[FiveseveN]

it also will block some bass frequencies.

What do you need 10 Hz for? You can set the cutoff as low as you want but our ears don't go all the way down.

in general unneeded bandwidth is the devil's playground.

Absolutely! Limiting bandwidth is a good and often necessary thing.

[GibsonGM]

I personally always use a cap. For the bass cut as well as some level of protection from a DC source getting in and screwing up the bias of a circuit etc.

If you switch up where the pedal is located in the chain, if no input cap, a leaky output cap from another pedal can mess things up for you.  Any source of DC NOT a signal can fry your circuit without one, too - tho admittedly, ending up w/DC power source on an input isn't likely, tho it COULD happen. 

Back to bass...when mixing recorded tracks, often you'll find that guitars are given a cutoff as high as 250Hz!!  To make room for real bass instruments. Raw tracks can sound very trebly, thin, until you hear them in the context of a band.  Even jamming, bringing down 100Hz and below gets rid of 'flubbiness' and boom; makes the guitar a lot tighter.    My opinion, YMMV, but it's a common one....so a mild cut due to presence of an input cap is mostly a welcome thing    

[antonis]

For an absolutely symmetrical dual supply, input cap could be eliminated..

[DIY Dood]

Thanks, all. Makes sense.

One followup question/thought that these replies triggered, from the other frequency direction: When you have an input cap and an input resistor hanging off that to ground, is that forming an R/C network that's filtering the other end - the highs?

Here's a simple circuit for reference: LPB-1



Is the combination of C1 and R2 acting as a high freq filter in addition to dealing with the DC voltage coming from the battery to the base of Q1?

[ElectricDruid]

No, that's the same situation you'd see with an op-amp. You've got a input capacitor, and then a resistance to ground, which makes a highpass filter, which will remove some lows.

If it's "CR" then it's highpass. If it's "RC" then it's lowpass. You just swap the two parts over.

[antonis]

Is the combination of C1 and R2 acting as a high freq filter ?

Actually, the High Pass (Low Cut) filter above is formed by C1 and the equivalent resistance (impedance) from its right leg to GND..
(which is the parallel combination of R1, R2 and h_FE times (re + R4).. - for reasons beyond the scope of present thread.. )

[DIY Dood]

So in this case, is R2 there as part of a voltage divider (combined with R1) to the base of Q1 and it's accidentally acting as a CR high pass filter?  Or is it there designed as a high pass filter and not doing voltage dividing work at all? Or designed in as both a voltage divider and a high pass filter in one?

[antonis]

R1 & R2 are there for Q1 bias..
(Base at about 800mV -> Emitter at about 200mV -> Emitter current = Collector current about 500μA -> Collector voltage at about 4V, give or take..)
Their values have nothing to do (intentionally) with C1 but with bias configuration stiffness and stage input impedance..

Given that, C1 serves both for input DC blocking capacitor and capacitive element of CR HP filter..
(of 57Hz corner frequency, considering about 500μA Collector current and h_FE about 200..)

[GibsonGM]

It would seem to be more that you need to have R1 & R2 there for bias in this configuration, but there is a consequence - they interact with C1 'just because they are there', due to physics - so designers have learned to manipulate this "phenomena" to their advantage and use their presence to filter out lower frequencies that aren't desired.  There are many such cases of needed components acting in other ways that are taken advantage of, such as coupling caps in amps...

If that effect was not wanted, all one needs to do is use a big input cap and 'all' frequencies will pass.

That is just about the case here, as Antonis points out, the corner frequency is 57Hz, which is below any real useable frequency that a guitar produces...so 'everything' is passing (for all practical purposes).     

Having a cutoff at least a bit above 50/60 Hz (line frequency, pick your country's) is good, because it rejects power line hum from less than perfectly filtered power supplies...~100Hz is a good compromise, because the slope of the cutoff isn't great and you'd like to be 'down some decibels' before you encounter that 50/60Hz region - AND 100Hz on guitar is 'flubby', ha ha.

[antonis]

With all the respect to your Highness..
Hum is caused by ripple (dips) of double the mains frequency, Sir..
(full-wave rectification..)

[FSFX]

With all the respect to your Highness..
Hum is caused by ripple (dips) of double the mains frequency, Sir..
(full-wave rectification..)

With all respect to you too
The mains power supply ripple from a linear power supply is usually double the mains frequency as it is likely too have a full wave rectifier.
However, inductively coupled hum induced in magnetic pickups and so present on the input to a pedal will be at the mains frequency itself.

[antonis]

Of course but I just answered to Sir Mike's focus on particular cause of hum ..
>it rejects power line hum from less than perfectly filtered power supplies<

[FSFX]

>it rejects power line hum from less than perfectly filtered power supplies<

But I thought the subject here was the input coupling capacitor, not PSRR and power filtering.

[GibsonGM]

Yes, yes....if you've rectified, you've doubled the frequency.  Or, you could be standing in a bar, and picking up line freq. thru induction via your pickups

You caught me - most of my thinking in these terms goes back to tubes.  Heater supply rejection    

The best thing is that if you set your cutoff to reject 50/60Hz, then you are certainly more than safe at 100/120Hz!!   And again, not a lot of super-useful guitar down there. 

A thought occurs to me...what if you had line frequency mixing in like a clean signal, superimposed upon the rectified and doubled ripple?       Just kidding.