Is the input cap always needed

[wayfaerer]

Is the input DC blocking cap always needed? I'm making the following assumptions:

• The effect is for my personal use, not being released into the market—I don't have to worry about edge cases such as compatibility with some weird pedal with no output cap.
• The effect is guaranteed to be immediately after a guitar with passive pickups, or after another pedal with an output cap.
• The circuit does not rely on the DC blocking cap to remove bass.

This is something I've always wondered, but now I'm working on a really, really tiny circuit that will go directly into a 1/4" plug. I see that there's no DC blocking cap in Donald Tillman's FET preamp cable here: http://www.till.com/articles/PreampCable/. Just wondering if there's a downside to this.

Another way to phrase this: Where would the DC voltage come from that an input cap is supposed to filter out?

[aron]

AFAIK you don't need it with those assumptions.

[GibsonGM]

I guess if a guitar had active electronics, DC could leak from there. But for your purposes, knowing 'what's up' and all...as Aron said, no need for an input cap, nope.

  Many tube amps don't use them either. 

[willienillie]

Don't know what circuit you're talking about, but in many the input goes to the base or gate of a transistor, which often has a small DC voltage relative to ground.  In that case, no input cap would put DC across your guitar pickup and result in a scratchy volume pot, if plugged directly in.

[wayfaerer]

Thanks for the responses.

Don't know what circuit you're talking about, but in many the input goes to the base or gate of a transistor, which often has a small DC voltage relative to ground.  In that case, no input cap would put DC across your guitar pickup and result in a scratchy volume pot, if plugged directly in.

Interesting. Under normal circumstances, would it hurt anything?

I was going to use this buffer from the AMZ Basic Buffers page: http://www.muzique.com/images/buff3.gif, except without the 0.1uF cap, and instead with a 20K resistor in series with the incoming signal as Tillman does here: http://www.till.com/articles/PreampCable/images/plugschematic.gif. He says this is "for protection from static discharge as the plug will spend some time waving around in the breeze". My transistor expertise is non-existent... do you think what you described might happen in this case, or is it a case of "try it and see"? (I'd prefer to avoid try it and see because I'm planning to have a PCB made).

[PRR]

> this buffer from the AMZ Basic Buffers page:

No link to the page?

Anyway I assume Vr is a small positive voltage for better signal handling in the source follower. Yes, it will also work if "Vr" is zero (DC-shorted by guitar), for very small signals. The way AMZ did it is more generally better.

If it is the cost of the capacitor: use 10Meg gate resistor and 500pFd coupling (DC-block) cap.

[amptramp]

The difference between the AMZ and Tillman preamps is the AMZ one goes to a bias voltage that will appear on the input.  The Tillman bias is ground.  The Tillman does not need an input cap.  If you don't have an input cap with the AMZ buffer, changing the volume control on the guitar will change the bias on the preamp which may lead to interesting but probably unusable effects.  You need the capacitor because otherwise the voltage on the gate will vary too much and it is likely that the FET would be driven into cutoff.

[Mark Hammer]

Back in the day, CRaig Anderton used to design his projects with the option for an input with DC blocking, and either a second input or simply a toggle for situations where you knew for a fact there would be no incoming DC.

[R.G.]

One thing that I've picked up on is that my assumptions will usually fail when I'm in some situation where I can't do anything about them failing, or long enough later that I've forgotten making the assumptions.

Do you always have to have an input cap? No. But my best advice is to not fence yourself in so that you CAN'T add one and a pulldown in later when the assumptions fail.

[Rob Strand]

Is the input DC blocking cap always needed? I'm making the following assumptions:

You can do what you want in the comfort of your own home.  If you have problems down the track you know where to look.

Another way to phrase this: Where would the DC voltage come from that an input cap is supposed to filter out?

Like most things: evil happens as a result of two wrongs.    Someone designs a pedal with no input caps.  Another guy designs a pedal with no output cap.   When the two meet there is trouble.

[PRR]

http://www.muzique.com/lab/buffers.htm

Reading:

  "Probably the easiest buffer is the basic jfet common drain amplifier. ....This configuration has the least number of parts but is limited in that if the input voltage exceeds the gate-source forward voltage plus the bias voltage at the source, the signal will be clipped.
  "The basic jfet buffer shown in the last paragraph may be improved upon by connecting the gate resistor to a bias voltage instead of to ground. ....The Vr point on this circuit connects to a bias voltage source as shown in the first paragraph...."

While he does not say this, if you draw-out the buffer with Vref supply it is clear that you need the gate cap; else stuff on the input jack can short-out the Vref that you lovingly put in to improve performance.

[wayfaerer]

Very helpful replies... thank you! I can see now that the input cap serves a greater purpose than what I'd originally thought.

While he does not say this, if you draw-out the buffer with Vref supply it is clear that you need the gate cap; else stuff on the input jack can short-out the Vref that you lovingly put in to improve performance.

Can you clarify in what situation the Vref could get shorted without it? As a semi-noob, I'm not seeing it. (Schematic again, for reference: http://www.muzique.com/images/buff3.gif).

[Mark Hammer]

Do you always have to have an input cap? No. But my best advice is to not fence yourself in so that you CAN'T add one and a pulldown in later when the assumptions fail.

Which is precisely why pretty much every pedal manufacturer includes an input DC-blocking cap.

[ElectricDruid]

Do you always have to have an input cap? No. But my best advice is to not fence yourself in so that you CAN'T add one and a pulldown in later when the assumptions fail.

Which is precisely why pretty much every pedal manufacturer includes an input DC-blocking cap.

Yep! If you're designing pedals and someone buys yours and plugs it in and has problems, they're going to blame your new pedal, not the rubbishy old thing they've had on the board for ages that has no output cap and about 0.5V DC going out!
So you adopt "defensive driving" as a strategy and make sure that your box works as well as it can more or less whatever it's plugged into. That mandates caps at both ends, pull-downs, the works...even if that is mostly repeated by the next box downstream.

[PRR]

[wayfaerer]

Thanks for the diagram... I get what you mean now!

[Eb7+9]

a generally capacitor-less screamer idea: dual rail adapted, a fully dc-coupled TS re-design, using NPN/PNP buffers towards cancelling final offsets ...