A Word on Buffering
Guaranteeing good neighbour relations
last update: Feb 27, 2012

Copyright 2010-20 by H. Gragger. All Rights Reserved. All information provided herein is destined for educational and D.I.Y. purposes only. Commercial re-sale, distribution or usage of artwork without explicit written permission of the author is strictly prohibited.The original units  with their associated  trade-names are subject to the copyright of the individual copyright owner. The Author is by no means affiliated with any of those companies. References to trade names are made for educational purposes only.By reading the information provided here you agree to the Terms of Use.


Early Pedals Had Technological Compromises Built Into Them
Analog Stomp Boxes Are Handmade
Old Habits Are Hard To Change
Ignorance / Lacking Technical Knowledge
Design Objectives
When Is A Buffer Indicated?
Can Buffering Cause Problems Or Make Things Worse?
Remote Powering, a.k.a. "Phantom Powering"
Sound Samples

cookie monsterthis is a free image from profilekiss.com Me want Buffa!

This guy is obsessed with buffering.

With the couple of stompbox clones I recently made I saw myself frequently confronted with pretty bad designs from a perspective of proper input/output impedance, despite the simplicity to do it right.

It is not that the solution to this is unknown; it is thus fully justified that the kind reader asks: "Why does not every manufacturer do it right form a start?"

This has several reasons which we will shed light upon in the following:

Back To Index

Early Pedals Had Technological Compromises Built Into Them

Pedals were coming into scope as soon as the first germanium transistors appeared at reasonable prices. Unfortunately little was known back then about the processes involved in mass-production of transistors compared to todayīs standards, so transistors were spare, leaky, low gain, sensitive and expensive.

Exotic schematics were devised to wring as much boost as possible out of as few devices as possible. Like in tube circuits, this lead (amongst other factors) to compromises in input impedance and/or output impedance.

While those circuit topologies certainly contribute to a deviceīs signature sound, those lead to generic problems like treble loss, muffled sound, funny humps in control sweep and all sorts of incompatibility issues with other devices.

Today, a semiconductor device costs nothing, so fixing this problem by putting a sonically transparent stage in front or after is no problem.

Unfortunately, those designs never get revisited, they stay the same over the decades and keep their unique set of problems.

Back To Index

Analog Stomp Boxes Are Handmade

I recently bought a digital multi-effects device with incredible functional power compared to what a single effects device can do. Funnily both devices may end up with the same price. How come?

The first one has one mass-produced PCB inside that has no more than a handful of VLSI chips on it that are being  placed by automated processes. Despite the fact that they are built like a tank they come out extremely cheap compared to an analog pedal that can only do a diminishing fraction of the things.

Analog pedals are a niche product usually made by small freak companies with an artist touch, not seldom by musicians with a green thumb. Components are often hand selected, hand wired and require fine-tuning. The designs are often rip-offs or close copies of existing designs. So the price goes up for every additional few components and again, the designs keep their problems.

Back To Index

Old Habits Are Hard To Change

Of course them pedals all work more or less and people have gotten used to the ball and chain on their legs. In fact some even demand that the pedal stays the way it is, although proper buffering only adds to the pedalīs versatility.

A few examples:
  • A given circuit has an extremely low input impedance. The problems this leads to are explained in my Poker Face article. Is an unusual volume control behavior something desirable?
  • A passive guitar looses treble with long cables. (Now this is not a stompbox, but has a comparable output impedance problem. The issue is addressed in my guitar modification page) Is a guitar signal that is lame from a start desirable?
  • A given device has high input capacitance and sucks treble. And this only happens when you dial the guitar volume back. Is this what I need?
  • A stompbox loses sparkle when driving long cables. Now this may take away some of its sharpness, but would I not rather want this to be under my control?
  • Mojo components
Now this is certainly one of the show stoppers.

Back To Index

Ignorance / Lacking Technical Knowledge

And, last but not least, some people just donīt know it better or donīt care. It probably also depends on the music type played, because some gear arrangements just swamp everything coming into them.

Back To Index

Design Objectives

A  buffer wrapper to a stompbox device should create ideal impedance relationships with the successor and predecessor, while maintaining the deviceīs virtues. In other words, it should help the device to unleash itīs full potential without any burden, while all the other devices are kept happy too. Sometimes this involves some additional simple circuitry (such as load or drive resistors).This approach should create constant conditions for the device to operate reliably and repeatably in virtually all environments and generally will mean not only a vast improvement in tone, but also versatility.

Back To Index

When Is A Buffer Indicated?

I do not advocate general buffering. In fact half of the stompbox renderings I have done get along without extra buffer. Why?
Either because they are inherently well designed (such as the Great Cheddar) or they do not create apparent problems in my rig.

But when you encounter problems such as:
  • permanent tone degradation or gradual tone degradation during volume control sweep
  • a device does not get on well or not at all with another device in front or after (oscillations etc.)
  • sudden unusual humps on the guitarīs volume control when connected to the device
you should take a buffer into consideration. While you are at it, look if the stompbox you have problems with does proper bypass-switching, because this may be another can of worms.

Back To Index
Can Buffering Cause Problems Or Make Things Worse?

Theoretically a buffer is transparent, meaning it has unity gain and does not contribute any tonal coloration or distortion. Now everything you connect to a guitar changes tone in some way, it is just a question whether this subjectively improves something or not. As quantum theory has it: "even looking at a system changes the way it behaves..."

A  j-fet buffer (my favorite), if biased correctly, has very little current draw, high input impedance, sufficiently good drive. It is also  very low distortion (and if, second order), low noise, very fast, and very low cost.

A BJT might be lower on the output impedance, but this is usually overkill and higher current draw.

An op-amp also works perfect, but might take higher supply currents and might have speed issues. Also, op-amps are known for their coloration (have a look at all the tube screamer discussions...) although they should be transparent. This is an alley that has not been explored enough yet...

And yes, MOSFETs work too...

Look at Ray Marsdenīs 4 part series of "Fet Principles And Circuits".

I agree that, like with all things, excessive buffering may be counter-productive, but creating correct impedance relations on the in- and outputs is not exaggerated. If somebody explicitly wants some of the effects that parasitic capacitance or wrong termination cause, one can always implement this as a switchable option.

Back To Index
Remote Powering, a.k.a. "Phantom Powering"  

Initially I used a buffer box next to my Stratocaster. The box is attached to the guitar strap with an old mobile phone bag. This is unobtrusive with an electric guitar.

Later I wanted to use the buffer box with my lapsteel, because this suffers (like any guitar with a volume pot and a long cable) from tone-sucking when the volume gets turned down. Unfortunately the box is not that elegant dangling down the side of the lapsteel.

I again looked into some remote powering schemes, a.k.a. phantom powering, although the latter already nails you to a quasi hardware standard.

Don Tillman suggests a preamp cable that is remote powered (with the supply in a different enclosure...). While this is a crafty design, it suffers from a few substantial drawbacks:
  • The circuit has gain. Gain is the #1 enemy of any battery driven circuit. There is no gain needed for a guitar signal out of a coil-based pickup.
  • It uses self-bias for the j-fet. This biasing scheme allows only for small negative signal swings and is thus prone to clipping.
To the rescue of the circuit: there is no other all-singing-all-dancing circuit I could find that would allow for remote powering.

A perfect remote powered buffering circuit would need to fulfill the following criteria:
  1. High input impedance, at least medium output impedance
  2. remote powered by either 9V battery or +48V phantom power
  3. low current consumption in 9V mode
  4. +1 Gain (buffering only) or low gain at least
  5. optimized voltage swing (balanced to half supply)
  6. as few components as possible on the "sender" side to fit inside a guitar plug
Unfortunately those requirements cannot simultaneously be fulfilled.
Jensen produce a line of audio transformers for musical instruments use. Amongst their schematics they have AS-004 that describes the remote powering scheme for a buffer . This could be adapted to fulfill all of our criteria - except #3, #5 and #6.

Note that this schematic would work for a +48 Volt remote power, but not if the remote supply were only 9V. The battery has to be close-by.

Upshot: Remote powering is good, is desirable, but not all criteria for a good transparent buffer can be fulfilled without substantial sonic sacrifice. My lapsteel will have to live on with the dongle that cannot be remote powered...

Back To Index

Sound Samples

Some sound files that demonstrate the merits of a buffer in front of a heavily loading effects device and the merits of true bypassing thereof. As effects device my breadboarded (yet unbuffered) FTM clone was used.

The subsequent recordings have been done using the following setup:
  • Strat modified according to specs, neck+bridge pickup
  • J-FET buffer directly after the guitar except where noted
  • FTM clone.
  • 59ī Bassman emulation (slightly turned up gain) with 4x10 Bassman Speaker emulation (Digitech RP-500), with EMT Plate reverb
  • PC hard disk recording over a Turtle Beach Santa Cruz sound card.
(Names may be copyrighted by the associated copyright holder)

Note that on my breadboard a simple output-side effect switching (as it was customary in the early days) is implemented. This does no muting of the effect input as contemporary 3-pole switches do and the effect permanently loads the guitar, even in bypass mode. On the demos the FTM is always off, any crackling distortion you hear is due to heavy loading and bleed-through effects from the non-muted input.

I have done recordings with the guitarīs volume up fully and about 1/4 reduced (always the same position), to demonstrate the effects of different volume drops due to loading effects. Sometimes the volume of the recorded track gets very low. This has not been equalized to demonstrate how much influence the load has.  Please also note how much the tracks differ tonally.

The direct to board recordings go straight into the RP-500, but have a long cable in between. Not too bad a loss, but audible. The buffer is attached directly to the guitar (strap).

Back To Index

Update History

  • Feb 27, 2012: addes section on remote buffering
  • Nov 19, 2010: first release
Back To Index


(c) 2010-2020 AQUATAUR Musik & Elektronik