Please Educate Me On The Buffer Circuit

[frankie5fingers]

Over the years I've bought a couple of pedals that had a switched buffer circuit.  I never really thought much about it; Ive read that essentially it "preserves" the signal, or rather; returns it to the "strength" it had when it entered the pedal.  When switched off, take the Badgerplex for example, there's a slight (very) volume drop and of course a little of the highs rolled off etc.  Now here's the question.  The Weener has the switchable buffer, right?  I didn't have any reason to switch it off so I didn't.  I socketed R2 so I could use a 47K or 33K if the other mods caused a volume drop.  With the 47 or 33 and buffer on, it was really too far above unity, maybe even causing Q1 to clip.  With the 68K and buffer on, it's right where it should be.  I don't understand the buffer circuit.  Here's the question.  My insatiable urge to tweak has me wondering what would happen if I left the buffer out and played around with r2 to get the volume back - what else would it affect tone-wise?  Thanks

[jkokura]

Great question.

First off, many people misunderstand buffers, including a whole shwack of DIY guys as well, so don't feel bad if you don't know. Second, the most important thing to know is that the most important feature of a buffer is about impedance, which has little to no audible effect.

Essentially you guitar puts out a high impedance signal. This means that there's a lot of electrical load on the signal, meaning that it takes a lot of 'work' for that load to travel down the cable. On a short cable, that doesn't matter at all, there's little to no impact on what you hear. However, if you're using a very long cable, you'll begin to actually hear a depreciation on the signal being put out by your guitar.

To demonstrate this, plug a 10' cord from your guitar straight into a good amp. Turn it up a fair ways and then play for 5 minutes. Now take a 50' cord and do the same. Then go back to the 10'. Can you hear a difference? Try 100'. Even worse.

The solution to this problem of high impedance is called a buffer. What a buffer does is take the high impedance signal of your guitar's pickups and puts out a low impedance signal. Low impedance signals travel down cable with little to no issues at all. That's why you can string mic cables end to end and run snakes across stadiums and get no depreciation in the signal. Doesn't matter if it's 10', 100' or 1000', it should be fine.

A well designed buffer should be unity volume, and should have high enough headroom that your guitar pickups don't overload it. Most of the buffers available as DIY projects around here qualify easily. You should be able to run a buffer at 9V, because it should be imparting little to no impact on the tone of your signal.

A secondary use for a buffer is to assist in the bypass switching that you may have in some pedals. In this arrangement, your signal is always buffered even when the pedal is bypassed. In many cases, this allows many manufacturers to use less expensive methods for bypassing their effects, while simultaneously protecting the user from making the mistake of wrecking their tone with cables that are too long.

In the case of pedals with buffers that can be bypassed, essentially that should mean that the pedal becomes true bypass when the buffer is disengaged. True bypass should mean that the signal only travels through mechanical means and is never touched by any circuitry during it's transit. Most of the time, my experience is that this isn't true and it's really just marketing trickery or perhaps some sort of workalike method that a manufacturer uses. In some other cases though, it really is a choice between buffered or true bypass. I can easily build a simple toggle switch that does that exact thing into my pedals and you can too.

So, long story short, that's what buffers are about.

Jacob

[frankie5fingers]

Now that explains why a buffer at the end seems  more effective than "driving" the chain at the front.  This is interesting, but naturally leads me to more questions.  Is there a variable facet in the buffer boards ie, do all buffer circuits knock the signal to the SAME low impedance or does it change to compensate for the varying degrees of pickup impedance?  If I understand it, the guitar signal is the primary culprit, for lack of a better word.  The buffer boards are relatively small and run on 9v right?  Why don't I load one (battery powered) into my guitar - would that eliminate the need for one down the line or does even a low capacitance (~ 21pF/ft) cable and other effects create the need to balance the impedance?

[crash]

Here is a good read: http://www.moollon.com/techntip/impedance_e.asp

[jkokura]

Good stuff Crash.

Now that explains why a buffer at the end seems  more effective than "driving" the chain at the front.

That actually doesn't make sense to me. I often hear about people running two buffers, one at the beginning and a second at the end of the pedal chain. From my point of view, that shouldn't be necessary. In fact, I'm in favour of running a buffer about as early in your chain as possible. I think that because once your signal has been transformed into a low impedance signal it's done and you don't have to do it again.

One thing I should note is that most pedals when they're turned on act as a dedicated buffer. In fact, if you never operate with no pedals on then you probably don't need a buffer. You just need to be aware of your impedance.

Also, yes your output impedance can be changed by adjusting a few of the values in a buffer. Reading up on some electronics theory will help you understand it.

Jacob

[frankie5fingers]

That IS a great read, thanks Crash.  I'll have to read through it a couple of times.  Yeah Jacob, I see what you mean, I guess there's nothing other than psycho-acoustics to support a buffer at the end vs one at the start, is there?  Yet, it's another pretty active debate.  Interesting...
Thanks guys, for the insight and the site.
Frank

[jkokura]

Yeah Jacob, I see what you mean, I guess there's nothing other than psycho-acoustics to support a buffer at the end vs one at the start, is there?  Yet, it's another pretty active debate.  Interesting...

I'm not sure. Part of the problem is that some buffers also include some tone altering circuitry included in them. The actual buffer itself is/should be truly transparent, meaning that imparts absolutely no adjustment to your tone at all. However, it's obvious that some actually do change your tone. If that's the case, I'd rather people call them a preamp than a buffer, but since buffers are quick cash in today's market I'm sure the manufacturers want to propagate the myth that you need more than one buffer.

jacob

[frankie5fingers]

OK.  Now if we take Brian's Weener... wait, strike that.  If we take the schematic in the Weener Wah buffer circuit build instructions, am I correct in thinking that since a buffer is a current or voltage amplifier; in this case, a current amplifier using a transistor, that because C7 and R12 come before Q3 that they have a function other than buffering, perhaps tone shapers?

[stecykmi]

In this particular case, since C7 and C8 are fairly large caps, their primary purpose is DC blocking. They're not intended to alter the frequency response significantly (they are technically high pass filters, but the corner frequency is so low they act as all-pass filters).

I believe R12 is a pull-down resistor to bias the transistor gate's voltage at 0V (which is a typical value for a JFET).

It's very common to see a transistor in this arrangement, called the common collector or emitter follower. There are many examples in madbean builds alone. Any of the three main types of transistors (BJTs, JFETs and MOSFETs) can all be configured as buffers in nearly identical circuits (only biasing needs to be changed). https://en.wikipedia.org/wiki/Emitter_follower

Another common way to buffer a signal is to use an opamp with a gain of 1. The boneyard, for example, uses a opamp-based input buffer, IC1B on the madbean schem. It's a very common configuration, you'll notice it all the time.