Skip to content
Need help with guit...
 
Notifications
Clear all

Need help with guitar buffer circuit...

7 Posts
3 Users
0 Reactions
3,017 Views
(@jeghers)
Active Member
Joined: 16 years ago
Posts: 8
Topic starter  

You folks who know electronics, please advise me! I am wanting to design some buffer circuitry for a homebrew fancy guitar selector gadget. I know that op-amps will give me more flexibility in circuit design over JFETs, but some people have observed that op-amp buffers sound "colder and more sterile".

Do you have any opinions? Would a JFET circuit really preserve the guitar's tone better? If there is a difference that could be noticed, I think I'd be willing to do the more involved JFET design, though I'll need to know how to boost the gain (I'll want to boost the signal be a few db), which of course op-amps can do more easily.

Also, seems to me a JFET circuit might consume less power than using op-amps -- Agree? Disagree?

I know this is a rather ethereal question, but if any of you know about this, please advise.

/Mark


   
Quote
(@gnease)
Illustrious Member
Joined: 20 years ago
Posts: 5038
 

Some general notes:

There are low power opamps, and some of these probably have FET inputs (either JFET or MOSFET). Even a discrete FET buffer will require some reasonable bias that will result in some current draw. If you are going for Class A, that will not be insignificant.

"Cold and sterile" is a term usually mis-applied to amps that do the job without coloring the signal. All those Class A input FET and Tube amps some love are probably adding color in the form of even order harmonics (FETs are square-law devices), and that makes the tone "warmer." What do you really want? Amplification without coloration in order to let the subsequent stages (EFX, amp) color the sound? If so, then "cold and sterile" might be exactly what you need. Another reason "cold and sterile" may be applied to an opamp circuit -- even with FET inputs, is that balanced driving of the input differential pair will suppress even order harmonic distortion. With the usual feedback arrangements to linearize the amp, all that will be left (in addition to the signal) will be very low level odd order harmonic distortion => unfairly tagged "sterile"

Make sure whatever design you choose has enough head room to handle the max signal without voltage peaks coming near the rails -- again, unless you want coloration at higher volume (guitar or buffer amp) levels. These days, a number of internal amp and high quality effects use 15 VDC or greater voltage supplies to achieve good dynamic range on the high level end (low end depends on noise figure of the amplifier circuit. Most audio-specific, FET input monolithic amps are fairly low noise.

Avoid using a lot of negative feedback around opamp designs that have significant delays (lots of stages). Too much feedback around an amp with significant delay results in transient intermodulation distortion. Very unmusical. Read the app notes on this.

When designing a buffer, prioritize the reducion of any type of intermodulation distortion over reduction of harmonic distortion. The former sounds nastier (in a bad way).

Realize that by adding a buffer, you will be lowering the output impedances of your guitar pups. Good for driving cables and preserving frequency response, but that difference in impedance may change the way your guitar interacts with your amplifers input -- for better or worse.

-=tension & release=-


   
ReplyQuote
(@jeghers)
Active Member
Joined: 16 years ago
Posts: 8
Topic starter  

My goal is that I do not want to color the guitar sound. I am designing a switching device (to select multiple guitars) that includes gain control (with the ability to boost a few db). Later devices will color the sound, so I want this device to ONLY select which guitar without changing the tone. The point of the buffering is A) to present a high input impedance to the guitars so their passive pickup/tone circuits are not loaded down and colored in any way and B) low output impedance to drive the amp/pedal that it feeds into really well.


   
ReplyQuote
(@ricochet)
Illustrious Member
Joined: 21 years ago
Posts: 7833
 

If y'all were talking about vacuum tubes I might have some useful input. This solid state stuff's a passing fad. :D

"A cheerful heart is good medicine."


   
ReplyQuote
(@jeghers)
Active Member
Joined: 16 years ago
Posts: 8
Topic starter  

Make sure whatever design you choose has enough head room to handle the max signal without voltage peaks coming near the rails -- again, unless you want coloration at higher volume (guitar or buffer amp) levels. These days, a number of internal amp and high quality effects use 15 VDC or greater voltage supplies to achieve good dynamic range on the high level end (low end depends on noise figure of the amplifier circuit. Most audio-specific, FET input monolithic amps are fairly low noise.

This is for signals direct from guitars, would 15VDC really make a difference in that case compared to 9VDC? If I we able to get away with 9VDC, then I could design it with the option of power from a single 9V battery (vs external power). Of course, if it would make a substantial difference, then I'd probably opt for 15VDC (reluctantly)
Avoid using a lot of negative feedback around opamp designs that have significant delays (lots of stages). Too much feedback around an amp with significant delay results in transient intermodulation distortion. Very unmusical. Read the app notes on this.

I am not clear on this. Do you refer to the feedback resistor in the classic op-amp circuit where the 2 resistors determine the gain? I need a little clarification here.

FYI, I am planning as few stages as possible. I want it simple: the guitar feeds into a high-impedance input, goes thru a solid-state switch (e.g. CD4066 or similar), and provide a volume pot that has a little gain boost available. No fancy preamps, etc.
When designing a buffer, prioritize the reducion of any type of intermodulation distortion over reduction of harmonic distortion. The former sounds nastier (in a bad way).

What kind of design practices would likely incur intermodulation distortion?
Realize that by adding a buffer, you will be lowering the output impedances of your guitar pups.

Huh? I'm afraid I don't follow. My goal is to have a very high impedance load that affects the passive pups circuitry minimally. Am I missing something?

Thanx,
/Mark


   
ReplyQuote
(@gnease)
Illustrious Member
Joined: 20 years ago
Posts: 5038
 

Make sure whatever design you choose has enough head room to handle the max signal without voltage peaks coming near the rails -- again, unless you want coloration at higher volume (guitar or buffer amp) levels. These days, a number of internal amp and high quality effects use 15 VDC or greater voltage supplies to achieve good dynamic range on the high level end (low end depends on noise figure of the amplifier circuit. Most audio-specific, FET input monolithic amps are fairly low noise.

This is for signals direct from guitars, would 15VDC really make a difference in that case compared to 9VDC? If I we able to get away with 9VDC, then I could design it with the option of power from a single 9V battery (vs external power). Of course, if it would make a substantial difference, then I'd probably opt for 15VDC (reluctantly)

Yes, it can make a difference. Especially if you are using a single-ended, Class A JFET or MOSFET amp design with source DC feedback (resistor) and some gain. You want to avoid the rails with some margin. With a fading 9V battery, you may only have an output swing range of a 9V (battery) - 2V (source resistor DC feedback drop) - 2V (fading life) = 5 V range minus some margin for linearity. The budget gets tight when all this is factored in. Also consider that a 'bucker may produce 2 Vp-p. At 6 dB gain, that's an output signal swing of 4 Vp-p. Getting tight.
Avoid using a lot of negative feedback around opamp designs that have significant delays (lots of stages). Too much feedback around an amp with significant delay results in transient intermodulation distortion. Very unmusical. Read the app notes on this.

I am not clear on this. Do you refer to the feedback resistor in the classic op-amp circuit where the 2 resistors determine the gain? I need a little clarification here.

Yes, that would be the feedback mechanism for a voltage gain opamp. There is a finite time for the feedback to act. If the gain reduction due to negative FB is great (say 80 dB down to 6dB) what do you think happens in that short time it takes the feedback to stabilize the situation? For a brief instant (delay through amp) the gain is 80 dB and the output slews toward clipping at a supply rail, until the negative correction finally "arrives" back at the summing node. So the amp is briefly "open loop." This is one of the reasons piezo pups sound so crappy through the budget equalizers. The sharp transient edges go uncorrected for a brief, but audible time in circuits where the FB loop delays are significant and the correction (FB ratio) is especially large. There are audio opamps designed to be used as unity buffers (lots of negative FB) that are designed with very low delay (fewer stages) to avoid this situation. That will be discussed in the app note.
FYI, I am planning as few stages as possible. I want it simple: the guitar feeds into a high-impedance input, goes thru a solid-state switch (e.g. CD4066 or similar), and provide a volume pot that has a little gain boost available. No fancy preamps, etc.

Again, many opamps contain many internal stages. If you decide to use an opamp, pick one with few stages => less delay. The opamp app note usually will point out it's low delay as a design advantage.
When designing a buffer, prioritize the reducion of any type of intermodulation distortion over reduction of harmonic distortion. The former sounds nastier (in a bad way).

What kind of design practices would likely incur intermodulation distortion?

The type known as Transient IM (TIM) is caused by delays in feedback loops that are appreciable compared to the edges of the signals of interest (guitar signals). Other, more "conventional" IM depends upon the amp design, bias, signal amplitude. I'm suggesting you pay attention to the app notes and/or specs for whatever opamp or device you chose to make sure the guidelines for suppressing/mitigating IM are observed. AFA what IM does: It cause multiplication of signals applied to the amplifier's input, producing additional tones at the output. Example: Inject 440 Hz and 500 Hz simultaneously into the amp. Intermodulation distortion will cause additional tones to appear at the output at 380 Hz and 560Hz (these are third order and higher products), as well as 320Hz and 620 Hz (fifth order +) and so on. These tones are usually low in level, but the ear is quite good at picking them out as esp dischordant.
Realize that by adding a buffer, you will be lowering the output impedances of your guitar pups.
[a/quote]

Huh? I'm afraid I don't follow. My goal is to have a very high impedance load that affects the passive pups circuitry minimally. Am I missing something?

Thanx,
/Mark

You need to look from both directions: sink and source. The high Z input buffer is good for the pups' "view of the world". But you are forgetting that the amp normally "sees" the pups as source impedances, and you are altering (lower substantially) that impedance when you add the buffer. That change can effect overall tone. If it does, you may wish to add and output circuit to your buffer system that will emulate a pup's impedance from the amps "point-of-view."

-=tension & release=-


   
ReplyQuote
(@jeghers)
Active Member
Joined: 16 years ago
Posts: 8
Topic starter  

Can you point me to a good op-amp with the short delay you've mentioned?

Oh, and thanx for the helpful replies :)

/Mark


   
ReplyQuote