Opamp noise, sound cut off

[Tim Age]

So..... several months later, I made another attempt to work with opamps. To check whether I could do it without messing shit up, I built the simplest possible circuit: opamp buffer. Supply voltage, signal into non-inverting input, output fed back into inverting input, done. Now as far as I know, that should simply replicate the signal, input equals output, gain 1, etc etc. Breadboarded the whole thing, plugged everything in, bass guitar into input, output to amp, played - and instead of bass sounds, I got an odd, quiet, very clipped sound upon playing a note that cut off after a second or less. Everything should be wired up correctly, I rebuilt the whole thing a few times, same results every time. Any idea what the issue is?

I tried a 741 first but was told its supply voltage was 12V, used this 074 instead (datasheet says ideal supply voltage 5V, been told it runs on 4.5 no problem)

[duck_arse]

head over to amz and read this page:
http://www.muzique.com/lab/buffers.htm

you need to bias the opamp input, and isolate both the input and the output from the DC shift the bias produces. and go back to the 741, there is no reason it won't work right, at least for a starter circuit.

[GibsonGM]

No worrying about those supply voltages so much, either!   Use 9V, that is standard and fine for all of these.    As the Duck says, you will learn how to bias the opamp (with usually 4.5V, from the same 9V supply), and you will be installing a couple of blocking capacitors which also allow you to do some tone control.

Take each bit at a time.  Build the buffer, then work on what each section is really doing....

Good luck, welcome, and have fun!   

[Tim Age]

head over to amz and read this page: http://www.muzique.com/lab/buffers.htm

Will do!

and go back to the 741, there is no reason it won't work right, at least for a starter circuit

For what I ultimately have in mind, I'll need 3 amps, hence the 074.

bias the opamp (with usually 4.5V, from the same 9V supply), and you will be installing a couple of blocking capacitors which also allow you to do some tone control

Kinda thought the schematic I posted already did that (10KOhm voltage divider, virtual ground, filter caps), but I'll look into the article duck dude posted!

[GibsonGM]

Nope, that is just a dual polarity power supply.  Opamps were intended, long ago, to work off a plus and minus supply.  Since we have 9V batteries and that's easy, they figured out that you can use your 9V + for power, and just ground the " - " power terminal (not the " - " input, lol).

Since that will only allow you to amplify POSITIVE signals (like a rectifier, the negative part would be cut off), you must 'bias' the input with a DC level, usually 1/2 the supply voltage.     The AC signal will 'mix' with the DC, and be raised up above zero.   The purpose of the capacitors is to block that DC bias voltage from previous and following stages, as well as allow you some tone shaping. 

Read up at AMZ and GEOFEX, Tim!   The 10k, 10, and 1 cap will be common, but note a bias supply is VERY different in how it's used...

[GibsonGM]

Check out the image below. Note the 2 10k resistors and the cap at upper left; that is the bias supply.   The resistors, being equal-valued, divide the voltage in half.  The cap acts like a 'reservoir', and decouples this bias voltage from the power supply, which is the 9V batter. 

You'll see where the point marked "4.5V" is connected to the opamp and other components, elevating the incoming signal by 1/2 the supply voltage so both neg. and pos. parts of the signal can be amplified.   It's a bit different than the dual supply in your image above.

[Tim Age]

Ok so I gather what I tried to build here was this, minus the filters?



My power supply is almost identical to theirs

but for the cap value and them omitting a second cap.

And apparently what I'll want is this?



Or will the first pic do if I add the filter caps and resistors?


Basically what I think I understood from these explanations is that the +4.5 need to go into the signal, not the op amp, to make it fully positive, and later be filtered out again, yes?

[antonis]

Last schematic is prone to noise, due to big supply resistors..
(with all respect to Jack..)

2 X 10k with 1M from Vr to non-inverting input should be OK..
(you can also use 470k - 510k & 220nF cap with same results..)

"Filter" caps are mainly placed there for DC stop (AC only coupling) to prevent any signal upset..
(they're secondary used for high pass filtering..)

Basically what I think I understood from these explanations is that the +4.5 need to go into the signal, not the op amp, to make it fully positive, and later be filtered out again, yes?

Not exactly..

Whatever goes into the signal also goes into the op-amp.

+4.5V set the mid-point of signal swing headroom, for a single supply (+9V / - 0V) fed op-amp..
(signal can - IDEALLY - equally vary between 0V & +9V..)

"Filtering" is only needed when next stage is biased to any point other than +4.5V or/and there is a DC decoupling cap..
(signal from op-amp out carries its own AC waveform PLUS +4.5V DC ..)

[Tim Age]

Now I don't understand shit anymore.... every tutorial on the web never mentions any biasing, this AMZ link describes a circuit almost identical to what I tried... I don't get it. Might as well give up.

[MaxPower]

A matter of perspective I guess. The Vref sets the dc operating point for the op amp. Similar to how transistors are usually biased so the collector sits at about 4.5v (dc). The ac signal swings around the dc operating point.

And yes, the dc is filtered out from the ac signal later.

Texas Instruments has some PDFs about op amps you might want to read. Analog Devices has some useful stuff as well (how to minimize noise for example). A good book on electronics is Electronic Principles by Malvino. An older edition is fine (1980s or newer). They can be found for about $5 last time I checked.

[antonis]

 

Keep calm and try to understand the difference between DC & AC amplifiers.. 

[GibsonGM]

Abandon what you 'first built'.  Just build the AMZ buffer, as Jack shows it in the schematic!

Make this:


Notice that you have 9V on the top at V+, which is where your battery is connected, and 4.5 at "Vr". Never mind what you built before. This is your bias supply.  Of course, battery "minus" goes to the bottom, ground.


Then make this with your 741:


This is the actual buffer, there are no filters (?)   Connect "Vr" from the bias supply to the 1M resistor where it says "Vr".  Make sure you have an input cap (the one on the left), and the output cap on the right.   Value is not THAT important right now, anything to get it working!   

Plug in and try it out. Make sure the 9V goes to chip's "+" pin 7, and that pin 4 is grounded.

Don't give up, you are VERY close to having it working!  And we'll help you understand what is going on with this "bias" crap.

[EBK]

I'd like to briefly take a step WAY back to your original circuit, if I may.

That ground indicated on the input and output jack sleeves.....   Did you connect those to the same place you drew the ground symbol in your power supply circuit?  Or, did you mistakenly connect those to the negative terminal of your battery?

[MaxPower]

Google: texas instruments op amp pdf

That should bring up links to a few of the Texas Instruments's PDFs.

[GibsonGM]

If you don't 'bias' the opamp with DC this way, your AC guitar signal will look something like the bottom signal in this this pic (but not exactly, for other reasons):


The bottom half of the signal is cut off, because without 'bias', the opamp cannot amplify something that's going below zero (keeping it simple here).

So, to let it amplify BOTH parts of the wave, the plus AND minus halves, we add DC to raise the signal 'up', shown in the top pic.   The 1M resistor adding the 4.5V to the opamp does this; it is a property of nature that allows the 2 voltages to mix with each other (and to be brought back to their original form after we do work on them).    This DC is removed after the signal is amplified, by the output capacitor, and the signal again rides around the zero volt level, which is what MAKES it an "AC signal".  If you OFFSET the signal this way, it looks like this:



The reason you mostly use 4.5V for the bias in a 9V circuit is to allow the maximum amount of amplification to both sides of the AC signal (headroom) for that given power supply. 

Quote from: EBK on Today at 01:01:05 PM

    I'd like to briefly take a step WAY back to your original circuit, if I may.

    That ground indicated on the input and output jack sleeves.....   Did you connect those to the same place you drew the ground symbol in your power supply circuit?  Or, did you mistakenly connect those to the negative terminal of your battery?


Yes, the original SHOULD work, just would probably result in a crappy sound (if your input was hot), but functional.  Eric is correct in saying you should check that the 'ground' connections do go to the point between the R's and C's....

[Tim Age]

I'd like to briefly take a step WAY back to your original circuit, if I may.

That ground indicated on the input and output jack sleeves.....   Did you connect those to the same place you drew the ground symbol in your power supply circuit?  Or, did you mistakenly connect those to the negative terminal of your battery?

All ground symbols connect. from what i understood, the centre point / bias supply / Vr is called a virtual ground(?) i think and that connects to the sleeve on in and out jacks.

If you don't 'bias' the opamp with DC this way, your AC guitar signal will look something like the bottom signal in this this pic (but not exactly, for other reasons):


The bottom half of the signal is cut off, because without 'bias', the opamp cannot amplify something that's going below zero (keeping it simple here).

So, to let it amplify BOTH parts of the wave, the plus AND minus halves, we add DC to raise the signal 'up', shown in the top pic.   The 1M resistor adding the 4.5V to the opamp does this; it is a property of nature that allows the 2 voltages to mix with each other (and to be brought back to their original form after we do work on them).    This DC is removed after the signal is amplified, by the output capacitor, and the signal again rides around the zero volt level, which is what MAKES it an "AC signal".  If you OFFSET the signal this way, it looks like this:



The reason you mostly use 4.5V for the bias in a 9V circuit is to allow the maximum amount of amplification to both sides of the AC signal (headroom) for that given power supply. 

Quote from: EBK on Today at 01:01:05 PM

    I'd like to briefly take a step WAY back to your original circuit, if I may.

    That ground indicated on the input and output jack sleeves.....   Did you connect those to the same place you drew the ground symbol in your power supply circuit?  Or, did you mistakenly connect those to the negative terminal of your battery?


Yes, the original SHOULD work, just would probably result in a crappy sound (if your input was hot), but functional.  Eric is correct in saying you should check that the 'ground' connections do go to the point between the R's and C's....

Ohhhhhh ok that makes sense.... so mathematically speaking, if this was a graph or some, I'm adding +4.5 to the equation, run it thru the amp and the caps at the end do a -4.5 and lower it back down?

This is the actual buffer, there are no filters (?)   Connect "Vr" from the bias supply to the 1M resistor where it says "Vr".  Make sure you have an input cap (the one on the left), and the output cap on the right.   Value is not THAT important right now, anything to get it working!

Got 0.1uF, 1MOhm and 47uF, that should do then for now?

[GibsonGM]

You're just about there, Tim.  The cap blocks AC....voltage of ZERO frequency can't pass thru it.   {A tiny amount inevitably does due to leakage, but that's the idea}     It doesn't really "minus" anything, just lets the AC pass thru.    The 2 don't really mix, so they can be played with this way.  Yes, you are 'scooting' it up to work with it, then it comes back down because the DC gets left behind...

Caps..oh yeah...use the .1u for the input, and 47u on output, good to go!    Give it a shot, let us know if it works!!    If you're using the AMZ  "vr" version we've been talking about, ground IS the battery minus terminal, by the way, just to be sure.     The split supply you first showed, ground is the midpoint....not this, tho!

[Tim Age]

Phew. Finally worked. Didn't have a 10uF so I used a 1uF for the power supply and a 47uF and 1uF in series for the output (resulting in ~9uF), worked out nicely. Thanks for the help yall (and sorry for throwing a bit of a fit earlier, bad temper and no sleep don't mix, I'm afraid :/)

[GibsonGM]

No problem, Tim!  With some reading, looking up opamp basics on Youtube, you will quickly see what it's all about.  Focus on single-supply, which is what most of the things we do in pedals involves.      You have buffers, inverting and non-inverting gain stages to mess around with!   AMZ has tons of stuff on this.

[antonis]

Phew. Finally worked. Didn't have a 10uF so I used a 1uF for the power supply and a 47uF and 1uF in series for the output (resulting in ~9uF), worked out nicely.

Well done, Tim..!!

Although you could use 47μF for power supply and 1μF for IN/OUT..
(always use the bigger between availiable caps for "smoothing" power supply - voltage ripple rejection is proportional to cap value and LPF resistor value should be lower, resulting in lower voltage drop..)