Positive Feedback in Amp?

[eev_carl]

Hi,

I'm building a bass guitar preamp and am starting with the input and output stages.  I have a pair of op amps on a single 1455 IC, both unity-gain followers.  The circuit is powered by a split power supply.  There are 0.1 caps on +V to GND and -V and GND.

After about 90 seconds of playing or not playing, a repeating sound gets increasingly loader.  You can hear it in the attached link to a WAV file.  I'm on US power.

https://www.bekwam.net/audio/simple_audio_buffer_noise.wav

I was wondering if someone could help me identify the problem.  I killed the lights and extra gear besides the power supply and amp.

Thanks,
Carl

[Benta]

Where's your DC bias? The first amp is floating completely.

[schmitt trigger]

The circuit is no preamp. With the feedback as shown, you have two unity gain stages in series.

In addition to the lack of bias.

[eev_carl]

The circuit is no preamp.

I'm going to add a gain stage, a clipping stage, and a tone control stage.  I figured I'd start with the input and output.

[eev_carl]

For the first stage, should I put a resistor to GND.  If so, what's a good value?  I was confused because most of the schematics I read that had a bias were single supply and needed to account for the negative input.

[eev_carl]

1M for the resistor seems to be working.  Thanks!

[Zero999]

Yes, a 1M resistor to 0V would be ideal. The bias current is typically 70nA, so the output would settle at roughly 70mV, with a 1M resistor to 0V.
https://www.njr.com/semicon/PDF/NJM14558_E.pdf

If the op-amp is AC coupled, then the capacitors should be non-polarised, otherwise they may introduce some distortion as there's insufficient DC bias arccos them. An electrolytic capacitor behaves like a poor diode and is very non-linear when the voltage across it is revered. The input capacitor can be reduced to 10nF, for a lower cut-off frequency of 16Hz. The output capacitor depends on the load impedance, which might be unknown: 10µF is a reasonable value, which would give a lower cut-off of 16Hz into a 1k load. A non-polarised 11µF capacitor can be made by connecting two 22µF electrolytic capacitors in series back-to-back.

[HB9EVI]

I spare me OPAs for other parts of a pedal; input and output buffer can be easily done with an emitter follower; it's enough hi-Z for the guitar pickups and fairly low-Z to drive every realistic load in that environment

[Zero999]

I spare me OPAs for other parts of a pedal; input and output buffer can be easily done with an emitter follower; it's enough hi-Z for the guitar pickups and fairly low-Z to drive every realistic load in that environment

Is the emitter follower bootstrapped to increase the input impedance further?

Getting an input impedance of 1M and an output impedance of 4k7, is easily doable with an ordinary BJT and bootstrapping.

This type of bootstrapping is an example of positive feedback. The output is fed back to the input, via C2, adds to the voltage where R2, R1 and R4 join, this reducing the signal current flowing through R4. There's no oscillation or latch-up, because the gain of the amplifier is under 1 and a positive feedback with a loop gain of above 1 is required for oscillation or latch-up.

https://www.eevblog.com/forum/beginners/passive-components-in-guitar-pedal-input-stage/msg1753601/#msg1753601

[eev_carl]

two 22µF electrolytic capacitors in series back-to-back.

Is that the two positive sides connected?

[eev_carl]

easily done with an emitter follower

I have a 4 op amp IC i'm planning on using.  The other op amps are going to be for a clipping stage and a gain stage.  I'll have a passive tone control and a JFET bypass too.

I was also going to experiment with a source follower on the input and output stages.

[Zero999]

two 22µF electrolytic capacitors in series back-to-back.

Is that the two positive sides connected?

It makes no difference.