Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Reducing LM358 output noise

Status
Not open for further replies.

Mosaic

Well-Known Member
Hi all:
I have a possible case of random LM358 (12V single supply) noise causing instability on a PWM control circuit voltage.
the setup use 2 LM358 internal opamps.
1st stage measures current difference across a 0.1 ohm current sense via a low pass filter and amplifies by 10 (1k input & 10K FB)
2nd stage takes noninverting input from 1st stage and a target current reference voltage on the inverting input. Gain is x10 again with 1K input & 10K FB resistors.

The output of the 2nd stage passes thru a 1n4148 forward biased and is divided by half with a pair of 39K resistors and a .01uf filter cap to ground before feeding the PWM ref voltage to a TL494. The Opamp outputs in the range of 6 to 7.2V to control the PWM reference in the 2.65V to 3V region.

What is happening is I am occasionally getting instability in the regulation of the PWM and resultant current control, small deviations and occasional significant ones which would amount to a net OPA noise voltage of perhaps 0.01V to .05V. Sometimes its fine for hours, other times for seconds, op temps are <35C. Supply voltages are rock solid.

By breaking the control loop and feeding the TL494 with a trimmer controlled ref V....the instability goes away.
Thus I am wondering if the LM358 is operating in class B with only a 78K load on the output?

I am going to decrease this to around 5K, but I am just guessing here....
What current LOAD does the LM358 require to do Class A outputs?

I can consider an MC33172 replacement as a trial but I only have SOIC on hand...and the app. is DIP with underlying traces.
 
The LM358 is the dual opamp version of the LM324 quad. They were one of the first low power opamps so their output transistors have no bias and produce horrible crossover distortion. Their low power also causes them to be very noisy and have a poor slew rate that limits the full output level response to only 1kHz (most opamps go to 100kHz).
To reduce the crossover distortion the LM358 needs a DC load to the positive supply or to ground of 2k to 3k if it is feeding a high impedance.
 
The LT1013 opamp is also very old, is fairly noisy and has a very poor slew rate that cuts high frequencies at high levels.
 
Well, ok....I seem to have solved the problem.

Step1: Reduce the LM358 diff err. signal to unity gain. It's gain was set to 10X to reduce feedback pulling on the reference setpoint on the inverting input.
Step2: use the inverting Reference signal (1-10V) as the feed for a voltage follower PNP.
Step3: Drive a PNP base (linear biased) from the opamp err signal, tying the emitter to the follower supply voltage created in step2 thereby nulling most of the error signal deviation compared to the reference signal and preventing the unity gain OPA feedback (39K) from pulling the reference signal around.
Step4: Invert the output of the PNP with an NPN biased in its linear range and control the gain with a base trimmer.

Net result is the addition of 3 common bipolar transistors and a few resistors. Quite a bit cheaper than an instrument grade opamp.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top