• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 48 subscribers
  • Views 594 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMC6064: Simple non-inverting amplifier with gain of 2 is unstable

J D22
Prodigy 10 points
Part Number: LMC6064

Hi folks,

I built a 24-channel ADC signal conditioning module using six LMC6064AIM chips.  I am using them on a +10.5V single supply in a non-inverting 2x gain configuration to amplify low-frequency (< 10 Hz) signals from 0-5V to a range of 0-10V.  The input signal at the non-inverting pin is filtered with two RC stages consisting of 1kohm+1uF followed by 10kohm+1uF, so the input signals are extremely clean.  I started with two resistors on each amplifier:

  • 1 kohm from output to inverting terminal
  • 1 kohm from inverting terminal to ground

That's about as simple as it gets, yet some channels oscillate at 34-40 kHz with amplitude of 100 mV to 500 mV peak-to-peak.  The waveform on the scope flattens out considerably near the top of each cycle; it is a lopsided sine wave with a sloped flat top.  (No way to post the image at the moment.)  The output signal is centered around 2.4V, so it's not hitting the supply rails.

Bringing the power supply slowly down and back up restores a stable output signal for a time, but if  I power down the circuit and abruptly power back on, the oscillation returns.

I observe from the Bode plot in the data sheet that the phase margin is at least 45 degrees for a 10kohm load, but no Bode plot is given for lower resistances.  If I remove the V- to GND 1k resistors, the output is stable, and goes unstable again when I put the resistor back into the circuit.  I would expect the feedback gain of 1/2 to increase stability over the unity-gain configuration, and the amplifier is unity-gain stable, but another difference here is that the feedback gain of 1/2 introduces a DC load of the output.  With low feedback resistors I would expect input capacitance effects to be minimal, but I added 0.022 uF ceramic capacitors from output to inverting input in an effort to increase phase margin with no improvement.  Bypass capacitors are 10uF tantalum, .01uF ceramic, and .1uF ceramic soldered from pin 4 to pin 11.  They have no effect on the oscillation, either.  The supply rail is solid and clean using a linear regulator and a 47uF capacitor.

From the data sheet it would appear that this circuit is well within the output capabilities of the amplifier, since the maximum DC output current in this configuration is 5mA per channel at +10V output swing.  The amplifier has on problem reaching +10V output.

This is a hand-built prototype with the chips and components soldered to SIP adapters, so I would expect stray capacitance around the amplifier to be minimal.  The board has a copper foil ground plane.

I've dealt with unstable op amps before, but typically it was due to crummy layout or could be explained by the Bode plot.  The asymmetry and flat top of the waveforms out of the amplifiers lead me to think I might be running into a nonlinearity that's adding phase shift beyond what's given for 10kohm load.  Still, it seems that I am within the ratings of the op amp and should expect stable performance.  What am I missing?

Thank you!

  • 0
    TI__Guru* 93105 points

    Hi,

    The circuit behavior from your description appears to be marginally stable and when it does become unstable it oscillates near the unity-gain cross frequency. That is an indicator that the phase margin has dropped to the point where there isn't any margin remaining. The LMP6064 Gain and Phase vs Capacitive load graphs show that a matter of a few tens-of-picofarads on the output does reduce the phase margin significantly. You mention you are driving a 24-channel ADC signal conditioner so I would presume each LMP6064 is driving a number of ADC inputs. Depending on the type of  converters, SAR or delta-sigma, the input capacitances can be significant and will present a capacitive load to the op amp.

    Do note that the 1k feedback and 1 k input resistor do load the output with the equivalent of a 2 k load. Even though the LMP60604 has fair output drive capability, it really isn't specified driving a load impedance that low. That surely affects the open-loop gain/phase response of the amplifier.

    Please provide a detailed schematic for your circuit that includes the ADC information. Use the attachment icon from among the choices in the icon bar. It is the one to the left of the paperclip that sort of looks like a CRT screen. It takes common graphics formats.

    Regards, Thomas

    Precision Amplifiers Applications Engineering