Strange OP-amp behaviour - For amplifier experts.

[[email protected]]

Hello All

We have an LM224DG (ON semi) op amp amplifying the audio out of a GSM
module. Differential stage followed by inverting amplifier.

We always have problems with these designs trying to keep the hum
caused by the 217Hz pulsing of the GSM system (pulsing 2.5 - 3A for
the module when Tx) out of the audio circuits.

Whats happening in this design has me confused. The gain phase margin
of the inverting amplifier is causing oscillations at 40KHz. Ok no big
deal, change the gain and capacitive loading on input, output or
feedback and problem goes away.

The problem is that once I remove the 40KHz oscillations the GSM Hum
is much worse. The thing is you cant hear the 40KHz component and with
no GSM Hum you would nearly be tempted to leave the oscillations.


Does this operation make any sense? I just dont see it. Any
sugestions.

Many thanks for any input.


Regards


Denis
________________________
http://www.CentronSolutions.com

 

[Jim Thompson]

Hello All

We have an LM224DG (ON semi) op amp amplifying the audio out of a GSM
module. Differential stage followed by inverting amplifier.

We always have problems with these designs trying to keep the hum
caused by the 217Hz pulsing of the GSM system (pulsing 2.5 - 3A for
the module when Tx) out of the audio circuits.

Whats happening in this design has me confused. The gain phase margin
of the inverting amplifier is causing oscillations at 40KHz. Ok no big
deal, change the gain and capacitive loading on input, output or
feedback and problem goes away.

The problem is that once I remove the 40KHz oscillations the GSM Hum
is much worse. The thing is you cant hear the 40KHz component and with
no GSM Hum you would nearly be tempted to leave the oscillations.


Does this operation make any sense? I just dont see it. Any
sugestions.

Many thanks for any input.


Regards


Denis
________________________
http://www.CentronSolutions.com

Post your schematic, either via a URL link or place on newsgroup
alt.binaries.schematics.electronic

...Jim Thompson

 

[Winfield Hill]

Hello All

We have an LM224DG (ON semi) op amp amplifying the audio out of a
GSM module. Differential stage followed by inverting amplifier.

We always have problems with these designs trying to keep the hum
caused by the 217Hz pulsing of the GSM system (pulsing 2.5 - 3A for
the module when Tx) out of the audio circuits.

Whats happening in this design has me confused. The gain phase margin
of the inverting amplifier is causing oscillations at 40KHz. Ok no
big deal, change the gain and capacitive loading on input, output
or feedback and problem goes away.

The problem is that once I remove the 40KHz oscillations the GSM Hum
is much worse. The thing is you cant hear the 40KHz component and with
no GSM Hum you would nearly be tempted to leave the oscillations.

Does this operation make any sense? I just dont see it. Any
sugestions.

LM224, LM324, LM358, whatever, a very poor
choice for audio, most folks would say.
That's in part because it has a horrible
huge deadzone in its class AB output stage,
where neither the pullup or the pulldown
transistor is significantly working. This
deadzone causes distortion at higher audio
frequencies, etc.

I'll take a guess about your problem, and
say the 40kHz oscillation caused a dither
so one or the other output transistor was
always on, pulling up or down, thereby
creating a low output impedance, like you
would have with an ordinary opamp. But,
when you killed the 40kHz dither, now the
opamp's effective output impedance went up
and the output wiring became susceptible
to the GSM hum (whatever that is). You
can test my theory by subtituting a better
part for the LM224, such as LTC's LT1014i,
which has a properly biased output stage.

 

[Joerg]

Hello All

We have an LM224DG (ON semi) op amp amplifying the audio out of a GSM
module. Differential stage followed by inverting amplifier.

We always have problems with these designs trying to keep the hum
caused by the 217Hz pulsing of the GSM system (pulsing 2.5 - 3A for
the module when Tx) out of the audio circuits.

Whats happening in this design has me confused. The gain phase margin
of the inverting amplifier is causing oscillations at 40KHz. Ok no big
deal, change the gain and capacitive loading on input, output or
feedback and problem goes away.

The problem is that once I remove the 40KHz oscillations the GSM Hum
is much worse. The thing is you cant hear the 40KHz component and with
no GSM Hum you would nearly be tempted to leave the oscillations.


Does this operation make any sense? I just dont see it. Any
sugestions.

Many thanks for any input.

Bipolar amps can be really nasty with cell phone exposure. Their
numerous b-e paths act as rectifiers. Especially GSM is bad.

Try a FET opamp. Or at least bypass the inputs against each other and/or
GND really well. Just a few hundred pF, you want the ESR really low all
the way up to 2GHz and no strange resonances up there. Very small
inductors right in front help as well. Ferrite beads also but they
aren't too effective above a GHz.

Do you have a full ground plane? If not this could become a little
nightmare. Not hopeless but tough.

 

[John Larkin]

Hello All

We have an LM224DG (ON semi) op amp amplifying the audio out of a GSM
module. Differential stage followed by inverting amplifier.

We always have problems with these designs trying to keep the hum
caused by the 217Hz pulsing of the GSM system (pulsing 2.5 - 3A for
the module when Tx) out of the audio circuits.

Whats happening in this design has me confused. The gain phase margin
of the inverting amplifier is causing oscillations at 40KHz. Ok no big
deal, change the gain and capacitive loading on input, output or
feedback and problem goes away.

The problem is that once I remove the 40KHz oscillations the GSM Hum
is much worse. The thing is you cant hear the 40KHz component and with
no GSM Hum you would nearly be tempted to leave the oscillations.


Does this operation make any sense? I just dont see it. Any
sugestions.

Many thanks for any input.


Regards


Denis
________________________
http://www.CentronSolutions.com

Bipolar opamps love to rectify RF. Try a jfet or cmos amp and see if
that helps. LF347 is cheap and behaves pretty well.

John

 

[[email protected]]

Hi guys

Many thanks for the input. Will try a fet or CMOS opamp.

Denis
___________________
http://www.CentronSolutions.com