M
mook Johnson
- Jan 1, 1970
- 0
I have a buddy that is struggling with a design. Heres what he's trying to
do.
He recieves a 45 - 200Hz 500mV p-p sine wave through a signal transformer
and would like to buffer it and drive and high current version of the signal
to an output. Output load cna be anywhere from open circuit to .25 ohms.
So far he had a 6:1 transformer primary being driven by NPN-PNP totempole
and an OPamp to elimiante the xover distortion (aka simple audio amp). It
is desired that the secondary be allowed to have about 2Vp-p of 60-120Hz
common mode voltage on it from external sources so it can't be tied to
ground but it doesn not need to truelly be isolated.
What he is attempting to do is to AC couple off both sides of the output
coil and bring that differential signal back into to the primary side to a
second opamp. The feedback signal will be compared by this opamp to the
reference signal and any load dependant droop will be corrected out. with
all the elememts in the loop (opamp 1, opamp 2+ totem, 6:1 transformer,
etc, stability is a bit shaky. He even ran into the dreaded motorboating
problem.
Basically its a DC to AC linear inverter problem but with an isolated output
and stiff voltage requirements on the output voltage.
I have some ideas about using a isolated power supply to generate the rails
for a power opamp and us it to buffer the signal transformer. This would
generate quite a bit it heat even with the +/-5V rails most power opamps
require. The low load resistance is whats challenging.
How would you analog types go about attacking this problem?
do.
He recieves a 45 - 200Hz 500mV p-p sine wave through a signal transformer
and would like to buffer it and drive and high current version of the signal
to an output. Output load cna be anywhere from open circuit to .25 ohms.
So far he had a 6:1 transformer primary being driven by NPN-PNP totempole
and an OPamp to elimiante the xover distortion (aka simple audio amp). It
is desired that the secondary be allowed to have about 2Vp-p of 60-120Hz
common mode voltage on it from external sources so it can't be tied to
ground but it doesn not need to truelly be isolated.
What he is attempting to do is to AC couple off both sides of the output
coil and bring that differential signal back into to the primary side to a
second opamp. The feedback signal will be compared by this opamp to the
reference signal and any load dependant droop will be corrected out. with
all the elememts in the loop (opamp 1, opamp 2+ totem, 6:1 transformer,
etc, stability is a bit shaky. He even ran into the dreaded motorboating
problem.
Basically its a DC to AC linear inverter problem but with an isolated output
and stiff voltage requirements on the output voltage.
I have some ideas about using a isolated power supply to generate the rails
for a power opamp and us it to buffer the signal transformer. This would
generate quite a bit it heat even with the +/-5V rails most power opamps
require. The low load resistance is whats challenging.
How would you analog types go about attacking this problem?