three 100 nf ceramic bypass capcitors. I am measuring some instabilty in
frequency harmonics of 7Hz. What may be the reason.
The REF193 should not can any inherent instability in the configuration you
describe. The output of the REF193 is stable for all load currents up to 30mA
for cap load of 1uF or greater.
Ensure that you have allowed for the load and line regulation specs of the
REF193 in your calculations of expected ripple. Remember to consider maximum
and variations in load current.
Ensure the input voltage allows sufficient headroom (minimum headroom of 0.8V
for a 10mA output)
If, after checking the above, the output ripple is still higher than expected,
it is likely that the 7Hz spurious is coming from somewhere else in your
system. Review your system for switching power supplies or other magnetic
elements which could cause inductive coupling, long digital clock lines which
may couple capacitively onto the power supply, fast switching load currents,
residual ripple on the input to the REF193 (although the REF193 should offer
rejection >100dB at 7Hz)