I suspect that load resistance is low and PSoC opamp is overloaded. Providing schematic, project bundle and output traces is always helpful. Try to increase load to >1k. Another thing is that PSoC's Opamps are not rail-to rail, so the output will never goes closer than ~0.15V to zero (which itself caps output to ~85% of 0-1V input). Overall, PSoC's Opamps are not great, but can handle small signals upto ~1MHz, so there is something fishy is going on. Please check it with scope probe switched to 10x; it is possible that cable load capacitance is affecting the frequency cut-off.
The signal generator is set to High-Z and the op-amp output is not loaded. I did place a 100k load resistor there however. Scope probes are fixed at 10x and match the scope's settings.
Have attached the project. Changing the amplitude of the signal does not make any difference.
I spent some additional time (apologies for not responding quicker) to verify the behavior between the PSOC 5LP and the external opamp. In the process I've produced and verified the behavior with a schematic that produces a center offset of ~2.5V. (embedded below).
I've replicated all including any trace/wire lengths and continue to see expected behavior with the external opamp and the described unexpected behavior with the PSOC 5LP.
OpAmp_Test.cywrk.Archive01.zip 620.0 K
You were absolutely correct. I looked back to the schematic of the evaluation board and found that the only opamp that wasn't connected to other traces, passives, etc was opamp 1. Once I moved over there, I was able to keep a gain of 6 up to about 100khz.
That explains why one of the opamps was actually outputting a square wave in response to a sine wave.