Best way to "split" an analog signal

Hello, I'm getting an analog output from an op amp (connected as an integrator) and i want to drive it to an oscilloscope via a BNC connector, and also use as an input to a peak detector, also utilized with op amps.
Is it necessary to use a buffer? And which is the best way to split and route the signal in the pcb?
Thanks in adnance!

The maximum linear output current without too much loss from an opamp is only about 10mA peak. If the output is plus or minus 15V peak then its total load resistance must be 15V/10mA= 1.5k ohms.
If the input resistance of the 'scope or peak detector is only 50 ohms then you need a high current buffer.

With an output impedance being high on op Amp the capacitance on coax at 30pF/ft may cause LPF effect. You may not care with an integrator, but some high capacitive loads can might cause ringing in " some" OA's. The scope 1M load is not significant. If you want a lower impedance source to match impedance of cable, you could attenuate 5k to 50Ohm and tell scope you have a 100:1 probe. So connect 5k to BNC socket and 50R to gnd. and connect coax. For clean signals, keep ground connections short to BNC.

Thank you for your replies!
Actually the signals i'm dealing with have much smaller amplitude, 100-200 mV.
I was thinking of using a buffer after the integrator and then with a T-junction (actually Y-junction in pcb, I think it's better in terms of impedance missmatch) route the signal a) to BNC connector, with 50R to gnd and b) to peak detector.
If I keep pcb trace impedance relatively constant at 50 ohm will it be ok with 50 Ohm input resistance scope?

200mV into a load of 50 ohms is only 4mA that any ordinary opamp can drive without a buffer.

The integrator is a lowpass filter cutting high frequencies so matching the impedance is not necessary. Let the extremely low output impedance of the opamp intergrator drive the 50 ohm 'scope input and peak detector if the capacitance of the 'scope cable does not cause the opamp to oscillate.

You won't think about impedance matching unless you are working with multi MHz signals. The question is however very general, nothing said about signal speed.

Yes my mistake, I neglected to mention that. The signal is slow, a few hunderd kHz and maybe all these are not necessary at these frequencies, but i want to be sure that my signal won't be distorted. Even if i am exaggerating, i prefer to add a few extra components and extra caution to pcb layout to ensure that.

For high precision requirements, the external output should be driven by a buffer amplifier. A series resistor (e.g. 50 ohm) should be used to isolate the cable capacitance, with or without buffer amplifier.

Internal buffers can be useful for nonlinear loads or "noisy" loads like sanpling ADCs.

Let's rephrase question. It is NOT a splitter question. This only applies to frequencies that are high enough to require a splitter, pad, controlled impedance.

Source : integrating OpAmp, 100-200mV 100kHz
Destination: Scope 1MOhm via 50 Ohm coax.

Any special considerations for accurate signal capture?

Answer: Routine concerns are 20-30pF/ft capacitive loading of cable proximity to noise. If DC must be undisturbed when connecting coax. (Connecting Cap=> DC charge current dips voltage) if coax. capacitance affects attenuation from OpAmp high output impedance, any pad or buffer will improve response.

Otherwise determine source impedance and load capacitance effects ( determined by cable length as RC low pass filter) the nothing is required to sample signal on coax to scope.


Since loading is acceptable, no "splitting" is required !