beautypg.com

Common mode rejection ratio – Teledyne LeCroy WaveLink Differential Probe Series (13-25 GHz) User Manual

Page 89

background image

Operator's Manual

WL-HBW-A-OM-E Rev B

83

even though both outputs swing 370 mV, the maximum difference voltage between them at any
instant is still within ±370 mV. So, this signal could be measured with a differential amplifier that has
a differential mode range of ±400 mV.

Common Mode Rejection Ratio

The ideal differential probe/amplifier would sense and amplify only the differential mode voltage
component and reject the entire common mode voltage component. Real differential amplifiers are
not perfect, and a small portion of the common mode voltage component appears at the output.
Common Mode Rejection Ratio (CMRR) is the measure of how much the amplifier rejects the
common mode voltage component. CMRR is equal to the differential mode gain (or normal gain)
divided by the common mode gain. Common mode gain is equal to the output voltage divided by
the input voltage when both inputs are driven by only the common mode signal. CMRR can be
expressed as a ratio (e.g., 10,000:1) or implicitly in dB (e.g., 80 dB). Higher numbers indicate greater
rejection (better performance).

The first order term determining the CMRR is the relative gain matching between the + and input
paths. Obtain high CMRR values by precisely matching the input attenuators in a differential
amplifier. The matching includes the DC attenuation and the capacitance which determines the AC
attenuation. As the frequency of the common mode component increases, the effects of stray
parasitic capacitance and inductance in determining the AC component become more pronounced.
The CMRR becomes smaller as the frequency increases. Therefore, the CMRR is usually specified in a
graph of CMRR versus common mode frequency.

The common mode frequency in these graphs is assumed to be sinusoidal. In real life applications,
the common mode signal is seldom a pure sine wave. Signals with pulse wave shapes contain
frequency components much higher than the repetition rate may suggest. This makes it very difficult
to predict actual performance in the application for CMRR-versus-frequency graphs. The practical
application of these graphs is to compare the relative common mode rejection performance
between different probes and amplifiers.