RF Technology Exploration: Measurement Methods for Intermodulation Distortion

The intermodulation distortion in the transmit (Tx) path is added to the high power, and its characteristics are very important. Since there are generally no active components in the transmit path, its intermodulation distortion characteristics are called "passive intermodulation distortion (passive IMD; PIMD)". When designing a wireless communication circuit, PIMD does not change much in each functional unit, so PIMD will be measured first in the R&D stage.

Two-tone intermodulation test

The most convenient way to measure IMD is to combine two signals of equal power (*so called "two-tone"), keep a certain frequency interval between them, and input them to the input of the device under test (DUT). If measured with a spectrum analyzer, the output spectrum of the DUT will be similar to Figure 1, where the two largest signals are the amplified carrier signals; the other smaller signals are on both sides of the carrier, which are the 3rd, 5th, and 7th order mutual modulation products. The frequency intervals between all signals are equal.

Here are 10 useful measurement parameters:

●Carrier (C): This is the power of the carrier signal in dBm. It is similar to the Pout parameter, but C is measured using a spectrum analyzer, while Pout is measured using a power meter and only an RF signal is input.

●3rd-order intermodulation product (I3): This is the power of the parasitic 3rd-order intermodulation signal, expressed in dBm, and measured using a spectrum analyzer.

●Carrier to 3rd order intermodulation ratio (C/I3): This is the ratio of the carrier power to the parasitic 3rd order intermodulation signal power, expressed in dB.

●3rd order intercept point (IP3): This is the best indicator of the object under test, in dBm. This value usually changes with frequency tuning.

●5th order inter-modulation product (I5): Similar to the 3rd order inter-modulation product.

●Carrier to 5th order intermodulation ratio (C/I5): Similar to the carrier to 3rd order intermodulation ratio.

●5th order interception point (IP5): Similar to the 3rd order interception point.

●7th order inter-modulation product (I7): Similar to the 3rd order inter-modulation product.

●Carrier to 7th order intermodulation ratio (C/I7): Similar to the carrier to 3rd order intermodulation ratio.

●7th order interception point (IP7): Similar to the 3rd order interception point.

3rd level interception point

The nonlinear transfer function of a device or system can be expressed using a "Taylor series":

The third-order intermodulation signal is the third-order term from the f(x) series expansion above, so it is called the "third-order intermodulation product". The third-order input power increases faster than the carrier input power. The unit dBm indicates that the "third-order intermodulation product" is a logarithmic function, which is a relative value of a ratio in mathematics. In fact, the power of the third-order intermodulation signal increases three times faster than the carrier signal power.

. If the curves of the linear parts of "C vs. Pin" and "I vs. Pin" can be extended outward, the intersection point is called the "3rd order intercept point". However, IP3 is a theoretical value and cannot be achieved in actual design because the two curves are saturated before reaching IP3 (the slope approaches 0 and becomes a horizontal line). Usually IP3 is regarded as the "merit function" of RF devices. Most circuit design programs should actually be called "optimization programs". It is up to the designer to decide which areas should be optimized and to what extent. The result obtained in this way is called an "optimization function".

If, in theory, a 3:1 slope difference is assumed, IP3 can be calculated from just one power level. If a power sweep is performed and a graph is obtained, the IP3 calculated in the linear region will be constant (of course the 3:1 slope assumption must be correct). When the carrier and intermodulation signals are saturated, the IP3 value will usually drop, which means that the measurement of the intermodulation power will be wrong. At lower power levels, when the noise floor of the spectrum analyzer is reached, the IP3 will begin to change, which also means that the measurement will be wrong. Therefore, the correct measurement value should be within the power range where the IP3 remains unchanged.

Theoretically, IP3 is not a function of power level. However, at low power levels, its dynamic (linear) region is limited by the noise floor of the spectrum analyzer; at high power levels, due to saturation of the DUT or mutual modulation of the spectrum analyzer, its dynamic region is also limited. Therefore, if IP3 is considered a function of power, it will provide a good way to ensure that the measurement results are correct.