Talking about the Principles of Real Environment Simulation for RF Test and Measurement

This article attempts to explain such a point of view-and measurement should follow the principle of "how do you use, how do I measure".

The principles of testing and measurement are roughly the same. For example, if you want to measure the heart rhythm of a diver 10 meters underwater, then this test should be completed 10 meters underwater. The test results obtained under this condition are true and credible; other test methods, Even if the diver took the test immediately after he landed, the results obtained were deviated from the real data. The reason is simple-the test conditions have changed.

The same is true for RF testing and measurement, and similar things abound. For example, the VSWR of a filter is required to be less than 1.5, the insertion loss is less than 1dB, and the operating temperature range is -30~+60oC. Under the above conditions, the tester will not only use the vector network to test VSWR and insertion loss at room temperature, but also put the tested filter in a high and low temperature box for the same test. The criterion for passing or failing is that the electrical performance indicators meet the requirements within the specified temperature range.

The test of the above-mentioned filter is normal, but a closer examination of the filter's index reveals that there are still some topics that need to be explored.

Our habitual thinking

We list the main indicators of the above filters as follows:

1. Operating frequency range: 118-137MHz;
2. VSWR: not more than 1.5;
3. In-band insertion loss: not more than 1dB;
4. Power capacity: 50W (CW);
5. Working temperature range: -30~+60oC

To describe the above five indicators in another way: in the 118-137MHz frequency range, -30~+60oC temperature range and 50W continuous wave, the VSWR of this filter should be less than 1.5, and the insertion loss should be less than 1dB.

Here comes the question. Some people may say, "We used to measure with a network analyzer. How to measure VSWR and loss under 50W conditions?". Therefore, the description of the above five indicators has been changed to: In the frequency range of 118-137MHz and the temperature range of -30~+60oC, the VSWR of this filter should be less than 1.5, and the insertion loss should be less than 1dB; at the same time, the filter can withstand 50W continuous wave. In this regard, can we call it "selective blindness"? Obviously, this is a habitual thinking for a long time, because it is easy to obtain a high and low temperature box, but it is not easy to build a 50W test environment. But how should we answer the following questions?

? Now that the power capacity is specified as 50W, what are the criteria for judging whether it is qualified or not?

? Since it is required that the VSWR and loss within the specified temperature range must meet the requirements, logically speaking, under the action of 50W continuous wave power, these indicators must also meet the requirements, why are you turning a blind eye?

? In actual use, when 50W continuous wave power is applied to this filter for a long time, what might happen? Is VSWR and loss worse? Or will it cause ignition, breakdown and even failure of the device?

Maybe you will use a 50W filter to "bake" this filter, and then immediately use a network analyzer to test VSWR and loss, but this is like a diver's, the test conditions have changed, and the test results are unreliable.

Below we can cite two more to describe the habitual thinking in RF testing.

Passive intermodulation test conditions

One comes from passive intermodulation testing.

We know that passive intermodulation needs to be tested under specified frequency and power conditions. For example, a typical passive intermodulation index can be expressed as: -153dBc@2×43dBm, 925 and 960MHz. Regarding passive intermodulation testing, our habitual thinking is:

? Passive intermodulation must be tested under the action of two 20W carrier frequencies;
? The magnitude of the carrier frequency and the magnitude of the passive intermodulation are in a 1:3 (dBm) relationship;
? The amplitude of passive intermodulation is related to the operating frequency and must be tested at the relevant operating frequency;
? There is no calculation relationship between passive intermodulation and frequency, that is, the intermodulation value measured in the 900MHz frequency band cannot represent the intermodulation in the 1800MHz frequency band, and vice versa.

This common "selective blindness" phenomenon is manifested in the test power. We can often hear the following statements:

? 2×20W is the test standard for passive intermodulation;
? If the DUT is used in a 2W environment, the test power can be reduced to 2×2W to test its passive intermodulation; if the DUT is to be applied to 1kW, first use 2×20W to test its passive intermodulation, and then Calculate the intermodulation value under the condition of 1kW.

The test power of passive intermodulation 2×20W was originally derived from the nominal output power of 20W of the GSM base station, and now it has become a “standard” generally recognized by the industry. In fact, this is described in the IEC62037 standard:

For mobile communication systems, unless there are other requirements, it is recommended to load 2×20W (43dBm) on the DUT test terminal. Other systems may require different power levels.

This description clearly clarifies that the test conditions for passive intermodulation should conform to the real use environment.

And the power capacity of the cable

2 comes from the power capacity problems of microwave router components such as connectors and cables that we often encounter and have some doubts.

Figure 1 is taken from a microwave mechanical switch and describes the power capacity of switches of different specifications at different frequencies.

Figure 1. Microwave mechanical capacity

We can speculate that Figure 1 may come from some kind of simulation result, it may be an empirical value, and it may be experimentally verified on some frequency bands. However, no convincing experimental data was found to support it.

In addition, Figure 1 also illustrates that the power capacity is related to frequency, which also negates the previous DC substitution method. In the microwave frequency band, voltage and current have lost their exact meaning.

The idea goes back to the previous filter. The same question is what are the conditions for determining the qualification of this device when the power capacity is specified?

We have done an experiment to observe the relationship between VSWR and insertion loss of 0.086" cable assembly under different power and temperature conditions. The test frequency is 900MHz.

Figure 2 shows the change in loss of a 0.16m long cable assembly when loaded with different powers under normal temperature conditions. The greater the power, the greater the loss. Figure 3 shows the relationship between VSWR and ambient temperature under the same power. The higher the temperature, the greater the VSWR.

Figure 2. The relationship between cable input power and insertion loss at room temperature

Figure 3. The relationship between temperature and standing wave at the same power

You might say that the above indicators have not changed much. However, the experimental results show that it has changed after all. What's more, the 0.086" cable manual states that its power capacity at 900MHz is about 170W. Due to conditions, the experiment is only carried out at 100W, who can tell if it is close to its power What will happen in the extreme? All these need to rely on experiments to verify.

Concluding remarks

This article hopes to express the point of view-RF testing and measurement should simulate the real use environment as much as possible, so that the test conclusions drawn will be more practical. With the continuous development of technology, materials and instruments, the construction of various subdivision test systems has become possible. Of course, you can't just ask questions without a solution. In fact, the PM2000 series high-power test platform developed by BXT is an experimental interpretation of the views in this article. We will continue to publish experimental results in the follow-up, hoping to get guidance from our peers. In the next article, we will discuss with you the solutions to the problems raised in this article, so stay tuned.