Application of DM3068 digital multimeter in weak signal measurement

This article will introduce some common fault diagnosis methods through several examples such as ultra-low power circuit testing, weak power supply ripple testing, online current measurement, short circuit location, noise analysis, etc., and combines the advantages of RIGOL's new generation 61/2-digit digital multimeter DM3068 in weak signal measurement to introduce some common fault diagnosis methods.

Detecting ground loops

Modern electronic products often tightly integrate small signal analog circuits, digital circuits, and power circuits on a PCB. The circuit layout must not only meet the circuit performance requirements, but also be constrained by the structural design and comply with EMC specifications, all of which pose great challenges to the layout of the ground loop. Under the constraints of multiple constraints, there will be uncertainties in the PCB ground loop wiring during the design phase, which need to be verified during the testing phase.

Taking the circuit shown in Figure 1 as an example, U1, R3, and R2 form a common-mode amplifier. The design expects that the amplifier can amplify the input signal Ui by 10 times, and Uo is the output voltage of the amplifier. Due to the constraints on the PCB layout and wiring, there is a wire Rpcb between the "-" input terminal of Ui and the C end of the resistor R2. However, there may be a ground return current Ix introduced by "circuit X" on the PCB that flows through the wire resistor Rpcb. Affected by the ground return current Ix of "circuit X", there will be a voltage difference Ue at both ends of the wire Rpcb. In order to determine the impact of Ue on the measured signal, it is necessary to measure the size of Ue.

Figure 1: Common-mode amplifier circuit with ground return current interference

For precision circuits, even if Ue is only in the μV range, it will have a huge impact on the circuit. The minimum reading resolution of the DC voltage of the DM3068 digital multimeter can reach 0.1μV, and the measurement terminal is isolated from the chassis ground, so no additional DC current ground loop is introduced, which is suitable for measuring the μV DC error caused by the ground loop.

When measuring weak voltage signals, you need to pay attention to error sources such as thermoelectric potential, common mode interference, and electromagnetic induction. These errors are generally in the order of 10μV and will seriously interfere with small signal measurements. Using twisted or shielded test cables of the same material can reduce the errors of thermoelectric potential and electromagnetic induction. Before measuring Ue, you can first determine the size of the total error introduced by these error sources, use the "relative" operation of the multimeter to eliminate the interference of fixed errors, and then measure Ue to get a more accurate result.

First, connect the two test leads of the multimeter to the terminal C in Figure 1 at the same time. At this time, the reading of the multimeter is caused by the error sources of thermal potential, common mode interference, and electromagnetic induction. Observe its changes. If the reading fluctuates within a small range, it is considered to be a fixed error. Press the "Relative" menu key of the multimeter, the multimeter will record the current reading value, and subtract the reading value from each subsequent measurement result before displaying it, so that the interference of fixed errors can be eliminated. Then measure the voltage difference between terminal C and the "-" end of Ui. The reading is the Ue value after eliminating the interference of fixed errors, which can more accurately reflect the size of the real Ue.

Monitoring power fluctuations

If a circuit module is regarded as a black box, it will have at least one input port - the power supply. In circuit fault diagnosis, the power supply port is often forgotten or underestimated, so that some problems are characterized as "paranormal events".

Assuming that the circuit and signal input inside the black box are normal, if the output of the black box still has problems, then the power input should be checked. Commonly used power detection instruments include oscilloscopes, spectrum analyzers and digital multimeters, which can cover different measurement ranges (as shown in Figure 2). These instruments should be used in combination to fully observe the power signal and avoid testing blind spots.

Figure 2 Typical test ranges of different instruments

It is generally believed that multimeters are DC instruments, oscilloscopes are time domain instruments, and spectrum analyzers are frequency domain instruments, but this distinction is being broken. The new generation of multimeters has introduced time domain measurement functions. The following uses the data drawing function of the RIGOL DM3068 digital multimeter to introduce how multimeters cover the power supply test blind spots of oscilloscopes and spectrum analyzers.

Figure 3 is the waveform of the analog power supply voltage of a certain analog-digital hybrid circuit measured by an oscilloscope. Since the bandwidth of the oscilloscope is very large, most of the waveform is broadband switching noise introduced by the digital circuit, with an amplitude of 8.4mVpp. Generally speaking, the 8.4mVpp power supply ripple and noise meet people's "psychological expectations", so it is believed that there is no problem with the power supply (the power supply impact is underestimated).

Figure 3 Oscilloscope test results [page]

Figure 4 is the power waveform obtained by retesting the power supply voltage using the data plotting function of DM3068. The left side of the figure is the waveform of historical data, and the right side of the figure is the real-time waveform. From the real-time waveform, we can clearly see a sine wave with an amplitude of about 4.4mVpp. Further calculation shows that the sine wave frequency is about 50Hz. Such a strong 50Hz signal will cause great interference to the precision circuit.

Figure 4 Multimeter test results

The spectrum analyzer is limited by the frequency measurement range and frequency resolution, so it is difficult to find this 50Hz power supply interference. The high speed, high precision, low noise and strong high frequency suppression of DM3068 in low frequency time domain measurement just make up for the shortcomings of oscilloscopes and spectrum analyzers, and help reveal the truth of "paranormal events".

Using histograms to find hidden interference

When the signal/interference is extremely weak and submerged in the circuit's own noise, it can be exposed with the help of histogram statistical analysis methods.

DM3068 has real-time histogram statistics function, combined with low noise and large dynamic range characteristics, which is helpful for testing weak signals and interference.

Figure 5 is an example of using a histogram to observe a signal that is submerged by background noise. The left side of the figure is the time domain waveform of the circuit background noise (below, the vertical direction is the time axis direction. Same below.) and its histogram. The noise basically conforms to the Gaussian distribution and is considered to be white noise. The right side of the figure is the test result after adding a pulse square wave of about 3μVpp to the circuit. Comparing the time domain waveform, the signal waveform on the right is very similar to the white noise waveform on the left, and the voltage average value is also close, so the difference between the two waveforms cannot be intuitively determined. However, by comparing the histograms of the two, the difference between the two signals can be clearly found, and the histogram on the right can be inferred that the added signal has a low-level component, and the probability of the low-level component appearing is not high, which is similar to a negative pulse with a very small duty cycle.

Figure 5 Histogram showing submerged signals

Voltage and current testing of ultra-low power circuits

Ultra-low power circuit testing usually requires the instrument to be able to test nA-level weak currents, and the input impedance of voltage measurement tends to infinity. General handheld multimeters cannot measure nA-level currents, and the input impedance of voltage measurement is fixed at 10MΩ, which cannot meet the testing requirements of ultra-low power circuits.

Figure 6 is an intrusion detection circuit for an ultra-low power device. The normally closed switch S1 is used for intrusion detection. When the device housing is damaged, the switch S1 is disconnected. In this circuit, the diode D1 is used as an ultra-low current pull-up element, and its reverse leakage current Is is about 10nA. Once the housing is damaged, S1 is disconnected, and D1 pulls up the pin DET of the controller MCU, generating a rising edge as an intrusion trigger signal. The main test items of this circuit are the diode reverse leakage current Is, the DET level when the switch S1 is closed, the DET level when the switch S1 is disconnected, and the voltage rising edge waveform of the DET pin during the process of switch S1 closing to opening.

Figure 6 Intrusion detection circuit

Conventional instruments cannot effectively complete the above tests. The DM3068 digital multimeter has a minimum DC current resolution of up to 100pA, which can meet the test requirements of Is; the DC voltage 20V (the range is twice as large as that of competing products) and below has an input impedance greater than 10GΩ, and the input bias current is less than 100pA. Combined with its data plotting, level triggering and pre-triggering functions, it can capture and display the DET pin waveform in real time, and can easily complete the rising edge waveform and level test like using an oscilloscope. [page]

Finding Shorts in Circuit Boards

Manually soldered circuit boards often have short circuits caused by soldering shavings, and soldering shavings are usually hidden at the bottom of components and are difficult to find. Once the power supply and ground on the circuit board are short-circuited, all components connected between the power supply and ground become suspects. Checking one by one can solve the problem, but it is very laborious.

If there is only one short circuit on the short-circuited power supply, the position far away from the short circuit point has a larger resistance to ground due to the series connection of the PCB resistance. Therefore, the short circuit can be located by finding the position with the smallest resistance to ground.

As shown in Figure 7, Rp1~Rp(n) are the PCB resistors of the +5V power line, and the resistance value is 1mΩ; Rn1~Rn(m) are the PCB resistors of the GND ground line, and the resistance value is 1mΩ; C1~C5 are the decoupling capacitors of the +5V power supply. Assuming that there is a short circuit hidden under capacitor C2, the resistance measured at C2 is 0mΩ; the resistance measured at C1 plus the resistance measured at C2 plus Rp1 and Rn1 is 2mΩ in total; for the same reason, the resistance measured at C3~C5 is 2mΩ, 4mΩ and 6mΩ respectively. The resistance measured at C2 is the smallest, so it can be concluded that there is a short circuit under C2.

Figure 7 Equivalent circuit with short circuit

The higher the resistance measurement resolution, the higher the accuracy of short circuit positioning. PCB resistance is generally in the milliohm level, and most handheld multimeters have a resistance measurement resolution greater than 10mΩ, which cannot effectively determine the location of the short circuit. The resistance measurement resolution of DM3068 is 0.1mΩ, which can accurately locate the short circuit (for 1OZ thick, 5mm wide copper wire, it can be resolved to 1mm), making the above short circuit positioning method practical.

Online current measurement

It is not easy to measure the working current of a QFN or BGA packaged chip on a circuit board. It is difficult to find a place to cut the power line and insert the ammeter on a high-density multi-layer PCB; some digital chips require extremely low power supply internal resistance and cannot tolerate the insertion of test cables. At this time, if there are some breakthroughs left on the circuit board, the DM3068 low resistance measurement and weak voltage measurement functions can be used to achieve non-invasive online current measurement.

As shown in Figure 8, the VCC current Ivcc of chip U1 needs to be measured, and its current direction is from point A to point B. You can first disconnect the power supply of the circuit, and then use the resistance measurement function of DM3068 to measure the PCB wire resistance between points AB, then turn on the power supply and measure the voltage between points AB, and finally divide the measured voltage by the measured resistance to get the current. For example, if the PCB wire resistance is 4.8mΩ (1oz thick, 20mm long, 2mm wide wire), the measured voltage is 48μV, and the current is 10mA.

Figure 8 Schematic diagram of online current measurement

Online resistance measurement

Strictly speaking, online resistance measurement is not recommended, but frequently disassembling and installing resistance during circuit board debugging is indeed a tedious task. By analyzing the circuit during circuit debugging, you can find the conditions for online resistance measurement. Online resistance measurement only proves that the resistance value is the same as the expected value, so there is no need to worry about inaccurate measurement.

Taking the circuit shown in Figure 9 as an example, resistor R is connected in series between the output of logic IC1 and the input of logic IC2. The correct value of resistor R is 33Ω. It is suspected that the resistance value of R is abnormal and needs to be tested. By observing the internal equivalent circuits of logic IC1 and logic IC2, it can be found that logic IC2 only connects resistor R to the power line through a clamping diode. In other words, as long as the voltage across resistor R does not exceed the forward conduction voltage of the clamping diode inside IC2 (usually 0.5V), the current flowing through the input pin of IC2 can be ignored, and will not affect the measurement of resistor R.

Figure 9 A special case of online resistance measurement

Before testing, cut off the power supply of the circuit board and ensure that the power circuit is completely discharged. DM3068 uses the method of adding constant current to measure voltage to measure resistance. Check the manual of the multimeter to know the size of the current source used in each resistance range. For example, the 2kΩ range of the meter uses a 1mA current source. If the resistance R to be measured is normal, the voltage drop across it is 33mV, which will not turn on the IC2 clamping diode, so it can be measured correctly online.

Most semiconductor devices have similar diode isolation structures, so the scope of application of this method can be expanded according to the actual circuit conditions.

Simply put, you should choose a range that is 5 times larger than the resistance value of the resistor being measured to measure the resistance online. This method cannot be used when there are devices in the circuit with a reverse breakdown voltage less than 5V.

Conclusion

Although the challenges of debugging and verifying analog circuits are growing, the new generation of test and measurement tools have also made great progress in performance and functionality. Taking full advantage of the excellent performance of the tools can greatly improve the efficiency of fault diagnosis. DM3068 has unique advantages in the performance of weak current, weak voltage and small resistance measurement, and provides advanced analysis functions such as real-time data plotting, real-time histograms, internal triggering, etc., which can help ultra-low power and precision analog circuit verification and fault diagnosis.