Testing Protection/Clamp Diodes with the NI 655X Digital Waveform Generator/Analyzer

Most integrated circuits (ICs) and electronic devices require circuit protection from voltage spikes, surges, electrostatic discharge (ESD), and other overvoltage conditions. The most common way to provide this protection is with a clamp diode. This application note describes how to use the NI 655X with LabVIEW and NI-HSDIO to test for the presence of such diodes in a functional circuit. Using the method described below, up to 20 channels (40 diodes) can be tested simultaneously.

Description of Test

The presence of clamp diodes can be easily tested by applying a signal with a voltage that is high or low enough to forward-bias one of the diodes. After the voltage is set, a voltage measurement shows whether a diode is present and functioning. For example, Figure 1 shows a typical clamp diode circuit with attached testing equipment.Assume that the forward bias voltage for the diodes is 0.7 V. For the circuit above, when the power supply generates a voltage between -0.7 V (0 V – 0.7 V) and 4.0 V (3.3 V + 0.7 V), then very little current should flow through the resistor. The voltage Vdiodeis approximately equal to Vout. When Voutexceeds 4.0 volts (causing Vdiodeto also approach 4.0 volts), then the current through the upper diode (and the resistor) rapidly increases. You can see this relationship in the steep I-V curve of the diode in Figure 2 (in this case, the top diode is turned on).



The increased current causes a voltage drop across R1. When Voutincreases further, the current will keep increasing, causing an even greater voltage drop across R1. The effect is that Vdiodeis clamped to approximately 4.0 V.

If no diode is present, then the current only increases a small amount because of the finite input impedance of the device under test. The voltage at Vdiodecan continue to increase with Voutsince the small amount of current does not cause a noticeable voltage drop across R1.

Looking at the graph in Figure 3 you can see that as long as both diodes are off (that is, Vdiodeis between -0.7 V and 4.0 V), Vdiodeincreases linearly with Vout. Once Vdiodeis outside this voltage range, one of the two diodes turns on, and current begins flowing through resistor R1 and the diode.

You can see from the information on the graph that to check for the presence of the upper diode, you should set Voutto 5.0 V and then measure the voltage at Vdiode. If Vdiodeis close to 5.0 V, then the diode is not present; if Vdiodeis close to 4.2 V, then the diode is present and working correctly. Note that we have increased the power supply voltage from 4.0 V to 5.0 V. The higher voltage gives us a larger voltage difference between the two plots shown in Figure 3. Note that Vdiodehas increased from 4.0 V to 4.2 V. The reason for this change is that as the current through the diode increases, Vdiodewill increase slightly, as shown in Figure 2.

You can test the lower diode in the same way by setting the power supply voltage to -2.0 V.

Using the NI 655X Device as the Tester

Given the flexibility and accuracy of the programmable voltage capabilities on the NI 655X devices, you can use these devices to test for the correct operation of up to 20 diode circuits in parallel using the techniques described in the previous section. Replacing the voltage source and voltmeter in Figure 1 with the channel electronics of an NI 655X digital I/O channel results in the circuit shown in Figure 4.

The output impedance of the generation circuitry of the NI 655X device is known to be nominally 50 . This resistance takes the place of R1 in Figure 1. The input comparators detect whether Vdioderepresents the red or black plot shown in Figure 3.

You must first program the NI 655X to generate a voltage (Vout) that is sufficient to achieve a distinguishable split between the two plots in Figure 3. In our example, we chose 5 V and -2 V to test the upper and lower diodes, respectively. Since this test does not require high-speed signals, the NI 655X can perform static generation and acquisition operations. Figure 5 shows how you can initialize the static generation channels and configure the generation voltage levels. The generation voltage range of the NI 655X is limited to either -2 to 3.7 V or -0.5 to 5.5 V, so we must use two separate voltage settings for this test. Refer to the NI 655X specifications for a list of valid voltage ranges.


While the voltage range is limited for generation sessions, you can use the maximum voltage range for acquisition. Since the NI 655X has two acquisition voltage levels, you can set each level to the appropriate voltage for testing one of the two diodes. For example, set the acquisition high voltage level to 4.5 V, and the acquisition low voltage to -1.25 V. Figure 6 shows how these two voltage levels relate to the diode voltages.



You can detect the voltage at Vdiodeby putting the NI 655X into Valid/Invalid data interpretation mode for acquisition. In Valid/Invalid mode, data is interpreted as shown in Figure 7. When the signal voltage is between the two thresholds, a logic 1 is returned. If the signal is outside either threshold, a 0 is returned.



Setting the Acquisition Voltage High and Acquisition Voltage Low thresholds to 4.5 V and -1.25 V, respectively, allows the NI 655X to easily differentiate between the red and white plots shown in Figure 6.

Figure 8 shows how to configure the acquisition channels on the NI 655X. You must use the same channels for acquisition as you used for generation.



When you integrate the generation and acquisition sessions shown in Figures 5 and 8, the result should look like the complete LabVIEW VI shown in Figures 9 and 10.

Figure 10 shows the continuation of the LabVIEW code in Figure 9.


Once the generation and acquisition sessions are initialized and properly configured, the test can begin. The test first generates 5.0 V on the NI 655X channels. Next, the test performs an acquisition on the same channels. We adjust the generation voltage to allow the NI 655X device to generate -2 V, and we perform another read. We do not need to adjust the acquisition voltages after the first read because they were already set to accommodate testing for both diodes.

Analyzing the results is simple as well. Keeping in mind the Valid/Invalid mode discussion and the voltage graph in Figure 6, each bit in the U32 data shows whether a diode is present or missing. Since the presence of both diodes causes Vdiodeto never exit the range specified by the acquisition voltage limits, a reading of 1 shows that each diode is present. For example, if the Upper Diode Result is 0xFFFFE, and the Lower Diode Result is 0xFFFF7, then the upper diode on channel 0 and the lower diode on channel 3 are both missing, and all other diodes are present.

Note that this test does not detect whether the device under test has a short circuit between any of its data lines or either of its power rails. It is a good idea to perform a simple input/output test at normal operation voltages to ensure that the data channels are operating as expected.

Reader Comments | Submit a comment »

How much is the current
Hi, I read this application note and I think it is a good idea to perform diodes detection with HSDIO. But I have a question. How much will the current be to flow into the pin of DUT? According to my calculation, it is about 15mA ~ 16mA. It is large and may damage the diodes. Generally the current will not be allowed to exceed 1mA. Do you have any solution to solve this problem?
- Huagang Fang, NexFrontier Solutions. fanghuagang@126.com - Feb 13, 2008