How to capture intermittent signals

Finding occasional missing pulses in a signal using an oscilloscope

This page is about using automotive oscilloscopes covers how to "capture" waveforms that do not occur in a repeating and predictable way.

Sometimes during the automotive diagnostic process, it is necessary to check that mechanical position sensors are working correctly. Typically, the signals from these sensors will have voltage levels between 0V to 4V.

In the example below, we will investigate a digital crankshaft position sensor (CKP-Hall effect).

Typical signal of a properly operating sensor is shown in Fig.1

Fig.1

 Mechanical failure and / or incorrect (critical) adjustment of the air gap between the sensor and the gear may lead to a situation where one of the pulses is missing, but this can happen quite intermittently and not at every crankshaft revolution.

 We will look at two methods for detecting this type of malfunction.

• No triggering (no synchronization)
• Single triggering (single synchronization)

 Note: The selected time base values below are for reference only, the actual settings will vary depending on the number of teeth on the reluctor and on the engine speed during the measuring process.

 1. No triggering (no synchronization)

This method is applicable to oscilloscopes with large buffer memory (big memory depth). Most commonly these are the PC based oscilloscopes. They can use the huge computer resources with almost no limits.

 In this method we set the oscilloscope as follows:

• Trigger type: off or on (does not matter)
• Time base (time for one screen): 2 seconds
• Voltage range: 5V

 Start the measurement and wait for 2 seconds or more, then stop the measurement, we will then see the following:

Fig.2

 Because there are too many pulses and we are not sure if there is any pulse gap, we will use Zoom in by selecting a window that allows us to see the signal in more detail, for example, a 20mS time slice.

Fig.3

 If we don’t see a missing tooth in the selected window, we need to move the window to a new position. This process should continue until we browse through the entire signal or detect a missing pulse (tooth).

Fig.4

If we browse through the entire signal and don’t find a missing pulse (tooth), we need to do further testing by re-measuring the signal (2sec) and then browsing and zooming again until we find a missing tooth or until we see that such pulse does not exists in a certain (enough) number of measurements (2sec). This method is not always good, how would we feel if we had to flip through 1000 screens, for example.

 2. Method with a single triggering (single synchronization)

Single-shot capture method is more appropriate when the oscilloscope’s memory depth is not "infinite". This is the case in almost all oscilloscopes that don’t have a USB connection to the computer as these don’t rely on the huge hardware capabilities of the computer, and typically they have limited buffer memory.

It is worth noting that this method can also be applied to oscilloscopes with large buffer memory.

To use single triggering , you set the oscilloscope, so that it will wait and capture a screen only when the trigger conditions that you define have occurred. Any event that triggered the oscilloscope will have been captured and frozen on the screen.

In this method, we set the oscilloscope as follows:

• Triggering (synchronization) type: Single
• Sync front: Falling
• Sync level: 2V
• Time base (time for one screen): 20mS
• Voltage range: 5V

After the measurement starts, the oscilloscope will automatically stop after 20mS and we can check if there is a missing tooth on the screen.

If it is missing on the captured screen, we just need to run a new measurement and make new visual check. This process continues until we find a missing pulse (tooth) or until we make sure that such pulse does not exists in a certain (enough) number of measurements.

Observing intermittent signals

 This type of signal could be randomly occurring single pulses or packets of several pulses that appear at irregular intervals over time. A good example of such short packets containing multiple pulses, are the communication signals.

Fig.5

If we have a very large buffer memory oscilloscope, we can make a long record, stop the record, scale, and then look at the recorded measurement by browsing through many screens one after another.

A significantly more elegant method is by using a single trigger (single synchronization). When the oscilloscope is started the trigger waits for an event to occur. After the event occurs, it activates, writes one screen and stops. If we want, we can run the measurement again.

Fig.6

This is the only applicable method in car diagnostics when, for example, a random pulse that could occur within one or more minutes, is expected.