When using an oscilloscope the sample rate, timebase and waveform buffer must all be configured to allow the technician to review the captured data in detail with enough repetition for accurate diagnosis. This article looks at sample rate selection – the examples use a Pico Scope but other oscilloscopes have similar functionality.
How to determine the optimum sample rate?
The optimum sample rate is the rate which you can collect and store the required data in sufficient detail with enough repetition that is necessary to diagnose the circuit/component being tested.
Isn’t more always better?
The problem with selecting a high sample rate is the waveform buffer will fill up very quickly and your file size will increase to the point where you may not be able to share, upload or email files. Oversampling can also result in excessive noise.
The Waveform Buffer…
The Pico Scope Waveform Buffer can hold 32 Million Samples, so if you want to record for longer periods of time then you need to adjust the sample rate and change the default number of 32 waveforms.
To adjust the maximum number of waveforms that can be stored in the waveform buffer, you need to select Tools > Preferences:
Then change the number in the maximum waveforms box to 100+ using the keyboard or up down icons.
Default Sample Rate…
Fast signals, ones that occur many times a second, such as injectors and crank & cam sensors require a greater sample rate than slower signals, such as air mass meters.
Most oscilloscopes allow the user to adjust the sample rate; some may automatically adjust the sample rate based on the timebase selected.
The Pico Scope has a default sample rate of 1 million samples per frame. This can be used in the majority of cases, but there are many exceptions.
Setting the Sample Rate…
The Pico Scope allows the user to select the maximum number of samples using a sample rate control function in the capture set up tool bar.
This control box sets the maximum number of samples that will be captured for each channel. The actual sample rate is displayed in the properties tab and may be different from the number of samples selected, depending on the timebase. To view the properties tab, right click in the main screen and select View Properties.
Adjusting the Sample Rate to suit timebase and test requirements…
Taking an injector as an example, you may choose a timebase of 20 ms (2ms per division) to capture single injection events during snap acceleration testing.
Although the number of samples selected is 1 MS, the actual sample rate is 40MS/s filling the buffer in just a few seconds.
The test takes around 10-15 seconds to complete; snapping the throttle and allowing the engine to accelerate, decelerate and return to a stable idle. So the sample number is too high in this case. A more suitable sample number would be 100kS, which would capture around 13-14 seconds of data with the same time base.
The properties tab indicates a sample rate of 5 MS/s so it is still capable of showing significant detail in the waveforms, while capturing over 100 waveforms.
What if you needed to record an Air Mass Meter output during repeated WOT acceleration runs in third gear?
Each test will take around 20-30 seconds, so you need to ensure the waveform buffer is capable of storing the data for the entire test; around 90 – 120 seconds worth of data. The sample rate chosen would need to take into account the new timebase of 50 seconds, 5 seconds per division. A sample number of 5MS results in a 100 kS/s sample rate and minutes’ worth of data.
The same test carried out with a sample number of 20MS and a timebase of 20 seconds (2 seconds per division), results in the default 1MS/s sample rate. The extra noise from the excessive sample rate is clear to see and the file is 4 times larger than the previous example.
Depending on the type of signal you are capturing, the number of waveforms you require for sufficient repetition and the time taken to carry out the test, you will need to change the sample number to achieve your optimum sample rate. Check the actual sample rate in the properties tab, as it will change if you alter the timebase during testing.
Remember to increase the size of the waveform buffer, sometimes 32 waveforms are not enough. Parasitic draw testing is a classic example of this.
To find out more visit www.gotboost.uk – here you will find course dates, training news and helpful video guides.