Data Analysis Fundamentals: Logging Rate
Watch This Course
$129 USD $64.50 USD
-OR-
Or
8 easy payments of only $8.06
Instant access. Easy checkout. No fees.
Logging Rate
03.38
| 00:00 | - One of the decisions you'll need to make about each channel you want to log is what logging rate or logging frequency you should use. |
| 00:08 | This simply defines how often the logger will collect a sample from the sensor and store it in the log file. |
| 00:15 | The logging rate we choose will affect the accuracy of the data we're collecting as well as the amount of logging memory we will fill up during a track session. |
| 00:24 | Usually the logging rate or sample rate as it's sometimes referred to is expressed in hertz using the abbreviation, capital Hz. |
| 00:34 | A logging rate of 10 Hz means that the datalogger is collecting 10 samples every second or in other words one sample every 0.1 seconds. |
| 00:43 | The next obvious question then is which logging rate should you use? And the answer depends on the type of sensor you're logging. |
| 00:52 | What we really need to consider is how quickly the channel is likely to change. |
| 00:57 | For example a temperature sensor like an air temperature or engine coolant temperature isn't likely to change very quickly so it's fine to log these at a relatively low rate. |
| 01:08 | Normally in the region of 1-2 Hz. |
| 01:11 | If we were logging damper position on the other hand, then the damper is going to be changing position very rapidly and we're going to need to sample the input at a much higher rate. |
| 01:22 | If we look at some of the other typical sensors such as engine RPM, throttle position, brake pressure or wheel speed, generally a logging rate of around 25 Hz is a pretty good starting point. |
| 01:34 | If our logging rate is too low we can miss important data that can be misleading. |
| 01:39 | For example, if we're logging a channel like damper position and we're logging at a rate of 2 Hz, the car could traverse a large bump at high speed and the resulting damper movement could be completely missed in the log file if the bump occurs between two adjacent samples. |
| 01:57 | On the other hand, logging at frequencies that are higher than necessary generally does nothing to improve the quality of your logged data. |
| 02:04 | Instead, your log files will become unnecessarily large which can be limiting depending on the amount of internal storage you have inside your logger. |
| 02:13 | While most logging systems will allow you to define the logging rate for individual channels and select something appropriate, that's not always the case and some systems will log every channel at the same rate. |
| 02:26 | In this case you'll need to select a logging rate that's going to suit your faster moving sensors and accept the fact that you're over sampling slower moving channels. |
| 02:35 | It can be quite easy when viewing log data to forget about the logging rate. |
| 02:39 | Particularly if you're analysing data from a logger you haven't configured. |
| 02:44 | So it's worth just checking on the logging rate. |
| 02:47 | You'll usually be able to check that quite easily by clicking on the particular channel. |
| 02:52 | For example, here in MoTeC's i2 software, we can right click on the channel we're interested in and it will show the sample rate. |
| 02:59 | In the case of the wheel speed channel here, we can see it's 20 Hz. |
| 03:04 | To see the effect of logging rate, we can also press the S key which will show us the individual samples for each channel rather than a nice linearised line connecting the samples together. |
| 03:16 | When we do this, it becomes easier to understand that the channel could have varied wildly between two of the individual samples and we would have no idea it occurred just by looking at the linearised plot. |
| 03:28 | Included as an attachment to this module is some recommended logging rates that will provide a suitable starting point for most typical sensors. |
