The BatLogger II

A bat activity logger for the Onset HOBO data logger

NOTE: The deployment phase of the Bat Logger II project has come to a close. The project's goal was to field up to 50 BatLoggers in a 5 year window to determine the usefullness of this type of device for monitoring mines, caves, and hibernal roosts, over protracted periods of time. Onset Computers has announced the discontinuation of the HOBO logger that was used in the BatLogger II ... and so has defined the end-point of the deployment phase of the project. Feedback to date has been positive for the device.

If you are interested in logging bat activity, you might be interested in this new project ... Bat Call Data Recorder

The circuit of the Simple Bat Detector can be adapted to log bat activity on an integrated Onset HOBO multi-channel data logger. This was the basis of the BatLogger II.

The BatLogger II was
designed to be a relatively low cost, compact, rugged, and reliable bat activity data logging system. The logger can record bat activity, air temperature, and even nocturnal light levels.

The low power consumption ( 15ma ) allowed the logger to be operate unattended for months using a 12 volt battery for power.

The ability to log bat activity based on ultrasound comes from the use of a PIC microprocessor that has been linked to the output of a frequency division bat detector circuit.

The PIC data converter continuously samples pulses coming from the detector, and screens out stray ultrasound pulses that are less likely to be related to bat activity.

The PIC accumulates a count of these ultrasound events, and generates a corresponding analog voltage.

The analog voltage is generated using a simple D/A resistor ladder circuit, as shown in the diagram below.

When the HOBO data logger samples and records the voltage data, it generates a handshake pulse that is monitored by the PIC. The PIC resets its accumulators and begins counting anew after each data sample is recorded. This allows the system to be used with different sample times, usually dictated by the data resolution required, the number of bats being monitored, or the length of deployment.



The nature of the voltage data steps is important to understand at this point. To keep the circuitry to the simplest and least power consuming level, the output voltage has a resolution of only 4 bits ... 16 levels. To make this as meaningful as possible, a table of exponentially increasing threshold values is programmed into the PIC. This allows for good sensitivity for a small number of calls, yet allows higher call volumes to also be recorded.

This scale looks like this:

Each level represents approximately double the number of call events as the previous step, following a 1-2-5 progressive sequence. The last step on the scale represents 50,000 call events in the sampling interval.



Below is a sample plot of about 3 months of data, logged at 5 minute intervals. The data is shown as it is formatted by the HOBOPro software. The top graph shows the overall picture of the bat activity, while the lower graph shows an expanded view of 9 days worth of data. One can easily see the pattern of activity ... the initial emergence surge in the early evening, the cyclic evening activity, and the morning return to the roost. Midnight is denoted by the light grey vertical lines in the graph.

Hopefully, these examples will convey what I have felt is the functionality and usefulness of the BatLogger II concept for bat activity logging. Please feel free to email me with any questions or comments.


Tony Messina - Las Vegas, Nevada - email:

Information on the BatLogger II was first published on this page in May, 2010 ... last updated February, 2014