Dynamic Logging Examples

As I mentioned in an earlier post, the Butterfly Logger now has a dynamic logging feature. This post elaborates on the benefits and costs of the feature using some data gathered during testing. Some examples of the output while using the feature is shown below in figures 1 and 2. In processing the output from the logger I created a couple of scripts to analyse the data and create graphs. These are included below as listings 1 and 2. While the new feature allows us to log for longer without the risk of missing events, it also costs us in no longer being able to predict how long a logger can be deployed for in the field.

A Graphical Overview

Figure 1: Dynamic logging example.

The two part graph shown above in figure 1 demonstrates the increasing and decreasing of logging frequency as the monitored signal changes more or less quickly. The signal being monitored in this example is the ambient temperature of my garage over a week in March 2011. This was done using the AVR Butterfly’s onboard thermistor.

The upper part of the graph shows the temperature plotted in order sampled as it is stored in the memory of the logger. The lower part of the graph plots the temperature against time. This is done using the date and time as recorded by the logger and using the following lines in gnuplot:

set xdata time
set timefmt "%Y/%m/%d %H:%M:%S"
plot 'dataFile.log' using 1:8 with impulses notitle

The temperature in the lower section is plotted as impulses simply to highlight the change in logging frequency.

Figure 2: Comparison of data plotted against time or sample index.

The above graph in figure 2 shows a much larger sample from the same logging session. In this plot you can again see the differences between the samples recorded in memory and the samples plotted against time.

You will notice how the samples actually recorded only ever differ by a fixed amount. This is evident through the constant slope of the upper section of the graph. This constant slope, either increasing or decreasing, is an artefact of the logging threshold value. The threshold value is set at compile time in the software.

The software also has a timeout parameter that establishes the maximum logging interval. If no timeout is set the system will only log when the threshold has been exceeded, however if the parameter has been set then the logger will record a sample after a fixed period time (even if the sample threshold has not been exceeded).

The minimum interval is specified by the sample period parameter as used when logging in the standard mode. The timeout value is simply specified as a multiple of this interval.

Figure 3: Dynamic resolution of the data expressed as number of samples per hour.

Figure 3 is yet further analysis on the same data from the previous two graphs. This plot shows the logging resolution over time. The resolution is expressed as the number of samples recorded per hour. This figure was calculated by processing the data and counting the number of samples recorded that hour. Where no data is recorded for a given hour a zero is recorded. A Perl script to produce this data from the logger output is shown below in listing 2.

This example gives us a good metric for measuring the effectiveness of dynamic logging. If you examine the graph you can see that, at its peak rate, the logger is recording at a resolution of 30 samples per hour. The logging interval for this session was set to once per second, giving us a maximum sample rate of 3600 samples per hour. As the rate is dynamic it makes sense to also look at the average rate over the whole period, which is just over 2 samples per hour. If we compare the average rate to the peak rate we can gauge the efficiency introduced by the dynamic logging feature.

TABLE 1: Numerical summary
Number of hours monitored: 168
Number of samples recorded: 350
Minimum sample rate: none
Maximum sample rate: 3600 samples per hour
Average sample rate: 2.1 samples per hour
Samples needed if using peak rate: 5040
Samples needed if using maximum rate: 604,000


The benefits of dynamic logging

By looking at the number of samples needed at the peak sample rate and comparing this to the number of samples used, we can calculate the theoretical saving in redundant samples. If we were to capture the same data using standard methods we would have used over 14 times more samples. As our storage space is limited, if we were using a standard technique we could only log for a much shorter time.

This reduction in space used can be taken advantage of in two ways. If the temperature characteristics in the garage remain unchanged the logger could record for 3 years without filling up the flash. This is much much more than possible using the standard technique. Another way to utilise this extra space would be to log more sensors over the same period as possible with standard logging.

Other than storage savings the major benefit of the dynamic sample rate is that dramatic events are not missed. If we had wanted to conserve storage space using standard methods we would risk the possibility of sudden changes in the data being overlooked by a much slower sample rate.
Using a dynamic rate we get the benefits of a higher sample rate without the storage cost.

The costs of dynamic logging

There are a couple of disadvantages to this dynamic sample rate technique. One of these is the reduction in resolution in our sensors. This is due to the threshold introduced before a sample is taken. The resolution of the ADC has effectively been reduced from 1024 to approximately 340 levels. For our example this reduces the temperature to a resolution of 0.1°C per level in the temperature range we are looking at. (< Due to the non-linear nature of the thermistor, the effective temperature resolution will change of over the range of the sensor >)

The logger will also consume more power due to being out of sleep mode more often. The logger needs to leave sleep mode often to check if it should be recording a sample to flash. Compared to just reading of the sensorsWriting to flash is by far the most power hungry activity of the logger when c, because of this I don’t anticipate the extra power requirements to be too great, although I have not actually measured the impact.

When using a fixed sample rate technique you can calculate the exact length of time a logger will be able to log for prior to deployment. With the dynamic sample rate you can no longer know how long you can log for when deploying a logger. To mitigate this issue you could either make a prediction based on prior knowledge or develop a method to alert the user when the memory is approaching capacity.

Using data previously gathered in an environment you can predict how long it is reasonably likely to log for. Using the example data used I would calculate the expected logging capacity based on 700 samples a week (< Doubling the average sample rate we saw in the data simply for reasons of contingency. >). Using this value would allow logging for something like 80 weeks or so. This is of course still 7 times longer than standard logging would allow. These types of calculations could potentially be incorporated to the firmware to make using dynamic logging more useful.


As promised above I have included the scripts used to produce the data and graphs. They might be useful to anyone getting started with using Gnuplot, BASH and Perl to automate graphing and analysis.


# script to plot logging results temperature against time.
# also plot temperature samples for comparison and analysis of
# the dynamic logging system.

#assume arg 1 is name of file with ^M's ^D's and non data lines already removed.


# look at number of points per hour
#cut -d\  -f2 ${1} | cut -d: -f1 | uniq -c > ${1}_res

# TODO: need to pad out hours with zero points
# DONE: use pad.pl to calculate number of samples per hour and pad any zeros
# 	Does not check for an entire day with out samples though.
cat ${1} | ./pad.pl > ${1}_res

gnuplot << EOF
set terminal png size 1024,768 enhanced font "/Library/Fonts/Microsoft/Arial,12"
set output "$1_resolution.png"
set origin 0,0
set grid
set yrange [0:35]
set title 'Logging resolution over time'
set ylabel 'No. of samples per hour'
set xlabel 'Hour of sampling' 
plot '${1}_res' u 2 w i  not

set output "${1}_comparison.png"
set multiplot
set origin 0,0.5
set size 1,0.5
set xdata 
set yrange [0:${maxtemp}]
set title 'Ambient Temperature'
set xlabel 'Sample'
set ylabel 'Temperature (°C)'
plot '$1' u 8 w l not

set origin 0,0
set size 1.0,0.5
set xdata time
set timefmt "%Y/%m/%d %H:%M:%S"
set ylabel 'Temperature (°C)'
set yrange [0:${maxtemp}]
set xlabel 'Time'
set title ''
set format x "%d %b"
plot '$1' u 1:8 w l not

set output "${1}_impules.png"
set multiplot
set origin 0,0.5
set size 1,0.5
set xdata 
set yrange [0:${maxtemp}]
set title '150 Samples of Ambient Temperature'
set xlabel 'Sample'
set ylabel 'Temperature (°C)'
plot '< head -150 $1' u 8 w l not

set origin 0,0
set size 1.0,0.5
set xdata time
set timefmt "%Y/%m/%d %H:%M:%S"
set ylabel 'Temperature (°C)'
set yrange [0:${maxtemp}]
set xlabel 'Time'
set title ''
set format x "%d %b"
plot '< head -150 $1' u 1:8 w i not

open $1_resolution.png
open $1_comparison.png
open $1_impules.png

Listing 1: BASH script to process the data into graphs.

#!/usr/bin/perl -w

use strict;
# simple script to process temperature logs and show number of samples per hour

# my old script did this...
## look at number of points per hour
##  cut -d\  -f2 ${1} | cut -d: -f1 | uniq -c > ${1}_res
# .. but that didn't account for hours with no samples at all.

# my variables...
my @lines;
my $line;
my $hour;
my $previous_hour;
my $first = 1;
my $count = 0;
my @fields;
my @time;

# read in all lines from stdin
chomp(@lines = <STDIN>); 

#process each line in turn
foreach $line (@lines) {
		# extract the hour value from the time, ignoring the date.
		@fields = split /\s+/,$line;
		@time = split /:/,$fields[1];
		$hour = $time[0];
		# set up the previous_hour for the first line to enable the count
		if ($first == 1){
			$previous_hour = $hour;
			$first = 0;
		if ($hour == $previous_hour){
			#increment the count of samples for this hour 

		} else{
			#  print our total and move on to the next hour
			print "$previous_hour \t $count \n";
			$count = 0;
			$previous_hour %= 24;
			# check for non consecutive hours and pad out with zero totals
			# assuming their hasn't been a total day without samples
			while ($previous_hour != $hour){
				print "$previous_hour \t $count \n";	
				$previous_hour %= 24;
			# remember to count this first new sample in our totals 

# done

Listing 2: PERL script to process the data logger output and calculate the number of samples per hour.

Butterfly Logger Firmware 0.31A

I’ve just completed another release of firmware for the AVR Butterfly Logger project. It can be download it at the project site on Sourceforge.

New Features

The main feature of this update is dynamic logging. I’ll post some examples of this in action soon but for now I’ll just give you a basic overview.

Dynamic logging means that the period between samples is altered dynamically at runtime depending on the state of the system being monitored. If the system is undergoing rapid change then many samples will be taken automatically. If the system is stable then very few samples will be taken at that time. This has the same effect as sampling continuously at a high sample rate but removing any repeated or redundant data except that the repeated and redundant data is simply never recorded.

When using the dynamic logging feature you will need to log time and date to enable you to be able to properly reconstruct the data later once downloaded.

The dynamic logging feature also has an optional timeout value so that if no change has been read from the ADC for a large number of log attempts then a log is forced to ensure a minimum resolution such as 1 sample every hour is maintained.

Currently the dynamic logging feature only monitors a single ADC channel. I have plans to extend it to monitor other sensors and also multiple sensors simultaneously in future releases.

Setup and options for the dynamic logging features are currently maintained in the file main.h. There are some comments in the file to explain the options available.

To save on program space the dynamic logging parameters can be implemented as a compile time option. If more flexibility is wanted the features can also be implemented with the parameters as runtime changeable options that can be altered via the serial port interface.

Bug Fixes

This version also includes a fix for an issue with updating the LCD readings after the flash was full.

Butterfly Logger Firmware 0.30C (bugfix)

Last week I released another firmware update for the AVR Butterfly Logger. This was a bugfix to the thermistor routines for calculating negative temperatures from the onboard temperature sensor.

You can download the source code from SourceForge.

The thermistor routines in the logger project were adapted from Martin Thomas’ port of the Atmel AVR Butterfly example code. This routine compares the ADC input from the thermistor sensor to a table of known readings. The table () contains a single entry for each degree celsius from -15°C to +60°C. The routine would simply step through the table until the ADC reading exceeded ( negative temperature coefficient (NTC) device. This means as the temperature increases the resistance drops. In our electrical configuration of the sensor this also applies to the voltage seen at the ADC input.>) the value in the table. The position in the table is then given as the temperature in degrees celsius.

When first developing the data logger in 2004 and 2005 I quickly wrote an interpolation routine that would extend the integer value given in the original with an extra two decimal places. After finishing writing the code I reviewed both the code and the corresponding output. At the time of that brief review everything appeared to make sense. As the original project had other more accurate temperature sensors available, the operation of the onboard temperature sensor was not critical. The onboard temperature sensor was instead intended as a fast secondary sensor for live feedback to the user via the LCD and so the code was not widely tested. Whoops.

This December (being the coldest on record in the UK for the last 100 years) showed up some of the flaws in my code. After some prompting from a couple of users of the project I set aside some time to investigate further. My investigation involved writing a test routine to simulate all possible sensor readings and then evaluating the results. This was an incredibly simple exercise, just a few lines of code and then plotting the results with Gnuplot. The output from the function is shown below in Figure 1.

Figure 1: Thermistor routines test output

This figure shows the full-scale results for the original and fixed routines along with the data points given in the original source. As you can see the interpolated line appears to follow the datapoints quite nicely until it drops below zero degrees.

At this point the linear interpolation is way-off and producing mostly garbage (you can see this clearly in the close-up below in Figure 2)

Figure 2: Close-up of the thermistor routines test output

Looking at the close-up in Figure 2 you can see the awful job the routine does when dealing with negative temperatures. If you look very closely at the positive temperatures you will also note a small error in that calculation as well. After plotting the output and going over the code it was very easy to spot the couple of errors. After the corrections the graph of the output (The green line in Figures 1 and 2) intersects perfectly with the data points given in the source code. If it were not for visual inspection of the results (via the graph) I don’t think I would have noticed this second error at all.

Further changes were also made to the routine so that when the readings are outside the range specified by the data table then they will record an over or under temperature error. This is signified in the log as +++.++ or ---.-- for over temperature or under temperature respectively. On the LCD, +++.+C or ---.-C is displayed. Previously the routine had just returned the maximum or minimum temperature from the table.

It had been at the back of my mind that there was something not quite right about that bit of code for some time. It was so simple in the end to write a routine to test the function with all possible inputs that I could kick myself for not doing it earlier. By plotting the results on a graph it became obvious how the code was misbehaving. The key lessons I’ve learned from this it is that a little testing goes a long way and that some simple visualisation can really help you understand a bug quickly.

More information about the AVR Butterfly Logger project can be found at the project website.

Any questions can be asked in the project forums.

Butterfly Logger Firmware 0.30

In an unprecedented move I’ve released another version of the firmware for the AVR Butterfly Logger within a few months of a previous release. This was a minor change just adding MAX6675 based J and K thermocouple support.

You can download the source code from sourceforge, with the changes from the previous version are listed below.

Changes summary:
- MAX6675 K-Type Thermocouple logging (adjustable averaging)
- MAX6675 available on 'T' via the USART.
- MAX6675 available on LCD via 'THERMOCOUPLE' Menu. 

The new library has configurable support for averaging readings but each individual reading takes 200ms so an 8 point average will take 1.6s to complete. If you use this averaging feature please make sure that your logging interval is long enough to ensure enough time for the complete reading to finish. While thermocouples are capable of measuring wide ranges of temperature both positive and negative the MAX6675 is limited to returning readings between 0°C and +1024°C. If the chip detects an open thermocouple then the system will log ‘-1’ as the temperature.

The MAX6675 connects to the AVR Butterfly via the SPI bus. This bus is also used by the onboard Dataflash and the Kionix acceleration sensors (when in use). As this bus is already available the only additional pins required is a single chip select (CS). The library currently only supports a single thermocouple input but the code could easily be expanded to suport multiple MAX6675s requiring a single CS line for each additional chip and thermocouple.

The SPI bus is available on PORTB and also via the ISP connector. An example breakout PCB for the MAX6675 was published here. Default wiring from the this PCB to the AVR Butterfly Logger is given below:

PCB MAX6675 Butterfly J403 (ISP) J405 (USI)
1 1 GND GND 6 4
2 7 MISO PB3 1
3 6 CS PE6 3
4 5 SCK PB1 3
5 4 VCC VCC 2

More information about the data logger project can be found at the project website.

Any questions can be asked in the project forums.

Butterfly Logger Firmware 0.29

I’ve finally managed to get another release of the Butterfly Logger firmware done. I was supposed to have released a lot of this a couple of years ago but it got lost in the ether somewhere.

Without any further ado the source code is here, with the changes from the previous version are listed below.

Changes summary:
- Added  'j' and 'J' commands to read SHT75 Humidity and Temp.
- Fixed the SHT75 clock waveform (now it looks  symetrical )
- Added ENABLE_SHT_CALCS option to print/store real world values for SHT75
- Tuned vref for more accurate results. (Battery voltage and dependant SHT75 calculations)
- Added Kionix accelerometer logging
- Fixed vref ( done by matthias.weisser )
- Make file uses no-inline-small-functions ( done by matthias.weisser )
- Direction ADCs are now refered to as Auxilliary ADCs

More information about the data logger project can be found at the project website.

Any questions can be asked in the project forums, here.

How to fix LCD contrast problems on the AVR Butterfly

I had heard a number of reports of contrast problems on the AVR Butterflies produced over the last year. Unfortunately I haven’t had an opportunity to look into this until now. This week while checking some MP3 player code on a new AVR Butterfly I ran into the problem for myself. Luckily it was a simple fix.

A quick search on AVR Freaks turned up the following solution.

Turns out adding the following to lcd_driver.c fixed the problem:

//updated 2006-10-10, setting LCD drive time to 1150us in FW rev 07,
//instead of previous 300us in FW rev 06. Due to some variations on the LCD
//glass provided to the AVR Butterfly production.
LCDCCR |= (1<<LCDDC2) | (1<<LCDDC1) | (1<<LCDDC0);

I inserted these lines as instructed into the file LCD_driver.c at the end of the LCD_Init() function just after gLCD_Update_Required = FALSE; (this is around line 170 for ButterflyMP3).

The change has fixed the problem on my hardware. Although I have updated the files in the CVS, I have not done a general release for either ButterflyMP3 or Butterfly Logger. You can get the modified files directly from the CVS system, here for ButterflyMP3 and here for Butterfly Logger

In addition to the changes given above in the post at AVR Freaks, the macro around line 45 in the file LCD_driver.h

#define LCD_CONTRAST_LEVEL(level) LCDCCR=(0x0F & level)

should be changed to something like

#define LCD_CONTRAST_LEVEL(level) LCDCCR=(0xF0 & LCDCCR) | (0x0F & level)

This change will ensure that the drive time setting does not get erased when using the macro to set the contrast at some time after initialisation. I have not tested this as none of my firmware sets the contrast after startup.

Why are these changes needed?

It appears that the LCD’s characteristics have changed and it now requires a much longer drive time of 1150μs. The original LCD needed only 330μs. I’m unsure of the effect of this on the overall power consumption other than it will increase the LCD’s portion of it.

If the original 330μs drive time is used, then some of the newer LCDs will be very dim and you may only be able to read them at an angle if at all. If you really want to squeeze a little more battery life out of your AVR Butterfly based project then you could have a go at tweaking this value back to 850μs or even 575μs and check the display for readability.

Heres an explanation of what those bits do from the ATMEGA169PA data sheet on page 248.

• Bits 7:5 – LCDDC2:0: LDC Display Configuration
The LCDDC2:0 bits determine the amount of time the LCD drivers are turned on for each voltage transition on segment and common pins. A short drive time will lead to lower power consumption, but displays with high internal resistance may need longer drive time to achieve satisfactory contrast. Note that the drive time will never be longer than one half prescaled LCD clock period, even if the selected drive time is longer. When using static bias or blanking, drive time will always be one half prescaled LCD clock period.

And the accompanying table from the same datasheet also on page 248.

Table 23-7. LCD Display Configuration

LCDDC2LCDDC1LCDDC0Nominal drive time
000300 µs
00170 µs
010150 µs
011450 µs
100575 µs
101850 µs
1101150 µs
11150% of clkLCD_PS