Saturday, November 29, 2008

Eagle Tree Systems eLogger V3 Review

The eLogger V3 from Eagle Tree Systems is an in-flight data logger which continuously records volts, amps, watts, and milliamp-hours to enable pilots to analyze the performance of their electric power systems. In addition, it supports a wide variety of optional sensors to record altitude, airspeed, component temperatures, RPM, and more.


Specifications



Dimensions
2.25%26quot; x 1%26quot; x 0.5%26quot;


Weight
0.7 ounces (20 grams)


Voltage
5 to 70V


Amperage
Up to 100 amps


Temperature
0 to 424F


RPM
100 to 50,000


Street Price
$70 for the eLogger, $40 for the PowerPanel, $15 for the brushless RPM sensor, $10 each for either temperature sensor, $30 for the altimeter




The eLogger is a small device which continuously records these values to its internal memory. Once back on the ground, these log files can be downloaded to a computer for graphing and analysis. The device can also be used to display live values for all of its sensors, which means it doubles as a wattmeter on the bench.


Eagle Tree Systems provided me with the eLogger, a PowerPanel LCD display, two different temperature sensors, and an RPM sensor for brushless motors. All of these can be connected simultaneously, and still have room for a third temperature sensor and either a second RPM sensor or a throttle sensor. In addition, there is an altimeter, an airspeed sensor, a servo current meter, and a GPS module available. The variety of sensors and the ability to use so many of them at the same time is one of the major strengths of this product. The eLogger competes with the Oracle Data Recorder from Medusa Research and the eFlightWatt Data Logger from Mile High Wings .








To start with, I decided to test the eLogger as a wattmeter, by connecting the PowerPanel LCD and hooking it up between the ESC and battery in my Kadet EP-42. The logger I requested has integrated Dean's Ultra connectors, but another version is available with leads if you prefer a different connector. I found that installing the eLogger first made it difficult to connect the battery - I'm used to the ESC wire's flexibility helping to align the Dean's connectors. An easy solution is to put Velcro on the back of the logger, keep it unattached until the battery is connected, and then secure it in the fuselage.


The factory settings showed volts, amps, milliamp hours, and temperature on the LCD, which was good enough for a bench test. Everything worked fine for a few minutes, but then I noticed the PowerPanel display was corrupted. I reset the power and got a message to configure the LCD. It turns out that I had hit a bug in the software which had already been fixed. A newer version of the firmware was sitting on the CD-ROM that came with the logger, but of course I had skipped it (and the manual) in my excitement to try it out.





A short firmware update later I was back in business. Eagle Tree tell me that all new eLoggers going out the door already have the fix. Bugs are a fact of life, and I'm happy to see that they are responsive to fixing problems, and that the logger can be updated at home with the included USB cable. My advice is to hook a new eLogger up to a PC first thing in order to configure it and check for the latest software.





Besides acting as a simple wattmeter, all of the attached sensors can display their live values on the PowerPanel, which is very configurable. From the PC software, you can choose which parameters are displayed and where, give them customized three letter names, and set up multiple pages of information. For example, eight parameters can be spread over two pages, which alternate every two seconds.


Even better, there's an option to show the maximum values attained since power on when the current is zero. This is useful both on the bench to check peak amps, as well as at the field after a flight. The latter is particularly useful if you want to try out different props but not bring a laptop out to your flying site. On larger models the LCD could be mounted in the fuselage or under the canopy to make it visible. For smaller planes or helis, the PowerPanel is best kept on the ground, and connected when needed.



The desktop software can be used in live mode, where all parameters are displayed in real time on large readouts, and optionally recorded. There's even a live mode for the graphing feature, which is extremely cool (although perhaps less useful than the gauges and digital displays).



Next I moved on to recording in-flight data. The Data Recorder application is used to configure the logger's behavior, which is retained in the absence of power. You have a choice of which parameters to record, and how often to log them (from 1 to 10 samples per second). These choices affect the total logging time available, as does the data compression used internally. Eagle Tree state a minimum of 45 minutes of recording time. I set up the logger to record volts, amps, RPM, and one temperature value, sampling 4 times per second. In this configuration I'd estimate 2.5 hours of data can be stored.





There are three different RPM sensors available for this system. The first is a magnetic unit which requires the installation of a sensor and one or two magnets on a spinning surface. The second is an optical sensor which reads alternating light and dark areas on the back of the spinner or other moving part. Both of these are somewhat permanent setups, and are better suited to larger planes.





The third RPM option, and the unit I requested, is an electrical sensor designed specifically for brushless motors. It has the easiest installation procedure of the three, which consists of stripping a quarter inch of wire and attaching it to any of the three wires between the motor and ESC. I found it was easy to stick the sensor wire into the female 3.5mm bullet connector on the ESC, then reconnect the male side. Back on the PC, a configuration screen asks for the number of poles in the motor and the gearbox ratio, if one is used. The second setting is also used to indicate the ratio of the main gear to the pinion in a helicopter. One nice feature is the ability to see motor speed and prop/head speed at the same time.


I tried two temperature sensors as well. One is a loop, which was easy to slip around a LiPo pack, but I found that it did not cinch tight enough, so I added a piece of tape. The other sensor is designed to be taped down, and can be used interchangeably. Both are appropriate for measuring the temperature of inrunner motors, ESCs, and batteries. I couldn't think of a good solution for an outrunner motor though, since the spinning can is bound to be hotter than the non-rotating base.


Before flying with the eLogger, I wanted to compare the static amp draw of my Reactor with two different props. I ran full throttle with the stock 10x4.5SF, then changed over to an 11x4.7SF and ran the motor again. I downloaded the log files to my laptop, which showed two sessions. Each time the power is connected to the logger a new session is created, which is convenient. As seen here, you can examine these sessions individually or see them all at once. Using the default graph, I was able to verify that the current increased from 17.5A to 19.2A - just under the 20A sustained the RimFire motor is rated for. This gave me the confidence to use the bigger prop.





With the logger wrapped in foam at the rear of the compartment, I slid my battery forward a bit more than usual to get the center of gravity right. I flew a variety of maneuvers and didn't notice any effect on performance from the extra weight. Between the bench test and the flight, I had about 16 minutes of data in memory which was reported as 10% capacity used. It took about 5 seconds to download over USB, and saved to disk as a 448K file. This FDR file format is just a big text file, which means it can be examined or modified with any text editor. Most users will never do this, but it's nice to know the data is not locked away in a proprietary format.



The graphing feature is really the heart of the Data Recorder app, and it revealed a couple things about the flight right away. The first is that the larger prop only peaked at 16.75 amps in the air, and seemed to average around 10A. The average reported in the legend takes into account the idle time before and after the flight, so it's a little low. But second is the temperature curve of the LiPo pack. Not only did it climb steadily in the air, it continued to get hotter for another minute back on the ground. This confirmed my suspicion that I needed a larger air intake in the cowl for cooling.


Back on the bench I used a Dremel tool to open up the hole in the fiberglass cowl. On my next flight the weather was about 10 degrees warmer, but the difference in the curve was significant. Not only did the peak battery temperature drop from 66 to 58 degrees, the increased ventilation let the battery stabilize and even cool a bit as I drew fewer amps around the six minute mark. I would not have known about this problem or have been sure of the solution without the eLogger.






Since the logger itself has no buttons or controls, the PC application is used both for setup and flight analysis. The main screen can be configured to show whatever parameters you like, and many like RPM can be shown as both a gauge and a digital readout. These configurations can be saved for different models. One thing I found confusing was that these models do not also save and restore the configuration of the logger itself, for example which parameters to record in the air, or how many poles the brushless motor has. It would be nice to save the complete configuration of the eLogger in these model profiles, and have the LCD briefly display the name of the model it is programmed for on power on.


If the default graph options are not to your liking, there is an incredibly powerful settings window which allows you to change almost every aspect, from colors to line thickness to axis labels. I really appreciated that one axis can be made logarithmic, which is useful for showing large values like watts and RPM at the same time. Although the controls aren't obvious, the graph can also be zoomed in and panned around with the mouse. Because you can spend a lot of time configuring a graph to look just right, I'd like a way to save all of these settings as a preset. That would be useful to jump between one graph of just amps and temperature, and another of all settings simultaneously.





Later in my testing I tried the altimeter sensor. This has an unusual calibration routine, which consists of putting it in the fridge for five minutes, then connecting a battery as it comes up to room temperature. The altimeter connects to the LCD port on the logger, but allows the PowerPanel to be daisy chained, so both add-ons can be used simultaneously. The airspeed and servo current sensors (not tested) can also be chained together in this fashion, which is a great feature.





I installed the altimeter along with both temperature sensors in a Great Planes Spectra two meter electric glider. By disconnecting the bullet connectors between the brushed motor and the ESC, it was easy to slip the loop sensor around the motor can. I used scotch tape to attach the other temperature sensor to the battery, and tucked the altimeter into a small free compartment. It was a tight fit to get the canopy back on, but it worked. I might have had more room by placing the eLogger in the battery compartment, but would not have been able to reach the motor. Eagle Tree seem to have thought of this, and offer an inexpensive 12 inch extension cable for the temperature and RPM sensors.





After the flight, I downloaded the log to find a few surprises. First, that 600-size brushed motor hit 40 amps when I tested it at full throttle on the ground. Note to self: throw that piece of junk out and go brushless. The second was a reported maximum altitude of almost 3100 feet! Looking at the graph this was obviously a bad data spike. I tried to use the Set Minimum and Maximum Parameter Values command to eliminate this spike, but it had no effect on the graph. It appears this feature only affects the logger as it is recording.


So, how to fix this spike? As I mentioned earlier, the FDR log files are just plain text, so I made a copy and found the bad data by hand. Just changing these 1.5 seconds of data resulted in a much better graph, and correctly showed the max altitude as 240 feet. The climbs are interesting to watch: the amps (in purple) come up first, followed by the altitude (in turquoise), then gradually the battery temperature (in brown) and the motor temperature (in yellow). One thing this graph shows is that the motor heats faster than the battery, but also cools quicker as it has better cooling.






Later I browsed the Eagle Tree site and found a list of all changes to the software , something I strongly encourage companies to provide. And wouldn't you know it, a newer version was already out with a fix for altitude spikes. The new software also improves the appearance of the graphs with a white background and logo. The eLogger software is being improved constantly, which is a tribute to Eagle Tree. They also host a forum and respond to questions on RCGroups , which is a great way to provide support and share answers with many people simultaneously.


Can you fly electric without a logger? Sure - and you can practice biology without a microscope, but you won't really know what's going on. As a sport flyer I have found the eLogger very helpful for setting up new models and making changes to existing ones. It's particularly effective at detecting excessive current and cooling problems. This is a very well thought out, expandable logger which doubles as a wattmeter, and is very reasonably priced. Highly recommended.


To learn more about the eLogger, visit the product page on the Eagle Tree Systems website .

No comments: