It was suggested some time ago that I should run tests on lipo packs from several different makers and compare the results. The problem was the cost of the packs, but this has been overcome by Rob volunteering to pay for them, so here are the results.
I should make it clear that all packs have been purchased at random (no specially selected ones!), and that Rob told me to post all the results just as they came and without further reference to him. I have no connection with GC and receive no payment.
The 3S 2200 pack format was chosen as the most popular type and eight different packs have been tested in two groups of four. One group is for 20C rated packs and the other for 30C rated packs.
I am aware that most modellers just want to see the final results and not the detail (Just as we all jump to the “Flying” bit at the end of a model review), so I have split the report into two tables with notes on what to look for in the results and a more detailed section on equipment and measurement methods. The latter part is for the more technically inclined (anoraks!) who are interested or might need convincing that the test methods and results are valid.
To avoid attacks from various quarters I am not putting my opinions in the report, just specifying how the tests were carried out and publishing results that were obtained. The reader should, from the notes below, decide on what he considers the most important criteria on choosing a pack and compare the results obtained in the tables from the various packs.
[img]
[img]
[img]
NOTES ON RESULTS.
Trying to choose a pack just by asking other modellers or seeing a model fly is impossible because there are too many variables such as model weight, motor, prop, ESC programming, temperature etc. Add our individual prejudices and brand loyalty and, as Basil said in Fawlty Towers, “You might as well ask the cat”. In my tests the packs are all discharged at maximum rated current down to 3.00V/cell under identical conditions.
This full rated constant current discharge that the packs are subjected to is more severe than they are likely to see in practice, but it is a consistent, flat-field measurement of the pack’s ability to deliver their claimed performance. They are only subjected to the max rated load conditions specified by the manufacturer, starting from a fixed ambient temperature of 25 deg.centigrade.
The pack voltage and temperature is measured at 15secs, at the 50% nominal discharge point and at the end of discharge. The pack ESR (Effective Series Resistance) is measured at 25deg. and the total Ah and Watt.hrs delivered by the pack are both recorded.
(1) PACK VOLTAGE – The higher this is at any point, the better, as it means the motor will deliver more power. If the ESR is high then the initial voltage on load (say at 15Secs) will be lower, and this will continue throughout the discharge, but at a reducing level as the internal heat dissipated in the pack will reduce the ESR and the voltage drop due to it.
(2) TEMPERATURE – High temperature is the killer of a pack, but it is very difficult to quantify Life v Temperature in an equation. It is not linear and a small increase in temperature is likely to shorten life considerably, particularly near the higher limits.
Electrolytic capacitors, which are electrochemical devices not unlike rechargeable cells, show a halving of their projected life for every 10deg.cent. rise in temperature.
If you are looking for a long life pack, then look for low temperatures.
The surface temperature rise over the discharge is quoted at the end of the results. I have made a nominal 55deg. rise the maximum I consider acceptable, equating to a max. surface temperature of 80deg. starting from an initial figure of 25deg. Even this figure is only a guess and may be much too generous, as the internal cell temperature will be higher, but cannot be easily measured. There are other factors affecting pack life, such as QC and electrochemistry of the pack but I have no way to quantify them.
(3) ESR - The voltage drop of a pack and the heat generated in the pack due to load current is a direct function of ESR and we should be looking for the lowest value. It is very temperature dependant and the very different figures shown at 25deg. tend to converge at higher temperatures.
(4) Ah – The Amp.hrs delivered by a pack is compared with its rated capacity and quoted as a percentage. The manufacturers figure will be at 1C so at full rated current anything over 90% is acceptable.
(5) Wh – Watt.hours is a true measure of stored energy delivered by the pack. A lower voltage under load will reduce this figure. Note that the packs with lower voltages deliver lower W.h figures.
(6) Plots show best and worst performing packs with others missed off for clarity. A voltage difference of 0.5V will give a power difference of about 10% on a typical setup.
EQUIPMENT AND TEST METHODS
(1) Constant Current Discharger – Is a modified version of the unit I built for Bob Smith, for use in his Q&EFI testing. It is capable of sinking up to 100Amps and 1600 Watts over a voltage range of 3 – 45V (ie 1 – 10 cells). The discharge current can be set in 1A steps up to the max rating and the test run is automatically terminated when the min. voltage limit of 3.0V/cell is breeched.
(2) Medusa Power-Plus Wattmeter – This is a standard unit which measures the Ah and Wh of the discharge run. I previously used a data logger and fed the results into an Excel programme to produce the figures with the Medusa as backup, but the figures agreed so consistently (within <1%) that I now just use the Medusa as it saves time and effort. I also plot out the Voltage and Power curves on a laptop from the Medusa’s RS232 output, but have not included them as I have noticed that few people have viewed them in previous postings.
(3) Tenma 72-7760 DVM – A standard 4¾. digit multimeter with an accuracy of 0.05% and which has been calibrated is used for all voltage measurements. These are all taken at the cell itself, via the balance connector and therefore all power lead and connector voltage drops are eliminated. The balance connector voltage is also used to terminate the discharge to minimise capacity errors.
(4) IR Thermometer - This is a handheld IR thermometer I am now using to obtain accurate pack surface temperatures. It has an accuracy of + and – 1deg.cent over the range being measured. I have previously used a DVM thermometer with a miniature thermocouple probe but found it difficult to apply it equally well to different packs and be sure of proper thermal contact even when using thermal grease.
(5) ESR Measurement – I use a modified Battery Performance Meter, which I built a few of some years back, in modified form to read ESR directly in milliohms. In essence it measures the pack voltage with a ‘Sample-and-hold’ circuit, applies a constant current load of 10C (20A) for 12 mS, but takes another sample-and-hold voltage reading at 10mS, just before the load is removed. The difference between the two readings is suitably divided so that the ESR can be read directly with a DVM. The 12mS pulse is long enough for the battery voltage to settle after the load application, but not so long that any heating effect can reduce the ESR of the pack.
The voltage measurements are again taken from the balance connector via a differential amplifier to eliminate errors from power lead drops.
GENERAL
I carry out the testing with an open mind and operate on a flat playing field, but I only test one sample of each pack and therefore cannot account for production spread. There must inevitably be some tolerance in all the parameters of each pack but the poorer performing will always say that their sample must have come from a bad batch; it is a standard excuse I have heard several times.
The above testing only demonstrates the ability of the packs to deliver their stored energy. It would be nice to compare pack life which many users are interested in, but that would involve running 100 or more cycle tests on each pack. I have access to the necessary equipment but not the inclination to carry out such testing due to the time and effort involved in such an exercise. As a short cut to the likely life of a pack I would suggest that temperature rise is the best indicator to look at.
A quick note regarding ambient operating temperature might also help in regard to pack life. Manufacturers seem to ignore this aspect which they obviously know plays a large part in Lipo performance and life; perhaps it is convenient to ignore it. All my testing is at a pack temperature of 25deg.cent and I soak any pack for at least 24 hours at 25deg. in a temperature controlled chamber prior to measuring ESR or running a discharge. The ESR of a lipo pack has a negative temperature coefficient, and this varies from one make to another. Packs with a high coefficient will perform badly at low temperatures and (I believe) be prone to damage by high discharge rates in winter conditions. (See previous post at viewtopic.php?f=121&t=1280)
I should make it clear that the above only tests the performance of the packs, there may be quality aspects of which I have no knowledge which may affect the life of the packs apart from the damaging aspect of excess temperature rise. If there is not then the old adage “You get what you pay for” definitely does not apply to lipo packs.
Having said all the above and despite their safety problems, lipos really are a fantastically efficient energy storage device, having tremendous power delivery/weight ratios and have transformed electric flight over the past year or so. Electric helicopters and 3D flying were unthinkable a few years back, whereas now they are commonplace.