The choice between the Samsung 35e and the Samsung 30q was not one made without research and testing. Aspects such as capacity, rated current draw, size, weight, availability etc, are all major factors. Many cell options were looked in to and considered before any batteries were built for testing. Two specifically, the 35e and the 30q, were made in to 10s4p batteries to be tested in GT boards. On paper, both cells would be suitable, however knowing that cells can perform differently in real world situations, substantial testing was going to be required.
As the cells are both 18650, the same battery PCB and structure could be used for each battery.
Real world testing not only involved range tests, but stress testing the batteries in a workshop using load testers and real world conditions such as steep hills and heavier riders.
A local hill is used for stress testing board components and batteries. It is a bit over 200m with a peak gradient of above 25%, while maintaining around 20% for majority of the hill. This was the first major test for the batteries in real world situations. Prior to testing, we were expecting the 30q to be the better performer based on manufacturer specifications.
The surprise then came from the 35e pack. We consistently had better performance on the hill from the 35e pack. This was admittedly a surprise to us as we did expect the 30q to perform better in this situation.
The hill tests were carried out multiple times, each time consistently showing the 35e the better performing, even as the voltages began to drop.
When it came to range testing, the batteries were originally tested using the same BMS settings from the GT series. This meant the cut off voltage was higher than the rated cut-off for these cells. This plays a big part when comparing the total performance.
The initial batteries were range tested multiple times, by multiple riders. The results were comparable with each other, with the biggest variations coming from the riders and the nature of the ride. A factor of this was believed to be the higher cut off voltage of the GT boards.
Once the voltage cut-off was adjusted to suit the cells, the 35e started returning better range consistently due to better performance at lower voltage.
When it comes to the “nitty gritty” of looking at these cells, a few things need to be considered and kept in mind while looking over lab test results and graphs.
When viewing cell performance graphs, it is important to note the current draw listed is a continuous draw through the entire test. When it comes to the skateboards operating in the real world, this is never the case. They will always spike higher during initial load, then the current draw will reduce, generally to a relatively low level while maintaining as cruising speed.
Why is this important? Essentially this means multiple current draw graphs need to be considered and almost merged to understand how the cell may perform throughout the entire capacity usage.
The 30q and 35e cells perform very differently under different loads and at different voltages. The 35e cells do drop voltage quicker initially, however retain a more consistent rate of reduction throughout the remainder of the capacity in comparison to the 30q.
An interesting point to note with the 2 different cells is how they perform in the lower end of their voltage range.
From the graph below, while the 35e cell drops to 3.3V sooner than the 30q, it retains more usable energy and capacity until it reaches its cut off. From 3.2V, the 30q has less than half the remaining energy of the 35e at the same voltage.
Translated to real world situations, this means that the board will operate in a more consistent manner throughout the mid and lower range of the battery voltage, allowing for better performance when the battery voltage is low.
The graph used for this example is based on 10a current draw test for to an individual cell.
In a 4p pack this can be translated to an equivalent 40a draw continuous on a board. To put this in perspective, to achieve that continual draw requires a street board with 97mm wheels to be ridden at full speed in GTR mode on an incline.
A GTR board with 97mm wheels, drops easily below the rated continuous draw of the cells while maintaining top speed in GTR mode on flat ground.
As seen in the second graph, the 35e show less voltage drop with the current at 8a, or board equivalent of 32a. When compared at the same voltage cut-off of 2.65V, the 35e still provides greater capacity than the other cells tested.
There is no doubt from the graphs that the 30q maintains higher voltage initially during the tests. However, with the how the boards operate and how the boards are typically used, the 35e has proven to be a better choice. While the theoretical max speed under load may be higher initially with 30q cells, the performance available right to the cut off voltage remains much more consistent and usable from the 35e.
The GTR boards have a cut off voltage of roughly 2.7V/cell. Or 27V for the complete battery. If the cut off voltage was higher, as the GT models were, the range would be comparable.
The 35e batteries have been undergoing testing in boards since 2017, with the last 9months focusing on prototype boards being heavily tested across 3 different countries. In total over 20 samples were initially made and tested. The testing was rotated between many different riders with the intention being to stress the board operating systems as much as possible. The batteries have proven throughout this testing to continue to perform exceptionally well.
Looking at paper specifications, the 30q cells in some respects, could be argued as being better. The 35e could equally be argued as being the better choice, dependant on the application. For our boards, through R&D, real world and long term testing, the 35e units have shown their worth, reliability and continued performance.