Internal Resistance or Power%

Years ago we (Pro-Trak) decided that it was very difficult to measure the internal resistance of cells in a pack accurately.  

However we realised  the effect of the  internal resistance of a cell changed the peak voltage that cell reached when charging.  Many might disagree with this assumption but we think the  information at the end will at least make you think about the connection.

To the average RC car enthusiast the IR figures mean very little, some will possibly know that the lower the value the better the cell will release it's power.  But what is a good IR figure?  17, 23, 11.  There is no standard that the average RC enthusiast can relate the figures to.

We needed a standard measurement  the RC enthusiast could relate to. Hence Power% has been a measurement provided by Pro-Trak since the beginning.

 From tests we decided that the average Ni-CD cell peaked at 1.66 volts, later when Ni-Mh cells were available we did the same tests and decided the average Ni-Mh cell peaked at 1.52 volts.  Using these values as the standard we could say a Ni-Mh cell that peaked at 1.52 volts has a Power % value of 100%.  Having established a standard we could now apply calculations to the peak volts to produce Power % readings. Ni-Mh cells that peaked below 1.52 would have Power reading above 100% cells that peaked above 1.52 volts would have Power readings below 100%

ie 

Ni-Mh cell that peaked at 1.65 volts would have a calculated power % of 92%
Ni-Mh cell that peaked at 1.47 volts would have a calculated power % of 103%

This calculation is also be applied to the pack to give a power reading. You would  never get an IR reading for the pack.

The power reading of the pack allows you to check it's performance every time you charge it.  If you notice the power reading is getting lower than the figures when the pack was new then it could be getting past it's best, if the remote leads are fitted to the individual cells you will be able to tell if it is a general problem with all the cells or 1 particular cell. 

We believe the Power % measurement is much easier for the average RC enthusiast to understand, if a pack has a power % of 103% then it is a reasonable pack, if it has a power % reading of 92% then it is not so good.

Anyway now to prove the peak voltage of a cell is a direct result of IR.

The following graphs were produced from raw data collected using a Turbo Matcher 4/35.

The raw data is available at the following web site.

This graph shows the IR and the peak volts graphed straight from the raw data.
You can see IR and peak volts go hand in hand!

This is the same data but sorted on IR.  As you can see the trend is for the peak volts to follow the IR.  However the peak volts (blue line) is a little wavy.   But I suspect this is because the IR is to whole numbers and the peak volts is to 3 places.

This graph is sorted by peak volts and clearly shows the peak voltage follows the IR. As the IR increases the peak voltage of the cell increases.

On this graph we have added a calculated IR from the peak voltage. 

From this data we would suggest that a calculated IR from peak volts is a more accurate indication of the IR of a cell.

On this graph we have added the Power% figure that Pro-Trak would produce from the peak volts.  If you pick a cell with a low IR you can see this is reflected in a high Power% figure, if you pick a cell with a high IR it will have a low Power% figure.
( the scale on the left is power % and on the right is IR)

We hope from this data you will realise that the Power % figure given by Pro-Trak is a very good indication of the IR of the cell and the pack.

The raw data used also had the following information on the quality of the cells.

0001 - 0012: 1st lot GP3300 
0013 - 0072: 2nd lot GP3300
0073 - 0084: Obscurity
0085 - 0100: Matcher 1
0101 - 0112: Matcher 2
0113 - 0124: Matcher 1
0125 - 0154: Matcher 2
  ( it was uncertain whether these obscurity cells were 1st issue GP3300s or second issue, but having plotted  the data we can safely say they were 1st issue cells)

This graph shows the difference using matched cells can make. The 1st GP3300s and the Obscurity cells were not so good as they had a high internal resistance.  The 2nd GP3300s were much better.  The 2 different cell matchers produced cells with about the same IR.

Here the data has been sorted on the descending IR of the cells.  You can see the 1st GP3300s and the obscurity cells are at the start of the graph where you would expect to see them. But there are also a couple of the matched cells down in the region of the 2nd GP3300s.

Here the data has been sorted on the descending peak volts.

Sorting Data

From the 5 different qualities of cells available in the raw data we would expect the 1st GP3300s and the Obscurity cells to be at the beginning of the graph and the Matched cells to be at the end of the graph.  The 2nd GP3300s we would expect to spread across the whole graph as they are not tested and there could be good cells and bad cells in this group.

On the following graph you will see exactly what we would expect.

Here the data is sorted on the Peak volts and you can still see the trend we would expect but some of the 2nd GP3300s are well into the matched class of cells.

Here the data is sorted on the Discharge volts, again the trend is what we would expect.

Here the cells are sorted on the Discharge Time and as you will see the results are not what we would have expected.  What we would consider the poorer cells have good discharge time and the expected trend is reversed. 

Here the cells are sorted on the Power% figure given by Pro-Trak, and the trend is back to what we would expect to see.

Here is what we would consider to be the best way of sorting cells.  The cells are first sorted by the Power% figure then on the discharge volts.

Here we see exactly what we would expect, some of the untested 2nd GP3300s in with the better cells.