Li-Ion BMS

Home resources white_papers wp_resistance_vs_impedance

White Paper - Resistance vs impedance

Cell impedance at 1 kHz is not the same as DC resistance

Introduction

When cell manufacturers list resistance in their specs, they usually are talking about AC impedance at 1 kHz.

Yet, what the user needs to know is the DC series resistance, because DC is what flows through the cells.

Even though both are measured in mΩ, 1 kHz impedance and DC resistance are really different parameters.

Manufacturers specify impedance at 1 kHz because:

  • Test equipment to measure it is readily available
  • It remains pretty constant over the life of the cell
  • An unusually high value is a good indicator of a cell with manufacturing errors
  • Its value may be a nice, low figure
  • And, frankly, most chemists do not understand the concept of DC resistance

Tha value of the cell's AC impedance is useless to the cell user. Reporting this value is entirely misleading to the user, who assumes it;s the value of the cell's DC resistance.

The user need DC resistance because:

  • It allows to compare cells to be used in a power application
  • It allows to estimate voltage sag under load
  • It allows to calculate the battery efficiency under load
  • It allows to calculate how much heat must be removed from the battery through cooling

You can read more about this topic in section 1.2.7 of the book Battery Management Systems for Large Lithium-Ion Battery Packs .

Data analysis

Can you infer DC resistance from AC impedance?

No, not really.

However, if you look at data from many different cells, you can see some correlation.

Scatter graph
Impedance at 1 kHz vs DC resistance, for a variety of Li-ion cells

It appears that, for most cells, the actual DC resistance value (at 50 % SOC, 25 °C) is within a range of 2:1 to 1:5 of the cell impedance impedance value reported by the manufacturer.

That is a nice correlation.

But, when you stop and think about it, it means that a cell that is specified to have a 1 mΩ AC impedance at 1 kHz, is likely (though not guaranteed) to have a DC resistance of anywhere between 0.5 and 5 mΩ.

That is a full order of magnitude, and not very helpful to determine if a cell is appropriate for an application.

Davide Andrea, Elithion, 1/23/13

© 2008~2017 Davide Andrea. All rights reserved, except where noted by CC mark. Page published on Jan 23 2013. Graphic design by morninglori