3

I am making a battery tester for lithium ion batteries. I want to measure the internal resistance, voltage and current.

enter image description here enter image description here

I am skeptical of my results. On batteryuniversity, I read that the internal resistance curve should start and end with higher resistance. The resistance should reach the lowest point in the middle of the curve.

I am using LFP18650, 9 V, and 1500 mAh battery packs.

As a quick review of the measurement of the Ri: Ri= U1-U2/I1

  • Load enabled for 19 seconds
  • Load disabled for 3 seconds

U1: measure voltage very close to the last moment of the 3 seconds*

  • load enabled
  • waiting 500 milliseconds

U2: measure the voltage very close to the last moment of the 500 milliseconds.

I2: current measurement at the exact moment of the U2.

Why am I getting this result? Totally not similar to a u-curve.

SamGibson
  • 17,231
  • 5
  • 37
  • 58
Khalil
  • 31
  • 3
  • How much sag during load, slew rate and time of ADC is important – Tony Stewart EE75 Jun 02 '22 at 13:02
  • 1
    @winny 500 milliseconds (time interval), between "with load" and "without". – Khalil Jun 02 '22 at 13:08
  • 1
    Thank you! Is 500 ms off + 500 ms on the total cycle time? If yes, that time constant may be too short to give a correct V1. Have you tried to do a step response of the battery to find out its RC time constant? – winny Jun 02 '22 at 13:10
  • @TonyStewartEE75 thank you, I am using just python code for a programmable load east Tester ET5410 and charger Velleman connected to my PC. Good to know (I am not an electrical engineer) I am a software engineer ;) – Khalil Jun 02 '22 at 13:15
  • @winny 19 s load enabled, then load disabled for 3 seconds measure V1 at the last moment of the 3 seconds. Load enabled again, waiting for 500 ms and measuring the V2 with the I2. and again the same loop – Khalil Jun 02 '22 at 13:24
  • 2
    Your method of measuring Ri is OK, and gives 'correct' results until... what's going on at the right-hand side of the second graph, current is decreasing? Why is it not constant? – Bruce Abbott Jun 02 '22 at 13:46
  • @BruceAbbott like the way while charging a lithium ion battery, using CC and CV. Do you mean that I should measure the Ri just in constant current mode? – Khalil Jun 02 '22 at 13:51
  • 3
    _"Do you mean that I should measure the Ri just in constant current mode?"_ - Yes, all the way to low voltage cutoff. – Bruce Abbott Jun 02 '22 at 14:01
  • 1
    Right, three seconds >> RC time constant of the battery. It’s not that causing the diminishing calculated resistance. – winny Jun 02 '22 at 14:25
  • 1
    Khalil measure V at the beginning , middle and end of each 500 ms period to discover errors, then use a constant current load, not a constant voltage that later decays with the low state of charge and thus neither CV or CC or CR .. warning bad results – Tony Stewart EE75 Jun 02 '22 at 14:29
  • There are actually two ESR's and two C values for RC1=T and RC2 = T2 from double layer effects. So duration of pulse and delay time should be short high current and long no current in seconds as dV/dt= Ic/C and C2 is ~ 10000 Farads for nom Li Ion 18650 but what is C1R1? – Tony Stewart EE75 Jun 02 '22 at 14:37
  • If you short a battery for <<1 second or less. Why does voltage return slowly from 0V? The answer is C2R2 – Tony Stewart EE75 Jun 02 '22 at 14:38
  • I would increase current to max like 3A ? then extend Toff to 3 seconds and measure start middle end or until Vout returns to near previous voltage. This is how you can measure and compute T1 and T2 for ESR1C1 and ESR2C2 both series RC's in parallel. one is charged from the other in CC mode then in CV mode the smaller C catches up with the longer T time. but only for < 10% of total charge ctime. – Tony Stewart EE75 Jun 02 '22 at 14:41
  • You'll need some help with the formulae to convert Ah to coulomb counting to C1, C2 with separate slopes for T1, T2 (p.s. there is also a T3) – Tony Stewart EE75 Jun 02 '22 at 14:45
  • Do you understand me? How can I explain how to measure these properly? A constant current pulse measured by voltage changes computes ESR directly V=ESR*CC pulse. The time intervals must measure over a longer period, at start middle and end to compute dV/dt= DeltaV/ESR/C = Ic/C. Capiche? There are two ESR * C =T time constants at least. Questions? Think like Pease. – Tony Stewart EE75 Jun 03 '22 at 17:03

0 Answers0