Internal resistance mystery

sail_power · Feb 05, 2003, 10:01 AM

Hi I like to know if someone have an explanation for the behavior
of my battery pack.

I did find out that during the first moments of the discharge, the
internal resistance was higher than after few minutes.

In order to verify this, I make a set-up to test the internal
resistance during a controled discharge. The basic discharge rate is close to 1C, and the voltage is measured after 1 minute, then at every 10 minutes, for temporary discharge rate from close to 1C up to close to 6C. As soon as the measurement are completed, the 1C discharge rate is restored. These data are put on a graph, and the slope of the 'curve' is the dynamic internal resistance.

The internal resistance per cell is 198mOhms after 1 minute, then
145mOhms almost constant until 80% dischared.

What will explain that the initial resistance is 35% higher?

The pack temperature does not increase much during the test, maybe from 20C to 30-35C, does this can explain the phenomen?

This is a 'cheap' no brand 7 X AAA NiMH pack, assumed capacity 550-600mA. It was used for few 'unsuccessfull' flight at about 6C, before assign to 3-4C task. But 145mOhms still look very high DC dynamic resistance.

Does someone have run such test with high quality AAA like the Sanyo 720mAh? I will be surprise if they really meet the 30mOhms listed in the spec, since this data is AC resistance.

Another surprise is that the dynamic resistance remain pretty constant on the 1C to 6C range. Maybe this is because each high current discharge last for less than 10sec, before the 'low rate' 1C is restore.

Any explanation is welcome.

Regards
André

gkamysz · Feb 06, 2003, 12:02 AM

Your data appears to describe a typical inexpensive NiMH cell.

It's hard to understand what you are asking but if you look at a typical discharge curve the voltage is lower when the battery is near discharged. This is what you have shown on your graph.

0.145Ohm for a 7 cell pack would be very good for a AAA cell. Sanyo 720 are listed with .040 Ohm resistance. I have measured it to be approximately the same when discharged warm at 3.5A(4C). Don't expect any other manufacturers AAA to have lower resistance than this.

I have never heard of dynamic resistance. I don't know what characteristic of the cell it describes.

Greg

sail_power · Feb 06, 2003, 08:47 AM

Sorry Greg, sometime I have hard time to make my questions clear, English is a second language to me.

1- What I don't understand the 'improvement' in internal resistance between the 1 minutes and 10 minutes measurement. The others curves are in a easy to understand pattern.

2- The 145mOhms resistance is per cell. The total for the 7 cells pack is simply not acceptable.

3- I did call this dynamic resistance, since it is taken from two differents load value, insteed of from no load / load. I think that it is more representative of the cell under uses.

You said that you did measure the Sanyo AAA 720, and find a value close to the 40mOhms. For me this is an 3:1 improvement.

Thanks for your advise
André

Brad S · Feb 06, 2003, 01:49 PM

Andre,
My guess is that the reason for the higher apparent resistance, shown as the slope of your plots, at the 1 minute time versus the 10 minute time is due to a measurement timing issue. If you are taking your voltages at the various rates in order from lowest to highest, there will be an ever-decreasing baseline cell voltage as the measurements are taken. This decrease will bias your calculated effective internal resistance numbers to the high side. At the beginning of the discharge, the rate of baseline voltage decrease will be greater than it is during the middle part of the discharge.

Your test approach is valid and can be a much better predictor of sustained-load cell voltage than AC impedance is. However, the effective internal resistance numbers generated by this type of measurement method are significantly dependent on the test parameters, especially voltage measurement points, load level durations, and temperature effects. Personally, even with these added complexities, I find the results from these kinds of dynamic load tests to be much more useful for characterizing cell performance in my type of eflight applications (variable throttle settings) than any constant current discharges or AC impedance numbers. For pylon racing, or other constant throttle setting applications, the constant current discharge plots are by far more useful than dynamic load testing.

Sorry, I've not done this type of testing on AAA cells so I don't know how your numbers would compare to other cell models.

Brad

Chris True · Feb 06, 2003, 02:30 PM

Nimh cells have lower resistance at higher temperatures. If you start with a cold pack you can see the resistance drop as the cells heat up initially.

I've seen the following behavior - start a motor discharge, initial current is 52 amps which fades over the course of 30 seconds or so to the point that the current draw is only about 46 amps. Then it spikes back up to 50+ amps as the cells heat up - holds this for a long while then tapers down again.

Andy W · Feb 06, 2003, 03:01 PM

What Chris said. When you were describing the symptom, I said to myself "this sounds like a NiMH pack" and sure enough you state that at the end. In addition, the external temperature of a pack in no way reflects what could be happening inside. It takes time for the heat to propogate out. My F5B and hard-pushed NiMH packs continue to heat up after I've landed, in some cases. I dump them in a cooling tube right away..
..a

sail_power · Feb 06, 2003, 03:57 PM

Andy, that's make sense, the internal temp probably increase after the first test at high load, then stabilize.

I did not tought that the NiMH cells can be that sensitive to temperature!

I take care to keep my measurements timing constant for all steps, specially at the higher current where the voltage reading will be change quickly if the load is maintained.

Brad, you are right. I am a 'variable throttle' kind of pilot. I prefer the long duration easy flight, and I also like to keep some 'emergency power reserve' when flying.

I did came to the conclusion that this pack is ready for retirement, and did order a new set of Sanyo 720 to built a new one. For sure I will run the same test on the new pack ( after the inial break-in period ) to get the new condition baseline.

Regards
André

gkamysz · Feb 06, 2003, 11:08 PM

After 50 minutes even with the brief high current periods I don't expect the temperature to increase.

Greg

Brad S · Feb 07, 2003, 09:54 AM

Agree with Greg...in this particular test, temperature probably isn't a big player. Most of the temperature increase (from about 20deg C to 35deg C max) would have occured during the last several minutes of the discharge. If any, the temperature change from the 1 minute to 10 minute measurement points should be very small. Still think that the slope difference between these two measurement points is primarily due to an interaction between the voltage measurement timing and the initially decaying cell baseline voltage.

Brad

Feb 05, 2003, 10:01 AM	#1
sail_power Registered User Join Date: Nov 2002 Location: Montréal, QC, Canada Posts: 44	Internal resistance mystery Hi I like to know if someone have an explanation for the behavior of my battery pack. I did find out that during the first moments of the discharge, the internal resistance was higher than after few minutes. In order to verify this, I make a set-up to test the internal resistance during a controled discharge. The basic discharge rate is close to 1C, and the voltage is measured after 1 minute, then at every 10 minutes, for temporary discharge rate from close to 1C up to close to 6C. As soon as the measurement are completed, the 1C discharge rate is restored. These data are put on a graph, and the slope of the 'curve' is the dynamic internal resistance. The internal resistance per cell is 198mOhms after 1 minute, then 145mOhms almost constant until 80% dischared. What will explain that the initial resistance is 35% higher? The pack temperature does not increase much during the test, maybe from 20C to 30-35C, does this can explain the phenomen? This is a 'cheap' no brand 7 X AAA NiMH pack, assumed capacity 550-600mA. It was used for few 'unsuccessfull' flight at about 6C, before assign to 3-4C task. But 145mOhms still look very high DC dynamic resistance. Does someone have run such test with high quality AAA like the Sanyo 720mAh? I will be surprise if they really meet the 30mOhms listed in the spec, since this data is AC resistance. Another surprise is that the dynamic resistance remain pretty constant on the 1C to 6C range. Maybe this is because each high current discharge last for less than 10sec, before the 'low rate' 1C is restore. Any explanation is welcome. Regards André Attached Thumbnails 7.5 KB · Views: 300

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Matching of cells with Apex internal resistance meter.	John Jennings	Batteries and Chargers	9	Nov 14, 2002 01:25 AM
Measuring internal resistance?	JBowling	Batteries and Chargers	3	Oct 09, 2002 12:20 PM
Qual Comm internal resistance	Rob Hair	Parkflyers	1	Feb 21, 2002 11:57 PM
internal resistance of 2/3aaa nimh?	singlespeed7	Parkflyers	0	Nov 05, 2001 06:03 PM
Battery Internal Resistance	E-Man	Power Systems	2	Jul 24, 2001 11:06 PM

Feb 06, 2003, 12:02 AM	#2
gkamysz Dieselized User Join Date: Feb 2000 Location: Chicagoland Posts: 4,764	Your data appears to describe a typical inexpensive NiMH cell. It's hard to understand what you are asking but if you look at a typical discharge curve the voltage is lower when the battery is near discharged. This is what you have shown on your graph. 0.145Ohm for a 7 cell pack would be very good for a AAA cell. Sanyo 720 are listed with .040 Ohm resistance. I have measured it to be approximately the same when discharged warm at 3.5A(4C). Don't expect any other manufacturers AAA to have lower resistance than this. I have never heard of dynamic resistance. I don't know what characteristic of the cell it describes. Greg

Feb 06, 2003, 08:47 AM	#3
sail_power Registered User Join Date: Nov 2002 Location: Montréal, QC, Canada Posts: 44	Sorry Greg, sometime I have hard time to make my questions clear, English is a second language to me. 1- What I don't understand the 'improvement' in internal resistance between the 1 minutes and 10 minutes measurement. The others curves are in a easy to understand pattern. 2- The 145mOhms resistance is per cell. The total for the 7 cells pack is simply not acceptable. 3- I did call this dynamic resistance, since it is taken from two differents load value, insteed of from no load / load. I think that it is more representative of the cell under uses. You said that you did measure the Sanyo AAA 720, and find a value close to the 40mOhms. For me this is an 3:1 improvement. Thanks for your advise André

Feb 06, 2003, 01:49 PM	#4
Brad S Registered User Join Date: Jan 2002 Location: Arab, AL Posts: 769	Andre, My guess is that the reason for the higher apparent resistance, shown as the slope of your plots, at the 1 minute time versus the 10 minute time is due to a measurement timing issue. If you are taking your voltages at the various rates in order from lowest to highest, there will be an ever-decreasing baseline cell voltage as the measurements are taken. This decrease will bias your calculated effective internal resistance numbers to the high side. At the beginning of the discharge, the rate of baseline voltage decrease will be greater than it is during the middle part of the discharge. Your test approach is valid and can be a much better predictor of sustained-load cell voltage than AC impedance is. However, the effective internal resistance numbers generated by this type of measurement method are significantly dependent on the test parameters, especially voltage measurement points, load level durations, and temperature effects. Personally, even with these added complexities, I find the results from these kinds of dynamic load tests to be much more useful for characterizing cell performance in my type of eflight applications (variable throttle settings) than any constant current discharges or AC impedance numbers. For pylon racing, or other constant throttle setting applications, the constant current discharge plots are by far more useful than dynamic load testing. Sorry, I've not done this type of testing on AAA cells so I don't know how your numbers would compare to other cell models. Brad

Feb 06, 2003, 02:30 PM	#5
Chris True Registered User Join Date: Dec 1996 Location: St. Mary, Maryland, United States Posts: 6,531	Nimh cells have lower resistance at higher temperatures. If you start with a cold pack you can see the resistance drop as the cells heat up initially. I've seen the following behavior - start a motor discharge, initial current is 52 amps which fades over the course of 30 seconds or so to the point that the current draw is only about 46 amps. Then it spikes back up to 50+ amps as the cells heat up - holds this for a long while then tapers down again.

Feb 06, 2003, 03:01 PM	#6
Andy W That sickening crunch.. Join Date: Jun 1999 Location: Marietta, GA Posts: 39,774	What Chris said. When you were describing the symptom, I said to myself "this sounds like a NiMH pack" and sure enough you state that at the end. In addition, the external temperature of a pack in no way reflects what could be happening inside. It takes time for the heat to propogate out. My F5B and hard-pushed NiMH packs continue to heat up after I've landed, in some cases. I dump them in a cooling tube right away.. ..a

Feb 06, 2003, 03:57 PM	#7
sail_power Registered User Join Date: Nov 2002 Location: Montréal, QC, Canada Posts: 44	Andy, that's make sense, the internal temp probably increase after the first test at high load, then stabilize. I did not tought that the NiMH cells can be that sensitive to temperature! I take care to keep my measurements timing constant for all steps, specially at the higher current where the voltage reading will be change quickly if the load is maintained. Brad, you are right. I am a 'variable throttle' kind of pilot. I prefer the long duration easy flight, and I also like to keep some 'emergency power reserve' when flying. I did came to the conclusion that this pack is ready for retirement, and did order a new set of Sanyo 720 to built a new one. For sure I will run the same test on the new pack ( after the inial break-in period ) to get the new condition baseline. Regards André

Feb 06, 2003, 11:08 PM	#8
gkamysz Dieselized User Join Date: Feb 2000 Location: Chicagoland Posts: 4,764	After 50 minutes even with the brief high current periods I don't expect the temperature to increase. Greg

Feb 07, 2003, 09:54 AM	#9
Brad S Registered User Join Date: Jan 2002 Location: Arab, AL Posts: 769	Agree with Greg...in this particular test, temperature probably isn't a big player. Most of the temperature increase (from about 20deg C to 35deg C max) would have occured during the last several minutes of the discharge. If any, the temperature change from the 1 minute to 10 minute measurement points should be very small. Still think that the slope difference between these two measurement points is primarily due to an interaction between the voltage measurement timing and the initially decaying cell baseline voltage. Brad