Shop our Airplanes Products Drone Products Sales
Thread Tools
Dec 02, 2015, 07:55 PM
Registered User
u2builder's Avatar
Thanks. That is very helpful. Could someone provide me with the equation that predicts the approximate variation of typical cell IR with temperature if that exists.

I would like to come up with a calculator that roughly calculates the IR of a battery pack (at 72F) , given the actual pack temperature at the end of a flight, the actual current draw at full throttle, and the user specified allowable voltage drop at that current.

I am finding that with some of my older cells IR is so high that the voltage drop at full throttle near the end of a flight may be a volt or more on my 3S batteries which would mean I could be perilously close to my 3.3V ESC cut off if I was taking a battery down to around 11 no load volts These packs have plenty of capacity in them when I land but my telemetry reports dangerously high voltage drops.

In real world terms I could record the voltage drop reported under max throttle, cool down the battery, and simply measure the IR, but it would be fun to try to calculate it. And it is basically winter so I really can't test things in the field right now anyway.

My general sense from flying this summer, is that my (older) 3S, 2200 mah batteries, which have cell IR's in the 12 range (pack IR's of 50) , are beginning to have too high a voltage drop even though they can carry more current than my motor draws according to the lipoly calculator. Voltage drop seems to be more of a limiting factor than current with the type of planes I am flying.
Sign up now
to remove ads between posts
Dec 02, 2015, 08:29 PM
ancora imparo
jj604's Avatar
Afraid not. I couldn't find any reputable information on the variation of model LiPo cell IR with temperature so some time ago I measured it.

I used an automated Waynometer to measure the IR of the centre cell in a variety of 3S 2200mAh packs as they were slowly cooled from 50˚C to just above 0˚C over a 12hr period. This was slow enough to ensure the pack was at uniform temperature and in any case the temperature was measured using a thin film thermistor inserted into the pack between the centre and an outer cell. All my results were consistent.

A typical graph is attached. You can see the curves are all a falling curve but they have different slopes and some cross over (e.g. at points Note1 and Note2).

I got pretty excited when I did the first test as it looked like a logarithmic curve and plotted against log IR I got what was clearly two straight lines with a distinct change of slope at about room temperature. I thought I had discovered something fundamental and useful. Alas it was not to be.

Further tests showed that the curves for different packs (chemistries?) were all of different shape and I was never able to find a straightforward equation of any kind that fitted all curves. To get a fit for some packs you had to use a very complex power curve with numerous coefficients and constants - not in any way useful, and each was unique.

Bottom line is that IMO no simple universal equation relating IR to temperature for all LiPo packs exists.

John
Quote:
Originally Posted by u2builder
Thanks. That is very helpful. Could someone provide me with the equation that predicts the approximate variation of typical cell IR with temperature if that exists.

I would like to come up with a calculator that roughly calculates the IR of a battery pack (at 72F) , given the actual pack temperature at the end of a flight, the actual current draw at full throttle, and the user specified allowable voltage drop at that current.

I am finding that with some of my older cells IR is so high that the voltage drop at full throttle near the end of a flight may be a volt or more on my 3S batteries which would mean I could be perilously close to my 3.3V ESC cut off if I was taking a battery down to around 11 no load volts These packs have plenty of capacity in them when I land but my telemetry reports dangerously high voltage drops.

In real world terms I could record the voltage drop reported under max throttle, cool down the battery, and simply measure the IR, but it would be fun to try to calculate it. And it is basically winter so I really can't test things in the field right now anyway.

My general sense from flying this summer, is that my (older) 3S, 2200 mah batteries, which have cell IR's in the 12 range (pack IR's of 50) , are beginning to have too high a voltage drop even though they can carry more current than my motor draws according to the lipoly calculator. Voltage drop seems to be more of a limiting factor than current with the type of planes I am flying.
Dec 03, 2015, 10:10 AM
Registered User
u2builder's Avatar
This is extremely helpful,

It suggests to me that all of these sample batteries tend to have a similar IR when "warm" at 50 C, or in other words that the warm IR does not depend all that much on C rating. Also, in my operating range, 20 to 50 C, the decrease in IR is fairly linear and not all that great a slope.

What I have noticed is that as my batteries age, probably due to age as much as cycles, their cell IR increases, say from around 3 or 4 to say 12 after a few years. This degradation seems fairly independent of C rating, as my more expensive so called 60 C batteries wind up with around the same aged IR as my 30 C and 40 C batteries over about the same time period. High C batteries seem to degrade just as fast in my limited experience.

As I have mentioned, and according to to the Lipoly tool, batteries with cell IRs up around 12 are causing a high voltage drop (roughly 1 volt per 3 cells at full throttle for my 3 S 2200-3300 mah packs) even though they can still carry more than enough current to prevent damage or overheating. A 1 volt drop becomes significant when a 3S pack is down to 11 no load volts due to the ESC cut off of around 10 volts. This suggests that having a reasonably low IR measured at 72 degrees (meaning a fairly new battery) is more important than having a high C rating. So called high C batteries are much more expensive than say 30 C batteries. It suggests to me that I would be better off to purchase low C batteries and replace them more.

I guess what I am saying is that a fresh new lower C battery is going to perform better than an aged high C battery, especially when it is warmed up to 50C which is probably a fairly typical flight temperature by the middle/end o a flight, at least for me.

Does this make sense?
Dec 03, 2015, 04:09 PM
Registered User
u2builder's Avatar
I also tried to curve fit the iR vs T graphs shown above and also found that none of the standard curves were a perfect fit as the IR tends to get quite flat as the temperature approaches 50C.

I decided that my batteries typically come down warm. I put a dozen the batteries on the warming shelf of my wood cook stove which was running about 38C. I turned them frequently and checked them until I felt that they had stabilized at 38C which "felt like" my batteries after a typical flight (I have never lit up the 45C temp indicator on my Revos). Not perfect but my wood oven was not an option. I had previously measured the cell IR at 22C. They varied between a low IR of 3.48 for a new Turnigy blue to 13 for a several year old G-Force Nano. I had some two year old Revo Platnum 40's at 5.44 in the mix. I calculate the "slope" of a line connecting these values. It ranged from about .5 for my older high IR batteries to a low of about .1 for my new Turnigy. I decided to divide the slope by the IR(22) and noticed that I got a pretty much constant value of about .033. I made a little formula for calculating the IR(38):

IR(38)=IR(22) - .033* IR(22)*(38-22)
IR(38)=IR(22) - (1-,033* 16) = .47 IR(22)

At this point it became obvious just looking at the numbers that the IR dropped about in half as the temperature increased from room temperature to about 38C (100F) for a variety of 3S packs. I went back and looked at jj604's curves for the same temperature range and observed "roughly" the same result.

I caution that my experiment was done without precise temperature control and the results should be considered "approximate" and is obviously only valid for the temperature range of 22 to 38. Still, it is interesting a dozen different batteries of different conditions all had a similar slope/initial IR value. I also observed that the difference between the pack IR and the sum of the cells varied from around 4.5 and 6.5 for my 3S cells with Deans connectors.
Dec 03, 2015, 04:39 PM
Registered User
Quote:
Originally Posted by u2builder
This is extremely helpful,

It suggests to me that all of these sample batteries tend to have a similar IR when "warm" at 50 C, or in other words that the warm IR does not depend all that much on C rating. Also, in my operating range, 20 to 50 C, the decrease in IR is fairly linear and not all that great a slope.

What I have noticed is that as my batteries age, probably due to age as much as cycles, their cell IR increases, say from around 3 or 4 to say 12 after a few years. This degradation seems fairly independent of C rating, as my more expensive so called 60 C batteries wind up with around the same aged IR as my 30 C and 40 C batteries over about the same time period. High C batteries seem to degrade just as fast in my limited experience.

As I have mentioned, and according to to the Lipoly tool, batteries with cell IRs up around 12 are causing a high voltage drop (roughly 1 volt per 3 cells at full throttle for my 3 S 2200-3300 mah packs) even though they can still carry more than enough current to prevent damage or overheating. A 1 volt drop becomes significant when a 3S pack is down to 11 no load volts due to the ESC cut off of around 10 volts. This suggests that having a reasonably low IR measured at 72 degrees (meaning a fairly new battery) is more important than having a high C rating. So called high C batteries are much more expensive than say 30 C batteries. It suggests to me that I would be better off to purchase low C batteries and replace them more.

I guess what I am saying is that a fresh new lower C battery is going to perform better than an aged high C battery, especially when it is warmed up to 50C which is probably a fairly typical flight temperature by the middle/end o a flight, at least for me.

Does this make sense?
In a word, Yes.

High C values really are, for the sports flyer, just a sales gimmick and an excuse to increase the price. 20C and 25C packs are likely to have genuine ratings, paying a lot more for "40C" or "60C". will at best give you a slight performance increase, so I agree with your analysis.

Although the Lipotool is conservative, it is intended to produce a max current value which you can take continuously throughout a full discharge which you would never do in practice. Who would discharge a lipo in less than 3 minutes? That is equivalent to 20C. So why do we need 40 C, particularly as you could take greater than 20C for short periods without a problem.

It is good practice to warm lipos in winter. As you have noted the general IR plots are approaching linearity at higher temperatures; operation at low temperatures is best avoided.

Wayne
Dec 03, 2015, 06:01 PM
Registered User
u2builder's Avatar
I can now answer the question I started with. What is the maximum safe cell IR for my batteries measured on the bench in actual flying conditions. Let's say I have taken my 3S 2200 mah pack down to around around 11 volts (based on my typical flight times for that pack or telemetry). Let's say my ESC cutoff is 3.3/cell = about 10. Let's say I don't want to drop below 10.25 so I can afford a .75 volt drop at full throttle and I know I am drawing 30A at full throttle and I expect my batteries to be at 38C based on experience.

.75V /30A =.025 ohms IR per pack =25 miil ohms x2(to correct to 22C) - 6 (for pack wiring) /3 cells = 14.6 mil ohms/cell would be approximate max.

This corresponds to what my telemetry is telling me when I fly batteries with this elevated IR. Time to junk them since I don't have low amp draw planes.
Dec 09, 2015, 05:49 AM
Registered User

ESR Meter


Hello Wayne,

I am very much interested in your ESR meter. Kindly let me know the details so that I can make the payment by Pay pal. I need the item to be shipped to India Mumbai, will give you the shipping address as soon as I receive your mail.

My email I/D is gejacob@gmail.com

Awaiting an early reply. I hope I am able to get the your ESR meter before Christmas!!!

Thanks and best regards.
George Jacob.
Dec 09, 2015, 02:24 PM
Registered User
Quote:
Originally Posted by gjay
Hello Wayne,

I am very much interested in your ESR meter. Kindly let me know the details so that I can make the payment by Pay pal. I need the item to be shipped to India Mumbai, will give you the shipping address as soon as I receive your mail.

My email I/D is gejacob@gmail.com

Awaiting an early reply. I hope I am able to get the your ESR meter before Christmas!!!

Thanks and best regards.
George Jacob.
George,

E-mail sent.

Wayne
Jan 09, 2016, 03:56 AM
Registered User
Hi Wayne

The thickness of the discharge leads won't make a difference to the cell measurements, correct? So if one has 14AWG wire and the other 12AWG, it won't make a difference to the cell measurements of the 2 batteries.

Thanks

Hennie
Jan 09, 2016, 06:00 AM
Registered User
Quote:
Originally Posted by hethjo
Hi Wayne

The thickness of the discharge leads won't make a difference to the cell measurements, correct? So if one has 14AWG wire and the other 12AWG, it won't make a difference to the cell measurements of the 2 batteries.

Thanks

Hennie
Sort of, maybe
If the cells voltages are checked at the balance leads, no.

The difference in wire size would impact the maximum current draw slightly with the same load across the whole battery. But, if the load current was set to be the same, there should be no difference.

I'm too lazy to try to get lab quality measurements.
Instead - - -
The same battery should have reasonably similar IR readings cell to cell.
IR reading twice what they were or are with a new lipo of the same size and type indicate
possible bad cells, or at least diminished capacity in terms of voltage at higher currents.

So, if a new lipo were to read ~4 milli ohms per cell, using my DP-8 charger,
and a similar lipo had readings of more than about 7 milli ohms, my next step would
be to use a slightly modified "old school" (carbon pile) automotive load tester, and see what the overall
battery voltage is under a significant load that is slightly greater or the same as I'd expect
during use. The other guideline is usually up to 1/2 the maximum "C" rating.
Obviously, high current load testing is not done for an extended period.

For instance
At takeoff and early climb out, my P-51 may have 70-75 Amp current draw.
The Lipo is a 6S5000mah rated at 40C.

First using the automotive tester, will the fully charged Lipo provide the needed voltage and current?
Assume first, 3.7v per cell minimum, and 70-75 Amp current (Yes, fly, no don't)
What very short term current results in about 3.2-3.3v per cell ?
Is this consistent with the rated maximum C rating?
Is there enough delta to indicate that a touch and go or an aborted landing is possible?
Is the rated maximum "C" real or a pipe dream?
The good 6S5000 lipos I use will provide twice or more of the take off current ~ 150A,
and still be well within the maximum "C" rating.
Average good lipo flight times are between 5 and 6 minutes with a gross weight of about 10 1/2 Lb.
A bad or marginal lipo (or a discharged one) will usually cause power loss and a resulting ESC "warble" after take off
when full throttle is set. (Get it on the ground while you still have about 1/2 power available) The warble indicates that the
average cell voltages are dropping to about 3.2-3.3v in flight when high throttle is set..

The "slight modification" mentioned changes the load tester's analog voltmeter from 16vdc to 32vdc full scale.
Last edited by chuck75; Jan 09, 2016 at 06:37 AM.
Jan 09, 2016, 06:16 AM
ancora imparo
jj604's Avatar
Actually, the answer is no. Wayne will no doubt respond but his meter is a constant current load with 4 wire voltage measurement - that is the cell voltage is measured at effectively no voltage drop.

So the connecting wire size makes no difference. The current is set by the meter not any connecting resistances.

The difference between the pack IR and sum of the cell IRs will however be a function of the total connector resistance.
Jan 09, 2016, 02:45 PM
Registered User
Hennie,

Yes, you are correct and John's explanation is spot on. The meter is a constant current sink so that it will always take the same current irrespective of supply voltage and series resistance provided that the sink is not saturated by a very low cell voltage. ( <2V).
The sense lead measures the voltage drop via the balance connector at the terminations of the cell so that any resistance in the leads and connectors is completely ignored.

Wayne
Jan 10, 2016, 01:00 AM
Registered User
Thanks Chuck, John and Wayne.

Why I wanted confirmation, is I measured at 21.8 degrees, a Gens Ace 1300mA 25C at 9.24, 9.72 and 9.52 and a Tattu 1300mA 45C at 9.64, 10.00 and 9.48. That say to me that the C rating for both are essentially the same, even though the Tattu is supposed to be 45C.

These are brand new batteries, not used at all.

I curious, has anyone seen a 3S 5000mA that lives up to its C rating? I bought a 25C and measured only 18C.

Hennie
Jan 10, 2016, 12:16 PM
Registered User
Ken Myers's Avatar

ESR meter anomalies


Admiral 3S 1100mAh 25C,
Pack ESR 70.8mOhm, cell #1 22.12mOhm, cell #2 23.92mOhm, cell#3 23.68mOhm, cells sum = 69.72mOhm, ESR - cell sum = 1.08mOhm
Pack ESR 68.7mOhm, cell #1 21.4mOhm, cell #2 23.32mOhm, cell#3 23.36mOhm, cells sum = 68.08mOhm, ESR - cell sum = 0.62mOhm
Pack ESR 64.5mOhm, cell #1 20.08mOhm, cell #2 21.96mOhm, cell#3 22.36mOhm, cells sum = 64.4mOhm, ESR - cell sum = 0.1mOhm
Pack ESR 67.5mOhm, cell #1 20.96mOhm, cell #2 22.76mOhm, cell#3 22.56mOhm, cells sum = 66.28mOhm, ESR - cell sum = 1.22mOhm

Turnigy 3S 1000mAh 30C,
Pack ESR 89.1mOhm, cell #1 29.04mOhm, cell #2 32.16mOhm, cell#3 28.8mOhm, cells sum = 90mOhm, ESR - cell sum = -0.9mOhm
Pack ESR 81.6mOhm, cell #1 26.36mOhm, cell #2 29.44mOhm, cell#3 26.24mOhm, cells sum = 82.04mOhm, ESR - cell sum = -0.44mOhm
Pack ESR 91.5mOhm, cell #1 29.4mOhm, cell #2 32.52mOhm, cell#3 28.88mOhm, cells sum = 90.8mOhm, ESR - cell sum = 0.7mOhm
Pack ESR 87.3mOhm, cell #1 28mOhm, cell #2 31.08mOhm, cell#3 27.5mOhm, cells sum = 86.58mOhm, ESR - cell sum = 0.72mOhm

Revo 3S 1000mAh 60C,
Pack ESR 27.3mOhm, cell #1 7.48mOhm, cell #2 7.4mOhm, cell#3 7.12mOhm, cells sum = 22mOhm, ESR - cell sum = 5.3mOhm
Pack ESR 27.6mOhm, cell #1 7.56mOhm, cell #2 7.48mOhm, cell#3 7.16mOhm, cells sum = 22.2mOhm, ESR - cell sum = 5.4mOhm
Pack ESR 30.9mOhm, cell #1 8.8mOhm, cell #2 8.72mOhm, cell#3 8.52mOhm, cells sum = 26.04mOhm, ESR - cell sum = 4.86mOhm
Pack ESR 27.9mOhm, cell #1 7.72mOhm, cell #2 7.6mOhm, cell#3 7.32mOhm, cells sum = 22.64mOhm, ESR - cell sum = 5.26mOhm

GensAce 3S 1000mAh 25C,
Pack ESR 30mOhm, cell #1 8.88mOhm, cell #2 9.04mOhm, cell#3 8.96mOhm, cells sum = 26.88mOhm, ESR - cell sum = 3.12mOhm
Pack ESR 24.3mOhm, cell #1 6.68mOhm, cell #2 6.68mOhm, cell#3 6.92mOhm, cells sum = 20.48mOhm, ESR - cell sum = 3.82mOhm
Pack ESR 29.7mOhm, cell #1 8.52mOhm, cell #2 8.76mOhm, cell#3 8.72mOhm, cells sum = 26mOhm, ESR - cell sum = 3.7mOhm
Pack ESR 25.5mOhm, cell #1 7.04mOhm, cell #2 7.12mOhm, cell#3 7.08mOhm, cells sum = 21.24mOhm, ESR - cell sum = 4.26mOhm

ElectriFly 3S 1000mAh 30C,
Pack ESR 66.6mOhm, cell #1 19.36mOhm, cell #2 22.28mOhm, cell#3 22.4mOhm, cells sum = 64.04mOhm, ESR - cell sum = 2.56mOhm
Pack ESR 51.9mOhm, cell #1 14.2mOhm, cell #2 17mOhm, cell#3 18.52mOhm, cells sum = 49.72mOhm, ESR - cell sum = 2.18mOhm
Pack ESR 62.4mOhm, cell #1 18.24mOhm, cell #2 20.68mOhm, cell#3 22mOhm, cells sum = 60.92mOhm, ESR - cell sum = 1.48mOhm
Pack ESR 57mOhm, cell #1 16.32mOhm, cell #2 18.76mOhm, cell#3 19.96mOhm, cells sum = 55.04mOhm, ESR - cell sum = 1.96mOhm

Dinogy 3S 1000mAh 30C,
Pack ESR 57.3mOhm, cell #1 18.32mOhm, cell #2 17.96mOhm, cell#3 17.76mOhm, cells sum = 54.04mOhm, ESR - cell sum = 3.26mOhm
Pack ESR 44.7mOhm, cell #1 13.76mOhm, cell #2 13.6mOhm, cell#3 13.84mOhm, cells sum = 41.2mOhm, ESR - cell sum = 3.5mOhm
Pack ESR 50.1mOhm, cell #1 15.6mOhm, cell #2 15.48mOhm, cell#3 15.44mOhm, cells sum = 46.52mOhm, ESR - cell sum = 3.58mOhm
Pack ESR 44.7mOhm, cell #1 13.6mOhm, cell #2 13.44mOhm, cell#3 13.4mOhm, cells sum = 50.44mOhm, ESR - cell sum = 4.26mOhm

Dinogy 3S 1000mAh 65C,
Pack ESR 54.3mOhm, cell #1 17.44mOhm, cell #2 17.32mOhm, cell#3 18.2mOhm, cells sum = 52,96mOhm, ESR - cell sum = 1.34mOhm
Pack ESR 44.4mOhm, cell #1 13.76mOhm, cell #2 13.72mOhm, cell#3 14.24mOhm, cells sum = 41.72mOhm, ESR - cell sum = 2.68mOhm
Pack ESR 51.3mOhm, cell #1 16.24mOhm, cell #2 16.28mOhm, cell#3 17.32mOhm, cells sum = 49.84mOhm, ESR - cell sum = 1.46mOhm
Pack ESR 42.9mOhm, cell #1 13.16mOhm, cell #2 13.12mOhm, cell#3 13.64mOhm, cells sum = 39.92mOhm, ESR - cell sum = 2.98mOhm

Note that the data is for different states of charge and NOT listed from oldest readings to most current. All temperatures were 22 deg C.

What I don’t understand is why the very poor batteries, the Admiral and Turnigy, have the smallest and ‘weirdest’ differences between the pack measured ESR and the sum of the individual cells.

I also do not understand why the middle of the road batteries, the ElectriFly and Dinogy, show significant differences in the pack ESR and cell sum difference.

I also have more data for these specific batteries that is demonstrating these same anomalies.

It appears that ‘better’ batteries have a larger difference between the pack ESR and the sum of the cells IR.

What’s up?
Jan 10, 2016, 01:06 PM
Registered User
Ken,

Short answer is that nothing is up.
I answered this same question a few weeks back but I cannot remember on which thread or whether it was a private message.
The figures to always use are the Cell figures because they use a Kelvin 4 wire measuring system and are accurate. Really the Pack figure is just intended for use as a quick check figure.

The reasons for the differences are several. One is that connectors we use are cheap and inconsistent and plugging and unplugging the connector between two readings will demonstrate that you get a slightly different reading almost every time. Cell readings are a lot more consistent as that error does not exist.
Other sources of error are in trying to read the connector and leads by subtracting one large sum from another; there only needs to be a small % error in the large sums to produce a much larger error in the answer.

The resolution and maths in the Pack reading is also limited by the capabilities of the microprocessor.

Just why it appears to tie up with different pack types would appear to be coincidental unless I am missing something obvious.

Wayne


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Best fo the West Jets Schwemmer Inland Slope Rebels 0 Oct 14, 2010 06:02 PM
Discussion Motor fo UH Haoye EDF Ducted Fan Unit 7 Blade 3.5inch RJPIoW Electric Ducted Fan Jet Talk 1 Sep 08, 2010 01:05 PM
Question New Low IR Lipos Cells nxtdoor High Performance 75 Sep 06, 2009 05:47 PM
Discussion ESR meter to test Nicad/NiMH ?? 1101 DIY Electronics 3 Jun 15, 2009 05:08 AM