Thread Tools
Mar 08, 2016, 01:18 AM
ancora imparo
jj604's Avatar
Thread OP
Discussion

Internal Resistance variation with temperature of Bolt and Graphene LiPos


This thread contains info already posted in a couple of other threads but I thought it worth starting a thread just on the topic for people who might be searching for the data. I do not believe it is available readily elsewhere.

These numbers are all for 3S, 2200/2400mAh packs but it is the behaviour, not the actual numbers, that are of interest. Smaller and larger packs will show the same result just higher or lower IR figures.

I did a bit of work on a number of different packs three years ago and have included two of the results here to show how things change in the LiPo world. Although I did some testing on other brands (Revo, Dinogy, GensAce, Haiyin, ChinaHobby) back then these present results are are all Turnigy packs as I got interested in how the new Bolt 4.35V packs and the new Graphenes compared with established "premium" HobbyKing labels.

I have plotted the value of the IR of the centre cell of the pack over a range of temperature from about 2.5˚C to 50˚C. The IR is measured on fully charged packs (although they drop by a few % due to the load test drain over the length of the test). The temp variation takes place over about 12 -14 hours so is uniform throughout the pack. It is measured with a thin film thermistor under the heatshrink and the pack is in a large insulting jacket of dry rice so the temperature is an accurate measure of actual internal cell temperature. Exactly the same equipment was used for all tests in both 2013 and 2016 including the Wayne Giles ESR/IR meter and the same thermistor. These curves are tests on a single pack but I have between 3 and 4 identical ones of each bought at about the same time, all of which were load tested and compared so am confident these results are representative.

In some cases I did the testing over two separate runs and the agreement in the overlapping temperature range was extremely high so I have shown only a single data point at any particular temperature.

The two solid lines are the values I measured for a standard Nano Tech45C and an A Spec Nano Tech 65C in 2013. The four plots with the crosses are the values I have just measured for a current NanoTech 45C, a current A-Spec G2 65C, and the new Bolt 4.35V 65C and Graphene 45C packs.

I have fitted a polynomial curve (the fine dotted lines) to the last 4 which allows accurate extrapolation down to 0˚C. The fit is better than R = 0.9999 for all four curves.

Comments:

1) The current ASpec G2 packs are dramatically worse than the old A-Specs, particularly at low temperatures.
2) The standard NanoTech 45C pack has improved across the board since the 2013 version.
3) The Bolt packs beats all the othes I have measured to date for low IR across the whole temperature range except for......
4) The Graphene pack which is uniformly the lowest IR. It also shows superior IR at low temperatures.

Take away from this is that:

1) LiPo cells seem to be improving significantly (except for the ASpec G2s).
2) If you fly at low temperatures and expect high currents you need to choose your packs carefully!

The results for both the Bolts and the Graphenes are better than anything I have seen to date. I would dearly love to carry out the same measurements on the new Revo Blend 435's and other Revo premium cells but unfortunately Revo cannot supply to Australia.

Hope some people find this data useful. It is not intended as a general judgment on the superiority or otherwise of Bolts and Graphenes. There are other opinion-rich threads already running for that.

John
Sign up now
to remove ads between posts
Mar 08, 2016, 05:47 PM
ancora imparo
jj604's Avatar
Thread OP

Same data over the common pack temperature range


Someone PMd and asked for the same data but over the more common range of temperatures they fly at so here it is. Makes it easier to see the differences.

John
Mar 08, 2016, 06:25 PM
Registered User
Quote:
Originally Posted by jj604
Someone PMd and asked for the same data but over the more common range of temperatures they fly at so here it is. Makes it easier to see the differences.

John
If you look at the Graphene and Bolt plots, you can see that at 35*C which is where you would expect them to be operating if treated reasonably, the cell IR is 2m.ohms which means that a 5000mAh cell would be <1m.ohm.
Thus a 6S 5000 pack gives the user a nominal 23 V supply with a source resistance of <10m.ohms assuming he uses sensible connectors such as Joe's 5.5mm solid bullets.

Translating that into efficiency figures means that the supply system up to the ESC is over 96% efficient @ 2KW and over 94% efficient @ 3KW which puts it into perspective. Pretty impressive figures.

Wayne

Ruined the other test though, John!!
Mar 09, 2016, 12:25 AM
Registered User
Nicely done John!

Your empirical and well-controlled testing corroborates my flight observations at my field. Winters are rather mild here in Southern California but I have a few packs that perform rather poorly when below ~60F and need to be pre-heated to obtain decent performance. Most notably are my Revolectrix Diamond 60C packs. They just do not like cool weather at all.

In contrast, I have used my Bolts and Graphenes when it was ~55-60F recently and barely noticed any difference from when used on warmer days. They're obviously my go-to packs and I now reserve my other packs that do not like the cold for warmer days as pre-warming to prevent damage due to internal heating can be cumbersome.

Mark
Mar 09, 2016, 12:53 AM
ancora imparo
jj604's Avatar
Thread OP
Mark, the Revo 60C I tested in 2013 has a virtually identical temperature curve to the Turnigy G2. It is tempting to say that they are the same cell from the same manufacturer.

As we found back then this cell has very high IR at low temperatures, but the curve crosses over many of the other brand "better" packs at around 45˚C and after that is lower.

This accounts for the poor performance at low temperatures yet their surprising success at high current rates IMO. At rates close to their maximum (about 30-35C in practice) they heat up very quickly and deliver good power with reasonable voltage drop.

In winter or at lower rates like 10-20C where many of us fly they are disappointing.

The newer cells like the Graphene and the Bolts are superior across the board however. They are simply in a different league. I would like to test the newer Revo Blend 435 but unfortunately we cannot get them here.

John
Mar 09, 2016, 08:50 AM
"An occasional user"
Superb tests jj604! Any change to test sub zero IR's? When flying/driving at -20C temps, Lipos often drop to sub 0C temps too. Like my fingers...
Mar 09, 2016, 02:05 PM
Registered User
Quote:
Originally Posted by DjZorlag
Superb tests jj604! Any change to test sub zero IR's? When flying/driving at -20C temps, Lipos often drop to sub 0C temps too. Like my fingers...
John has been struggling to get the temperatures down to 2*C because of the Aussie weather.
If you can get him to move back to Scotland where he originated, you might have a chance!

You definitely need some sort of warming device. There is a range of heater elements with very high positive temperature coefficients so that you don't need any control system; they will settle at a reasonable temperature.

Wayne

Wayne
Mar 09, 2016, 02:39 PM
"An occasional user"
How baut sitting in the freezer and doing the tests there?

I'm not really a serious flyer. More like a diy engineer who just tries new different things. So higher battery IR when flying in cold weather is not big deal. I am just curious to know what would happen in even lower temps.

If I were to better my battery/finger performance, maybe the first thing to consider would be better flying gloves / better batteries before adding external heating. But thanks for the tip! PTC resistors could surely come _handy_ in these heating applications!
Last edited by DjZorlag; Mar 09, 2016 at 07:45 PM. Reason: Made myself clearer
Mar 09, 2016, 07:38 PM
ancora imparo
jj604's Avatar
Thread OP
Hmmm... You have given me an idea.

l
Quote:
Originally Posted by DjZorlag
How baut sitting in the freezer and doing the tests there?

I'm not really a serious flyer. More like a diy engineer who just tries new different things. So freezing batteries is not a seriously big deal. I am just curious to know what would happen in even lower temps.

If I were to better my battery/finger temps, maybe the first thing to consider would be better flying gloves / better batteries before adding external heating. But thanks for the tip! PTC resistors could surely come _handy_ in these heating applications!
Mar 11, 2016, 02:33 AM
ancora imparo
jj604's Avatar
Thread OP

Turnigy IR down near -20˚C


Here is some interesting results. I left all 4 packs in my freezer overnight inside a fairly big container of rice to act as an insulator.
When I took them out I immediately measured the IR and Temp of all 4 then left the Graphene IR measuring while they warmed up naturally.

I did this as a trial run and got IR results from the Graphene a bit higher than the previous ones. So I did it a second time and I got identical higher IR so I assume it is due to the different measuring conditions as I had to make up a new thermistor lead and it may have a different resistance. In short it looks like the IR measured in this test is about 0.75mOhm higher than the high temperature runs. Conditions are also not as well controlled as the measurement is done in the open on the bench.

Overall however what is striking is how high the IR gets at about 17 degrees - and for all 4 packs. They all have IR values around 37 mOhm at -17 ˚C. And the plot of the graphene values suggest that it is heading rapidly upward. By the slope of this graph it looks like you would have infinite internal resistance at -19 to -20˚C.

One can only speculate but it seems a likely explanation is that that some freezing process in the cells is occurring at around -18˚ that effectively stops them functioning. At that temperature they are still producing a decent voltage when loaded at 16 Amps but I suspect they wouldn't at a few degrees lower.

Be interesting to see if I have 4 LiPo packs or 4 paperweights when I warm them up and use them!

John
Last edited by jj604; Mar 15, 2016 at 02:16 AM.
Mar 11, 2016, 03:12 AM
Registered User
John,

That is interesting.
The rise of 0.75 milliohms could indicate permanent damage to the packs although it could be measurement as you say.

Does this explain why the Inuit community are not big in electric flying?

Wayne
Mar 15, 2016, 01:06 AM
"An occasional user"
Quote:
Originally Posted by jj604
...
That's what I'm talking about! No wonder voltages seems to sag in sub zero temps.

Although the higher IR makes the battery produce more heat, it often isn't enough to keep the pack warm enough.
Mar 15, 2016, 02:09 AM
ancora imparo
jj604's Avatar
Thread OP
The 0.75mOhm difference turned out to be a simple error in transposing data. I am getting very good agreement across the temperature range when measuring at different times now. Where two runs have temperatures in common the IR numbers agree very closely.

John


Quote:
Originally Posted by Wayne Giles
John,

That is interesting.
The rise of 0.75 milliohms could indicate permanent damage to the packs although it could be measurement as you say.

Does this explain why the Inuit community are not big in electric flying?

Wayne
Mar 15, 2016, 02:14 AM
ancora imparo
jj604's Avatar
Thread OP
I will be updating soon with some better low temperature data. The graph in Post #10 is slightly misleading. I had not remembered that the maximum cell IR the meter can record is 37 mOhm. So those points on the graph for the Bolt, A-Spec and Nano pack are an artefact and too low. The IR is in fact even higher than that at low temperatures.

John
Quote:
Originally Posted by DjZorlag
That's what I'm talking about! No wonder voltages seems to sag in sub zero temps.

Although the higher IR makes the battery produce more heat, it often isn't enough to keep the pack warm enough.
Jun 07, 2016, 03:30 AM
ancora imparo
jj604's Avatar
Thread OP

Results of IR measurements at very low temperatures


Sorry, I didn't get round to posting these results earlier.

They show the variation of IR as the temperature drops. The spectacular rise in value below about -10˚C is notable.

One could speculate on what is going on: some components of the electrolyte are perhaps changing phase ("freezing"), the reason for difference in behaviour of the traditional A Spec pack compared with the newer Bolts and Graphenes.

But it would be simply speculation. What is clear is that operation below zero degrees is going to result in very significant performance loss! It is fairly evident that none of these LiPos will work below -20˚C.

There are small overlap discontinuities in the graphs in the range +5-+20˚C. This is because the readings were taken using two different methods but the difference over the common range is very small. Note that there are a lot fewer plotted points below zero degrees and the graphs are not as reliable as at higher temperatures because the temperature is changing at a higher rate. This probably accounts for the kinks in the curves.

John


Quick Reply
Message:

Thread Tools