Headway and Ping - IR measurements

[Tiberius]

Hi Everyone,

I finally got my battery testing together and found out it was dead easy to make some measurements of internal resistance. I've got a big electronic load that I've set up to toggle between 2 A and 10 A at 100 Hz. I can then look at the voltage of a cell or pack on the scope and see the square wave.

So far this is not automated, calibrated or traceable, but here's some preliminary results from two packs I tried.

First is 8 cells of Headway 38120 10 Ah. A 4 point measurement gives a delta V of 1.04 V, which works out as 16 milliOhms per cell. I can get into this one, since I built it myself, and the delta V looks to be evenly distributed among the cells.

Second is a Ping 36 V, 20 Ah pack. The 4 point measurement on this includes 6 inches of the Ping cabling, and the delta V is 720 mV. I'm not sure but I assume this is a 12SNP, where N is probably 4. Whatever the internal construction, if we translate that delta V into the equivalent of a 10 Ah cell, we get 15 milliOhms.

Not a lot in it, is there? The C rating is quoted as higher for the Headway cells, but a good chunk of it will end up as heat.

Both packs are about 9 months old, used and freshly charged.

Nick


Nick

 

[dogman dan]

Way over my head, but are you trying to say the ping pack is not much different in c rating from a headway?

When you see c ratings, I always wonder if each manufacturer is measuring different things. C rating for what? 1000 cycles, 100 cycles? All you ever seem to see is a 1c test anyway.

 

[Doctorbass]

The internal resistance measurement must be done with cell NOT freshly charged.

Ideally you must measure it at around 80% soc and ensure that the battery have sit few hours before to let it stabilyze.

The IR measurement of a cell must be done directly on the cell contact. You can not imagine how the resistance if the wier is high enough to indruduce error! a simple 10AWG wire have around 1miliohm er foot so if you have two wire(pos and neg) it's 2 miliohms... so with smaller wires size it increase alot.

Also RI measurement are usually done on discharge only

Doc

 

[Tiberius]

Way over my head, but are you trying to say the ping pack is not much different in c rating from a headway?

When you see c ratings, I always wonder if each manufacturer is measuring different things. C rating for what? 1000 cycles, 100 cycles? All you ever seem to see is a 1c test anyway.

Hi Dogman,

The IR measurement is not the same as the C rating, but they are related. Internal resistance is what causes the voltage drop and the heating when you draw current from the cells. The max current (C rating) a cell can supply without internal damage will be affected by other factors. That said, though, IR is something that can be measured, and you can't have high currents without low IR, so its a good guide to the "quality" of the cells.

The 1C test that is commonly done is to measure the capacity in Ampere-hours (Ah).

Nick

 

[Tiberius]

The internal resistance measurement must be done with cell NOT freshly charged.

Ideally you must measure it at around 80% soc and ensure that the battery have sit few hours before to let it stabilyze.

The IR measurement of a cell must be done directly on the cell contact. You can not imagine how the resistance if the wier is high enough to indruduce error! a simple 10AWG wire have around 1miliohm er foot so if you have two wire(pos and neg) it's 2 miliohms... so with smaller wires size it increase alot.

Also RI measurement are usually done on discharge only

Doc

Hi Doc,

These packs had had some hours since charging, so "freshly charged" was probably not the best phrase. But they were 100% SOC rather than 80%.

I agree with you about that the measurement should be done right on the cell. But these were measurements on packs and will include the cell to cell connections. In that case, the error from a few inches of wire will be tiny. Measurements on individual cells are what we would want for comparing cells, technologies and manufacturers, but pack level IR is what counts in the real world.

They were all discharge measurements. I've got the capability to set the lower and upper currents (up to 150 A or 750 W) that the load switches between and to choose 100 or 1000 Hz.

Nick

 

[liveforphysics]

Normally Ri testing is done at nominal voltage for the cells and at room temperature. Small temperature changes can make a very big difference in some cells Ri.


DocBass is correct, it should be measured at the tabs if you want the value for the cell. However, I've yet to just touch a cell to my motor to make my bike run, we generally run them in a pack, so if the connections in your pack are standard for your style of pack, this is perhaps a more valuable bit of test data than a per/cell Ri test. I like your method of testing too! That's a pretty good idea to watch the step difference on a scope.

My charger has a built-in Ri testing tool. Who knows what the accuracy is, could be off by a huge margin, but it indicates 8mOhm for a whole 10s40Ah LiPo pack, with the resistance of the pack leads and charging leads being included in this measurement. That would be about 0.8mOhm per 40Ah series group, or about 3.2mOhms per 10Ah cell (to set things relitive to your tests).

That would be about 1/4th the voltage drop, and 1/4th the energy wasted in heating the battery.

I'm very suprized to hear that your Ping cells test similar to Headways!

DocBass, we need you to get going on putting those new cell testing tools to use! And I need to quit forgetting to send you a care package with some LiPo cells in it.

 

[dogman dan]

I may never fully understand some of this stuff, but I like it that the ping cells have a decent internal resistance. I kinda knew that, just based on the small amount my pack heats up in use. But I am suprised that they rate even close to the headways.

 

[ZapPat]

The older headway cells did have about 16mohms per cell, so that makes sense. Ping's cells were better than these older headway cells, but the newer blue headways have about 9mOhms per 10Ah cell, a fair amount better than Ping's V2 cells (which I think I measured at about 15mOhms per 10Ah, similar to what Tiberius measured).

As for LiPo packs, I've repeatedly seen around 10mOhms Ri per 5Ah cell on some Zippy flightmax 15C packs I have here. That would make these Lipos have almost twice better Ri as new blue Headway's, and three times better than Ping's present cells. I have a large number of PingV2 cells lying around here now, so I'll remeasure the Ri of some individual cells to be sure.

Tiberius, did your Ping pack Ri measurements include the stock BMS? The FETs used on these are really cheap and have high resistance, so would be worth checking the Rds. Cool test setup BTW! I'll have to try out the high frequency dual current pulse test idea, and maybe try out different frequencies too. I just managed to implement a functional digital current control loop in my controller prototype, so this sounds like the pefect kind of thing to test it on.

 

[Kurt]

On the subject of new blue headway cells vs old headway brown cell. I understand some testing was carried out on the older brown cells and I have seen the result of both test. I am not disputing the results of the test.

I do think everybody is making to much of a classification of headway cells based on there appearance. I have both the brown cells and the blue cells and I was told by my supplier who asked headway directly if there is any difference between the brown cells and there blue cells and the answer was no.

What I make of this is that perhaps the very first of then gen 1 brown cells had high IR . Some changes were made to improve the IR. So first the charges were made to the cells then self. Then some time later they made the change to the blue packaging.

As a crude way of confirming this to myself both my blue and brown cell battery packs have the same voltage sag under high load.If one pack really was twice the IR I would be able to see it.

Kurt.

 

[ZapPat]

On the subject of new blue headway cells vs old headway brown cell. I understand some testing was carried out on the older brown cells and I have seen the result of both test. I am not disputing the results of the test.

I do think everybody is making to much of a classification of headway cells based on there appearance. I have both the brown cells and the blue cells and I was told by my supplier who asked headway directly if there is any difference between the brown cells and there blue cells and the answer was no.

What I make of this is that perhaps the very first of then gen 1 brown cells had high IR . Some changes were made to improve the IR. So first the charges were made to the cells then self. Then some time later they made the change to the blue packaging.

As a crude way of confirming this to myself both my blue and brown cell battery packs have the same voltage sag under high load.If one pack really was twice the IR I would be able to see it.

Kurt.

Do you mean that even your newer headway cells have around 16mOhms internal resistance, or that your brown ones and the blue ones are both close to 9mohms?

 

[Kurt]

Do you mean that even your newer headway cells have around 16mOhms internal resistance, or that your brown ones and the blue ones are both close to 9mohms?

Like I said I haven't measured them. What i am trying to say is don't judge the headway cells purely on the wrapping of the cell. I would be more inclined to say that my brown cells are 9mohms than my blue cell being 16mohms.

I say this because both seem to have very little voltage sag running with x5 and 48a controller and even towing trailers up long steep hills at full throttle they haven't ever got over ambient temp.

I am not doubting there was some average cells produced with hi IR in the very beginning.

Kurt.

 

[Tiberius]

Hi Guys,

I've been busy for a bit, and in another country, but I had another go at the internal resistance tests today.

Docbass said the Ri tests should be done at about 80% SOC. My test setup allows me to run the test continuously as the battery discharges, so I watched both the Headway and Ping packs as they ran down from 100% to about 70%. There was no significant change in the Ri for either.

I can also do the tests at 1 kHz instead of 100 Hz. Again, no significant difference for either.

ZapPat, the measurement on the Ping pack does include the BMS so it will include the Ron of any cut off FET in there. But compared to 12 series cell, I would expect this to be a small error. If the FET Ron is large, then the cells are even better.

Nick

 

[Lapwing]

Excellent data!

I am really pleased to see that your readings for Lithium internal resistance (IR) and state of charge confirm what I have found also using my old analog HP 4328A. At first I thought I was doing something wrong. :|

It seems generally with lithium state of charge does not affect IR measurably. That might explain why in a pack a low cell can be driven off a cliff so easily by higher capacity cells. It acts a wire even when driven to reverse polarity. Bad for the cell (it dies) but the rest keep on pumping, until they too go over the cliff. Keep in mind that temp does affect IR, so hot cells have less IR than cold cells.

Some interesting data for other chemistries, can be found here . http://www.batteryuniversity.com/partone-22.htm

See especially figure 5 & 6 which is where (for NiMh and lead) this idea of resting the battery comes from. I messed around with Nickel-Cadmium's and found that another factor profoundly affecting IR is temperature. Again it depends on the chemistry as to how much heating the batteries decreases IR. Makes a big difference on Ni-Cd's and lead. After charging Ni-Cd get pretty hot, so waiting temp equalises makes a big IR difference.

 

[Doctorbass]

Totally true!

That's why if we want to do an ACCURATE IR measurement we MUST have constant temp conditions during EVERY measurement and OVER EVERY MEASUREMENT that are compared!!

and.. that's why it is so difficult to prepare these conditions for me for every measurement i do for the E-S comunity!

Winter is the best to do that.. Temp control is easier!!

Doc