A123 Internal Resistance Question

What should I be getting as a measurement of IR for a good A123 cell?

I'm also interested in standard IR for other cells, like Moli, Konion, or lipo if anyone has that info around.

I'm starting to measure IR for my cells as part of the testing, but I'm not sure what I SHOULD be getting since all of mine are used.

Katou

katou said:

What should I be getting as a measurement of IR for a good A123 cell?

I'm also interested in standard IR for other cells, like Moli, Konion, or lipo if anyone has that info around.

I'm starting to measure IR for my cells as part of the testing, but I'm not sure what I SHOULD be getting since all of mine are used.

Katou

Click to expand...

7mOhm +-2mOhm for a 26650 A123 M1 cell.

Good LiPo was around 2-5mOhm per 5Ah cell, but the new Turnigy Nano's just blew my mind when I tested <1mOhm Ri at 25C discharge rate (100amps on a 4Ah cell).

katou said:

What should I be getting as a measurement of IR for a good A123 cell?
Katou

Click to expand...

At what state-of-charge?
At what temperature?
At what charge or discharge current rate?

All these have a huge effect on the IR. All the values are correct, but just for those conditions. I've done some very extensive testing of M1 A123 cells the past few months and the IR can vary from about 7mOhm to over 60mOhm, depending on the conditions.

For room temperature, >30% SOC, 5A or less charge/discharge rate, I get about 9.0mOhm to 10.7mOhm for a each cell using 12AWG equivalent cell-interconnections in a pack.

CamLight said:

katou said:

What should I be getting as a measurement of IR for a good A123 cell?
Katou

Click to expand...

At what state-of-charge?
At what temperature?
At what charge or discharge current rate?

All these have a huge effect on the IR. All the values are correct, but just for those conditions. I've done some very extensive testing of M1 A123 cells the past few months and the IR can vary from about 7mOhm to over 60mOhm, depending on the conditions.

For room temperature, >30% SOC, 5A or less charge/discharge rate, I get about 9.0mOhm to 10.7mOhm for a each cell using 12AWG equivalent cell-interconnections in a pack.

Click to expand...

70degF at nominal voltage is the industry standard to measure Ri AFAIK. At least that's where I take my measurements. Get a cell on it's last 5-10% capacity, and the Ri often goes up 10x or so.

Right, of course, V=IR, so it would vary as the other two.

I wish that the CBA would plot the IR too. That tester that you got Luke, it looks like quite the device. Sorta making me reconsider... Anyway, that's what I have for the moment.

In fact, isn't there a way to derive the IR from the graph of ah/volt that the CBA does put out? I'm not too electrically smart, but I have this feeling that it could be done by calculating instead of measuring.

I appreciate the benchmarks, I hope to report back on this in a week or so.

Any chance you all have the same figures for Moli cells and Konions?

Katou

Why? What sort of results did you get when you tested my cells?

If it were to stop and start the discharge process, and measure the current level and the voltage difference, then Ri could be calculated (some battery chargers try to do this, all I've seen so far really suck at it for low Ri cells, but do a decent job for high Ri cells). For continuous discharge, Ri can't be accurately calculated because the stage of charge is decreasing voltage while the current draw is causing voltage drop at the same time.

Har, har har. Your cells were fine, it's new ones I'm having trouble with.

Katou

katou said:

Har, har har. Your cells were fine, it's new ones I'm having trouble with.

Katou

Click to expand...

Coolio. Best of luck with them

I believe some industrial systems superimpose an AC current waveform on top of the (DC) discharge current and LPF the sensed Voltage in order to measure Ri throughout a full charge or discharge cycle.

liveforphysics said:

If it were to stop and start the discharge process, and measure the current level and the voltage difference, then Ri could be calculated (some battery chargers try to do this, all I've seen so far really suck at it for low Ri cells, but do a decent job for high Ri cells). For continuous discharge, Ri can't be accurately calculated because the stage of charge is decreasing voltage while the current draw is causing voltage drop at the same time.

Click to expand...

If I understood half of what you just said....

Sounds cool though. I'm a fan of automatic testing, all the decimals get carried properly, I just come back later and it's done.

Now, if someone will just make one that tests (and graphs) 10 at a time.

Katou

katou said:

If I understood half of what you just said....

Sounds cool though. I'm a fan of automatic testing, all the decimals get carried properly, I just come back later and it's done.

Now, if someone will just make one that tests (and graphs) 10 at a time.

Katou

Click to expand...

http://www.maccor.com/formationequipment.php

:wink:

Hi Luke,

liveforphysics said:

Good LiPo was around 2-5mOhm per 5Ah cell, but the new Turnigy Nano's just blew my mind when I tested <1mOhm Ri at 25C discharge rate (100amps on a 4Ah cell).

Click to expand...

How about a thread detailing the results with the Turnigy Nano's (hopefully including a destruction test video )? You know you want to do it .

Geez I hate it when vendors tell you everything about the product but the price!

http://www.maccor.com/formationequipment.php

How can I buy one if I don't even know how much it costs? If they're not too much, maybe I could buy two?

Katou

In general, if you have to ask how much something costs, it costs too much to afford unless you're a business that does work requiring such equipment (and even then, a lot of the time it still costs too much).

When there is no price listed they expect you to call their sales reps, then tell them about your company and business, how much money passes thru your hands each quarter, and then they figure out how much they can rake off the top of that and tell you that as the price. Or they laugh and hang up on you.

katou said:

Geez I hate it when vendors tell you everything about the product but the price!

http://www.maccor.com/formationequipment.php

How can I buy one if I don't even know how much it costs? If they're not too much, maybe I could buy two?

Katou

Click to expand...

I'm guessing $20K-$70K and up.
Even their smaller units cost $6,000.

I came across this discussion today about IR measurements and thought you might like to read it Katou.... at post #s 13, 14 & 15. hope it helps some :wink:

http://www.rcgroups.com/forums/showthread.php?t=1280537&page=2

27k huh?

Hmm, not *entirely* sure I can justify that one. Maybe I'll just stick with the CBA...

It probably measures IR pretty good I bet.

Can anyone recommend some good threads on testing cells for IR?

Katou

edit: thanks Scoot, I just saw your message. I will check it out.

I just got in 8 of the 2.3AH A123 cells from Hobby King (new, with clean 0.010" nickel tabs. Arrived after 8 days, scheduled was 4-6 weeks, box smashed to Chinese standards). They supposedly were charged to 50% SOC (whenever they escaped the A123 secret dungeons). All measured 3.30050V +/- 0.00010V unloaded!

Next, I did a DC Ri test by loading them with a 5.1115 ohm resistor+leads and measuring the voltage drop (0.00635V). After the load they measured 3.30035V. The Ri came out to be 9.8554 milliohms. Not too shabby...

I need to charge them up and see what the Ri is at known full charge... then I need to unbalance them and see what the switched capacitor balancer does...

All measurements were done with an HP3457A 7.5 digit multimeter calibrated to MINT (Mark's Institute of Non-standards and Technology) standards (which are arguably better than 90% of the world's national standards laboratories)

I think that I will build up a multi-channel cell tester using the 3457A, MegaDonkey controller interfaced to the 3457A GPIB interface, and some load resistors switched by relays. Might even cost less than $27,000 :lol:

27k huh?
Hmm, not *entirely* sure I can justify that one. Maybe I'll just stick with the CBA...

Click to expand...

katou, unfortunately, the CBA doesn't do well when testing for IR.

It's slow response time requires you to have current flowing for a while before it correctly reads the values and this will adversely affect your IR readings. You need to pulse the current for a very, very short time to get only one voltage value for the current being drawn.

And, even its pulsed-current mode (Pro model only?) turns the current on and full-off and doesn't allow for bouncing between two current values. Determining IR should be done using two different discharge current values to prevent interference from cell recovery effects, surface charge, etc. And always test a bit down (using a known discharge current and time) from a full charge too. A fully-charged cell gives hugely inaccurate IR testing results as the first few seconds of the discharge are completely different from the rest of the discharge curve (as can be seen on a discharge curve).

But, if only comparing your results against your own other tests, the slow response time and other issues might not matter. You don't need accurate results in this case, only very repeatable results. And the CBA is awesome for that. You won't be able to compare your results to anyone else's but it will help to sort out bad cells.