Li Battery

Nothing. It's training the battery charging circuit to recognize the particulars of that battery,

Could be quite a bit because the charger may shut down before completely charging the battery.

Den torsdag den 12. november 2015 kl. 18.03.35 UTC+1 skrev OldGuy:

afaik you shouldn't do that for Li-Ion, it was the old NiCd that needed that

-Lasse

That's what I found out too. You shouldn't run a Li-Ion down completely. The Li battery in my DSLR has been going strong for the last 8 years. The camera says it's in good shape.

They can be run down but there is a limit. If it's never run down, how do you know its capacity? Laptop charging circuits have gas gauges and need to run down the battery to calibrate.

If the battery sits on the shelf, then that self discharge does not show up on the gas guage. Plus the LI batteries I've used tend to increase with capacity in the first few cycles.

Cheers

Everything I've read so far says never fully discharge lithium.

But I think the gas gauge could be the reason I've never had to buy cells for my DIY ecig.

All my cells come from recycling bins in various shops, I've gone on the assumption that one weak cell causes the whole battery pack to be scrapped - but so far, I've never found that one weak cell.

Presumably a chip in the battery pack informs the user that their battery is worn out, and they throw it in recycling - whether it is or not.

That depends on what you mean by "fully", I suppose. Discharge to zero? No, that's not a good idea (but not as disastrous as SLACs). However, laptops won't allow this anyway. There isn't much energy under 3.1V (from memory) so there is no point in discharging below that point.

I suppose that depends on what you define as a good cell, too. I've had batteries go bad. I know they're bad because they discharge significantly faster than a new cell.

Charge/discharge efficiency is not 100%, nor is leakage perfectly matched in the set.

So, over time, some cells get lower than others. Without any cause at all, as far as the controller knows -- no terminal current was even flowing.

By recharging a flat pack, the most-charged cells go up from, say, 40% to

90%, while the least-charged cells go up from, say, 20% to 80%. There's less charge in the already-charged ones, and more charge in the less-charged ones.

Same thing happens on discharge, it's a decay process internally. Not perfectly divided, coulomb per coulomb, among cells.

So having

Back in the day, I built an automated test fixture to evaluate laptop cells. Ignoring the obvious exceptions of shorted cells and those that had been obviously overheated, I found that virtually every cell from a "bad" laptop battery pack would produce it's full rated number of electrons if discharged at low current.

The failure mode is increased internal series resistance. Battery management chips deal with that in different ways, but the result is pretty much the same...buy a new battery.

Do the math. Using some round numbers, if your laptop takes 60W and has a 10V battery, that's 6 amps AVERAGE. The difference between full and empty for a lithium cell is about a volt. 1/6 of an ohm series resistance per cell or cell pair renders your pack useless for powering a laptop. And the peak current may be much higher than that.

At far less resistance, you see the symptom that the battery gauge seems to be normal down to 50%, then instantly drops to zero, but the laptop may run for a long time at that level, if it's not working too hard...depending on how the battery management system is programmed. Even if it's a coulomb counter, the voltage drop caused by the peak current demand can trip the voltage level safety limit.

The ecig I use recovered cells on is just a little heating element - a small coil of resistance wire.

Commercial ones are from about 1.6 - 2.8 Ohms, the coils I rewind are somewhere around the 1 Ohm mark.

Maybe less than peak demand in a laptop - but not trivial either.

Laptop battery packs have a thin strip PCB with various chips etc on it, AFAIK: they use shunt regulation as a halter for any cells getting close to critical terminal voltage.

This must have a voltage sensing divider, I'm sure they do everything possible to minimise sensing current - but there must be some.

I guess the battery is imitation, made in China, which explains how they were able to add some instructions meant for NiCd batteries.

joe

You obviously have no - or little - experience with LiXX cells and their care and feeding. What krw said first up is correct - it is part of the device's capacity calibration process. It is very common advice with many appliances from reputable manufacturers. RTFM.

Also, Martin's observation: "LI batteries I've used tend to increase with capacity in the first few cycles." is what I have experienced in our testing of virgin cells. (I have designed a commercial LiXX charger, which project involved significant testing/cycling).

You always have to differentiate between the capabilities and behavior of the battery, and the behavior of it's protection circuitry.

When the battery protection circuitry is present, you're hostage to the firmware present and you've really no alternative but to do what the mfr tells you to do, right or wrong.

Sometimes the battery and it's control hardware/firmware are poorly matched (as is the 'gauge' monitoring of Li Ion cells) resulting in early death of the combination. This is not through fiery explosion, just faulty capacity reporting or poorly managed and ineptly deployed 'End Of Discharge' or 'End Of Charge' events.

There's often a disconnect between the battery protection firmware and that of the equipment in which it is used, even in dedicated single-product applications.

You might get useful help from other users of the same equipment, who've been deeling with it longer.

RL

RL

All I've seen are protection only: disconnect if the charge goes too high, or the discharge goes too low. And logging (charge reading chip) and info (EEPROM with mfg/sn#/key/..).

They certainly aren't large enough to dissipate excess charging power per cell!

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
Website: http://seventransistorlabs.com

Good, I stand corrected, thanks. What exactly is being 'calibrated'?

joe

Capacity.

The host device - in this case a laptop - provides both a readout of "remaining battery capacity" as a percentage (of full capacity) and provides low battery warnings and shutdown actions. To know what the capacity is, it needs to experience the limits and store the values.

At the end of full charge, it notes that as 100%. But where is 0% (of USABLE capacity)? That answer is found by full discharge, down to the pre-ordained (in the host) cutoff voltage - which is hopefully above the battery protection module's shutoff point.

Now it knows what the usable capacity is, and can quote numbers to the user, and make operational decisions based on those numbers. The only remaining kink in the scheme of things is that the capacity will increase over the first two or three cycles. Hence the recommendation that the battery be cycled fully a few times before use - as the user is unlikely to do this in service and therefore pre-emptive warnings and actions may be misplaced.