Recharging batteries ... in series.

I need a fresh look at this. I've spent countless hours and written several e-mails back and forth with Linear Tech about this and so far I haven't come up with a working solution. So maybe someone here can ...

I need to put two, single cell LiPo (or Li-Ion) batteries in series for the higher voltage. Each one will need to have its own charging circuit as well. Where I am stuck is, how do you tie that together in series? I can't simply tie one battery's negative pole to the next one's positive pole because when the charging circuit kicks in, current will flow all over the place between them. It just won't work. And on top of that, when the batteries are depleted and a DC Input is provided to recharge, that too needs to supply voltage to the system load, so I need to switch between the batteries and DC input.

I'm going to use the LTC3633 to generate the voltages and current needed for the system load. What I need now is to figure out how do I tie the batteries together so that their charging circuits don't screw with each other, and a way to switch between the batteries and DC input (when provided.) Again, Linear has some muxing ICs that can do the switching and also monitor the batteries to prevent an over-discharge, but they are all between DC, BATT1 and BATT2 ... and I need it between DC and both batteries in series, not a parallel setup.

Have you looked at the www.maxim-ic.com website? They have multicell battery charge control chips that handle situations like that.I think microchip.com does as well, for example:

"The MCP73842 combines high accuracy constant-voltage (0.5% max), constant-current regulation, cell preconditioning, cell temperature monitoring, advanced safety timers, automatic charge termination, and charge status indication in a space saving MSOP-10 package. The device provides a complete, fully functional, stand-alone charge management solution for applications utilizing dual series cell Lithium-Ion or Lithium-Polymer battery packs. Two preset voltage regulation options are available, 8.2V and 8.4V, for use with either coke or graphite anodes. The MCP73842 operates with an input voltage range of 8.7V to 12V and is fully specified over the ambient temperature range of -40°C to +85°C."

Yeah, that 73842 is for a single cell. I can find similar ones from Maxim and Linear. That's not the issue here. The issue is taking the two battery outputs and tying them in series in such a way that the charging circuit doesn't flow between both.

If I put two batteries in series, the charging circuit for the first battery will also flow into the second battery, no?

KirAsh4:
I can't simply tie one battery's negative pole to the next one's positive pole because when the charging circuit kicks in, current will flow all over the place between them. It just won't work.

That's not really true. When was the last time you charged a car battery (which has 6 cells) with a charger with 7 battery clamps?

Assuming that the battery is manufactured properly the cells should have an equal resistance between them and should charge evenly. When cells do NOT have an equal resistance or state of charge then cells with a lower voltage should charge more quickly than a cell with a higher voltage. Everything should stay in equilibrium -- although it may require longer charge times / trickle charging to do so.

Yes I do understand that it is typical with lithium polymer batteries to control the charge in each cell individually. It's just a matter the economics of the situation where a more complex / more expensive charger makes sense.

"The MCP73842 and MCP73844 are designed for
applications utilizing dual series cell Lithium-Ion or
Lithium-Polymer battery packs. Two preset voltage
regulation options are available (8.2V and 8.4V). The
MCP73842 and MCP73844 operate with an input
voltage range of 8.7V to 12V."

Charger uses thermal monitoring to keep it all safe.

You don't have the higher voltage availanble for the charging current?
Otherwise, need to implement some MOSFETs for voltage isolation & switching.

CR: the way it's planned right now is to have two batteries in series to go from 3.7V to 7.4V with the same Amp output. Take that voltage and pass it through a regulator that's going to give me a 5V and 3V3 rail @ 2A each (this from Linear). When the batteries need to be recharged, there will be a 12V DC input that will both recharge the batteries, as well as feeding the regulator (so the device continues to run.)

Chagrin: I will never assume that two batteries are identical. Car batteries aren't and will die over time too. By using single cells, at least I can monitor each one separately and keep them close to one another.

Any particular reason you want to use Lithium Ions .
Li batteries are good , but as you have discovered introduce a whole lot of issues when it comes to charging them.
Can you use NIMH types instead.
Far fewer charging issues.

No two batteries are the same, it doesn't matter what chemistry. While NIMH may be less of a headache than using LiPo or Li-Ion, the same rules still apply as far as keeping them balanced. Over time, they too will become unbalanced if you charge them in series, as will any pack that doesn't have a proper charging circuit to monitor each cell. The question becomes: how long will they last comparatively ...

Nimh batteries dont need balancing as NIMH chemistry can withstand overcharging if the charge rate is C/20 or less.
The battery just gets warm and dissipates the heat.

LIPOs cant be overcharged at all, so the complex charging regime needed.

KirAsh4:
No two batteries are the same, it doesn't matter what chemistry. While NIMH may be less of a headache than using LiPo or Li-Ion, the same rules still apply as far as keeping them balanced. Over time, they too will become unbalanced if you charge them in series, as will any pack that doesn't have a proper charging circuit to monitor each cell. The question becomes: how long will they last comparatively ...

I'm not at all convinced that is the case. Firstly, assuming they start out in a similar state of charge, why would they diverge? One reason would be that the capacities differ. In that case they might start out in similar states but cells would go flat at different rate; as they were recharged, they would converge again. That's no problem and is simply a fact of life, even with matched cells.

Another reason is that different cells will have different leakage currents and may spontaneously lose different amounts of charge over time. It seems to me that for most battery types, this effect would be completely corrected simply by trickle charging the batteries until all cells were fully charged, which is normal good practice on the types of rechargeable battery I've used (NiCd, NiMh). It may not be healthy to leave them cooking for days, but a standard rule of thumb is to trickle charge for sixteen hours at the 'twelve hour' rate so each cell is being given an excess charge anyway. This is completely harmless and I would expect it to ensure that every cell was left fully charged, even if they started out in slightly different states of charge.

The idea that all cells connected in series will inherently drift into differing charge states simply does not make sense.

The idea that all cells connected in series will inherently drift into differing charge states simply does not make sense.

Li-Po battery pack manufactures have a specific process they use to test and select and match individual single cells that will be wired in series/parallel combinations for the battery packs they sell. That is to make sure the the cells do have as close as possible discharge and charging characteristics rates as possible to fight the real problems that can happen if and when cells become 'unbalanced' from each other. They never recommend that an end user wire up individual Li-Po cells to obtain higher pack voltages/current capacity. If one does decide to build up their own packs from individual cells it makes it all the more important that they use a true 'balancing charger' which can help deal with cell to cell variations when charging. One only has to look at the high number of very sophisticated and powerful 'balancing charges' that are sold to the R/C hobby world to understand that Li-Po pack cell balancing is an important issue.

Lefty

retrolefty:
Li-Po battery pack manufactures have a specific process
...

I absolutely understand that's the case for LiPos, and how that has led to the need for each cell to be charged and managed individually.

My comment was in response to an earlier post suggesting that this behaviour was something that inherently applied to all types of rechargeable battery. It does apply to LiPos (and there may be other exotic battery types that it applies to as well) but there are many commonly used rechargeable batteries that it does not apply to.

PeterH:

retrolefty:
Li-Po battery pack manufactures have a specific process
...

I absolutely understand that's the case for LiPos, and how that has led to the need for each cell to be charged and managed individually.

My comment was in response to an earlier post suggesting that this behaviour was something that inherently applied to all types of rechargeable battery. It does apply to LiPos (and there may be other exotic battery types that it applies to as well) but there are many commonly used rechargeable batteries that it does not apply to.

Well the basic failure mode can effect an series wired rechargeable cells in that for instance one wires together some random cells that have different state of charge so that they start out in a unbalanced condition. Such a pack has a high chance of damaging cells during both discharge (causing cell reversal in some) and charging process (causing overcharging of some) from the get go. Even back in the days when hi current sub-C and C Nicad cells were the flight pack of choice in the R/C world, the competitive people would go to lots of trouble 'matching' their individual cells into packs. Of course they were really pushing the cells such that they were discharging them completely in a 10 min high current run, then letting them cool for only a short time until charging them up at a 10C or greater rate. Needless to say they tended to not get long cell life out of such usage abuse.

Lefty

It does apply to LiPos (and there may be other exotic battery types that it applies to as well) but there are many commonly used rechargeable batteries that it does not apply to.

From my experience, I don't quite agree with this. I use NiMH batteries exclusively, and
there is also a problem if the cells in a series pack are not closely-matched. I'm sure that
most RC guys who sell the pre-made packs match NiMH batteries beforehand.

http://www.promatchracing.com/whymatched.php

For my part, I simply buy packs of 4 NiMH batteries at a time at the department store,
namely standard brands like Kodak, Everyready, etc, and use them together, and it's very
common that one of these cells will start misbehaving after a while, and take much longer
to recharge in an individual charger than the others.

It's also happens that, if one cells in the series arrangement isn't quite up to par with the
others, then during driving loads, when the current is going through the cells in the opposite
direction to charging, the good cells can overwhelm the weak cell and discharge it, and
actually drive its cell voltage to zero, or even in reverse. I'm not alone in seeing this, it's
common,

See also,

Over-discharging
A complete discharge of a cell until it goes into polarity reversal can cause permanent damage to the cell. This situation can occur in the common arrangement of four AA cells in series in a digital camera, where one will be completely discharged before the others due to small differences in capacity among the cells. When this happens, the good cells will start to drive the discharged cell in reverse,

The weaker the cell, the faster this happens. 5 or 6 years ago, I build a jig that measured
individual cell voltages during load-driving, and could easily watch the weaker cells going flat.