Running a laptop off eneloop AA's

[philbx1]

BTW, I pulled apart a Dell battery last week and it had 6 Li-Ion's which looked VERY similar to these (without the text)
HK 18650's

OOPS - Actually, it may have been 9 of these. but at 9 x 3.7 = 33.3 it seems a lot of volts.

[Simon]

I like the look of the nano battery but not the price  and it ships from america i guess ? so best not go there, I'll have to keep an eye on battery developments. will there be the day that Li-ion batteries make it into mainstream use ? or are they just to erm.... volatile

[Simon]

were they directly in parallel or was there some control circuitry to isolate them ?

[IanB]

were they directly in parallel or was there some control circuitry to isolate them ?

It's completely OK to put lithium ion cells in parallel in a battery pack. For example if you put 3 x 18650 in parallel you would get something that behaves like a single cell with three times the capacity. It will charge and discharge just like one big cell.

However, you cannot do the same with NiMH like eneloops. Attempting to charge NiMH cells in parallel is a big no-no. Damaged batteries will result unless you design a really cunning charger.

[bfritz]

were they directly in parallel or was there some control circuitry to isolate them ?

They are mostly always charged as bank 1 bank 2 and bank 3 .

Every Li technology battery I have seen in my career that uses batteries in parallel, connects them directly in parallel.

Regarding circuitry to isolate them, I have never seen a battery pack that uses circuitry to isolate cells in parallel.  It is common to see areas in the conductive straps between the cells, where the strap was been made narrower.  The idea is that if a high current above some level starts to flow, the necked area will heat, and the strap will fail open, stopping the high current.

Most Li technology manufacturers will furnish a data sheet on the cells.  You usually have to ask for it, and it is almost always provided only under an NDA.  It will give details on the maximum number of cells that can be placed in parallel.  It is quite common to see two 18650 cells in parallel, and fairly common to see three in parallel.  I have not seen a manufacturer whose datasheet says configurations of more than four in parallel are allowed.

[Simon]

I guess that if more than 4 were put in parallel and they started failing you would end up in a position where 1 cell could be required to deliver 4+ times the intended current

[bfritz]

Having extra discharge current for one cell is not that big of an issue for Li, but having a charge rate above what they are designed for, can cause internal plating of metallic lithium... and that is very bad.  It clogs the graphite matrix, and usually results in shortening of the life, and can even lead to shorting across the seperator, and fires.

But, don't kid yourself that Li technology is the only cell with safety issues.  I have see NiMH and NiCd packs catch on fire and cause major damage.  I saw the entire passenger side of an auto burned out because a portable computer had been set on the seat, and charged from an inverter connected to the cigarette lighter socket, with the inverter connected to the AC adapter.

The majority of safety issues I have seen, mostly stem from construction issues between the cells.  The welded straps between the cells need to be insulated with abrasion resistant insulators anywhere that shorting could cause a problem.  Failing to do this can lead to shorts developing over years as the packs are subjected to vibration.  Such shorts can allow over 100A to flow.  It doesn't take long with a 100A short to cause a fire.  This can happen as easily with NiCd or NiMh as with Li technologies.

[bfritz]

Thankfully NiCd's have a fuse where if you short out the battery it just blows .

That fuse is external to the cells, and only limits current being drawn outside the battery.  Any battery technology can be wired with such a fuse.

Such a fuse does nothing to stop internal shorts within or between the cells, which is the typical problem seen in manufacture of packs.  This is often caused by inadequate insulation not designed for the abrasion that occurs over years of use.

[amyk]

Having extra discharge current for one cell is not that big of an issue for Li, but having a charge rate above what they are designed for, can cause internal plating of metallic lithium... and that is very bad.  It clogs the graphite matrix, and usually results in shortening of the life, and can even lead to shorting across the seperator, and fires.

The other thing is the voltage limit, 4.20V maximum (+/- 1%) for standard lithium-cobalt (3.6V for phosphate). Some cheap devices I've seen can't keep that voltage tolerance with their simple circuit and just go for a constant-voltage 4V charge which probably reduces the usable capacity (but prolongs the life a bit).

[bfritz]

Thankfully NiCd's have a fuse where if you short out the battery it just blows .

That fuse is external to the cells, and only limits current being drawn outside the battery.  Any battery technology can be wired with such a fuse.

Such a fuse does nothing to stop internal shorts within or between the cells, which is the typical problem seen in manufacture of packs.  This is often caused by inadequate insulation not designed for the abrasion that occurs over years of use.

An internal fuse , once you short it , it becomes useless .
But if it does short internally , this internal fuse cuts all contact , it's in the middle of all cells (inside the single cell alone)

I've never seen a NiCd or NiMh with an internal fuse.  I have seen some Li-Ion cells with PTC devices internal, but I don't consider that a fuse, it is a current limiting device.  I have designed a lot of packs, and looked at lots of cell data sheets, so I believe this is the exception rather than the rule.  Please point me at any datasheet that shows such a fuse.  I'm not asking because I don't believe, I'm asking because I haven't seen it.

[bfritz]

The other thing is the voltage limit, 4.20V maximum (+/- 1%) for standard lithium-cobalt (3.6V for phosphate). Some cheap devices I've seen can't keep that voltage tolerance with their simple circuit and just go for a constant-voltage 4V charge which probably reduces the usable capacity (but prolongs the life a bit).

That is a perfectly acceptable practice.  You give up some capacity per cycle, but end up getting a longer cycle life.

It is important to remember that even with this practice, it is still necessary to monitor the individual cell voltages of the series connected cells for safety purposes.  It is possible for the cells to be inbalanced, and one cell to become much higher in voltage during charge, as it has the least capacity.  If that cell goes overvoltage while the others are still below the overvoltage, it is possible to cause a safety event with the cell in overvoltage.

[G7PSK]

I have been following this thread with interest as I have often wondered about the pros and cons of various cell types and configurations. From what have read Lithium cells should not be run flat, The Panasonic toughbook's or at least the CF28 has a utility within the bios for refreshing the battery, which should be run about once a month, this discharges the battery pack and then recharges it, I have noticed that these units give very good battery life which I assume is partly due to this utility. Why don't other makes have something similar. Simon, on ebay you can get a power unit for laptops that runs on all voltages including a 12 volt input you could use one of these with a deep discharge lead acid battery.

[sonicj]

i run my macbook pro off of a 12V deep cycle lead acid during thunderstorms. pulls around 2.8 amps surfin' the net.
-sj

[IanB]

I have been following this thread with interest as I have often wondered about the pros and cons of various cell types and configurations.

Lithium ion cells of the regular kind used in laptops have more energy per weight and per volume than nickel metal hydride cells. In power tools too the lithium battery packs have more energy (longer run time) than the cheaper nickel cadmium packs.

From what have read Lithium cells should not be run flat,

It is OK to run them flat, where flat is defined as the point where no more useful energy is available. For a typical lithium ion cell this would be a resting cell voltage somewhere around 3.2 V, give or take. However, it is harmful to take the voltage much below this level and it may cause irreparable damage to the cell or even create a dangerous condition on recharging. Lithium ion cells and batteries always come with a battery protection circuit that cuts off the battery if the voltage gets too low.

The Panasonic toughbook's or at least the CF28 has a utility within the bios for refreshing the battery, which should be run about once a month, this discharges the battery pack and then recharges it,

There is no advantage to discharging and recharging a lithium ion battery. The battery comes with a useful working life, and cycling the battery unnecessarily just drains a few more grains of sand through the life timer and advances the battery further towards the end of its life. There can, however, be value in doing this to re-calibrate the battery management computer so it has an accurate measure of the capacity of the battery.

I have noticed that these units give very good battery life which I assume is partly due to this utility. Why don't other makes have something similar. Simon, on ebay you can get a power unit for laptops that runs on all voltages including a 12 volt input you could use one of these with a deep discharge lead acid battery.

The best way to achieve good battery life with lithium ion is to keep the battery voltage away from the extremes. Charging to 100% and discharging all the way to 0% will shorten the life compared to, say, charging to 85% and discharging to 15%. Laptops with good battery management will avoid topping up the battery unless it falls below about 90%, and they may even have a battery conservation mode where they will avoid charging the battery to 100% ever. Charging a lithium ion battery to 100% (a cell voltage of 4.20 V) and keeping it there constantly is the best way to shorten its life.

[bfritz]

I've never seen a NiCd or NiMh with an internal fuse.  I have seen some Li-Ion cells with PTC devices internal, but I don't consider that a fuse, it is a current limiting device.  I have designed a lot of packs, and looked at lots of cell data sheets, so I believe this is the exception rather than the rule.  Please point me at any datasheet that shows such a fuse.  I'm not asking because I don't believe, I'm asking because I haven't seen it.

http://homepage.ntlworld.com/wilf.james/nicads.htm
Fair 'nuff .

Dave, is the link above your answer to a NiCd or NiMH with internal fuse?

That posting does not mention a fuse within a cell.  That posting mentions that most NiCd's have a fuse within the pack.  This is very different from having a fuse within the cell.  A NiCd cell is cylindrical in shape.  Multiple cells are wired in series to form a pack.  Usually either the positive or negative lead that exits the pack, has a fuse placed in series.  Note that internally it is possible for the metal straps between the cells to short, even though external shorts are protected by a fuse.

See my Beautiful drawing below.   

Note that the short shown in red, will not trip the fuse, as the short current does not flow through the fuse.  This is why pack fuses do not provide protection against shorts within a pack, only those external to the pack.

[bfritz]

NiCd , even though packs have an internal fuse , it does not mean most cells doesn't have , i actually cut a busted NiCd open to find a HRC fuse inside on the positive side .

So, you posted that link to show that NiCd's have internal fuses... Yet the link never talks about internally fused cells.  And your explanation is that "...it does not mean most cell doesn't have..."

In my over 25 years of experience designing battery packs, and battery pack electronics, I have never seen a Ni pack with an internal fuse.  Most cells do not have fuses.  I was willing to believe some could, so I asked for some evidence, so you send me on a boondoggle, to an article that doesn't back up anything that was in question.

I'm sure you believe you saw a cell with one, have zero knowledge other than what you believe you saw, and were probably mistaken in your analysis.  It is possible to make a cell with an internal fuse, but doing this would have very little benefit to the cell manufacturer, as they would then have to stock different cells with different fuse ratings.  I just can't see a good reason to do it, can see lots of downside, and you offer no real evidence other than "I saw it".

Can you understand why your argument isn't very convincing?

By the way, I have quick access to cell data sheets from every major cell manufacturer.  If you give me a cell manufacturer, and model number, I can look this up quickly.  If most cells have this, as you say, then giving me a major manufacturer's model number for one, should not be difficult.

[amyk]

The other thing is the voltage limit, 4.20V maximum (+/- 1%) for standard lithium-cobalt (3.6V for phosphate). Some cheap devices I've seen can't keep that voltage tolerance with their simple circuit and just go for a constant-voltage 4V charge which probably reduces the usable capacity (but prolongs the life a bit).

That is a perfectly acceptable practice.  You give up some capacity per cycle, but end up getting a longer cycle life.

It is important to remember that even with this practice, it is still necessary to monitor the individual cell voltages of the series connected cells for safety purposes.  It is possible for the cells to be inbalanced, and one cell to become much higher in voltage during charge, as it has the least capacity.  If that cell goes overvoltage while the others are still below the overvoltage, it is possible to cause a safety event with the cell in overvoltage.

I've only seen that simple circuit using a single-cell low-capacity lipo (~150mAh). It was in a cheap MP4 player and just used a current-limiting resistor and voltage divider to provide around 4V at 50mA from the USB 5V line.

I've seen in the datasheets for a lot of battery charger ICs that they have "preconditioning" for recharging "severely depleted" cells but that confuses me a bit - is it safe to do that, because I've read that lion cells should not be discharged below ~2.75V, but apparently these chargers claim to be able to charge them properly with a very low current precharge?

Regarding powering a laptop from its DC input, all the ones I've seen have a buck converter to generate the system 12/5/3.3/2.5/1.8V rails (actually a lot of the more recent ones don't have a 12V rail) from the adapter or battery, and it usually accepts a rather wide range (e.g. 3-28V for a TPS51124), so anything >12V is usually OK - just make sure to take into account any fuses in the path and the higher current that may be drawn with a lower input voltage. For example my laptop has a 19V adapter and that's the voltage the MAX1977 buck converter gets when it's plugged in, but on battery the same rail only measures ~11V.

[nukie]

What is NiCd designed for? A lot of current. Some cells can dump 40A at a time. What's a 40A fuse going to look like? You can look it up in your car. I've never heard of a NiCd with a fuse. Some people actually short the terminals to vaporize those dendrites/whiskers. If there's fuse it would be on the device.

[bfritz]

I've seen in the datasheets for a lot of battery charger ICs that they have "preconditioning" for recharging "severely depleted" cells but that confuses me a bit - is it safe to do that, because I've read that lion cells should not be discharged below ~2.75V, but apparently these chargers claim to be able to charge them properly with a very low current precharge?

The preconditioning is safe.  Here is the way it works.  Most packs have a protection circuit that disables discharge at some undervoltage threshold, usually when any cell reaches 3V.  The electronics in the pack consume a very small current, usually in the 5-10 uA range.  This small current, along with the small self discharge present in Li chemistries could lead to a pack going to a fairly low voltage if allowed to sit uncharged for an extended time.  As long as the cell doesn't go below a safety threshold, usually around 2.5V, then charging to recover the cell is allowed, but typically must be done at a C/10 rate or less.  These preconditioning charge circuits apply this lower rate charge, and check that the cells comes above the approximately 3V undervoltage threshold within a reasonable time.  If it doesn't achieve the undervoltage threshold in a reasonable time, the circutit will typically blow a fuse that disables the pack for safety reasons.

Regarding powering a laptop from its DC input, all the ones I've seen have a buck converter to generate the system 12/5/3.3/2.5/1.8V rails (actually a lot of the more recent ones don't have a 12V rail) from the adapter or battery, and it usually accepts a rather wide range (e.g. 3-28V for a TPS51124), so anything >12V is usually OK - just make sure to take into account any fuses in the path and the higher current that may be drawn with a lower input voltage. For example my laptop has a 19V adapter and that's the voltage the MAX1977 buck converter gets when it's plugged in, but on battery the same rail only measures ~11V.

In general I see one of two pack topologies for most laptops.  The most common involves series connections of 3 Li cells, for a maximum pack voltage of 12.6V.  The other involves series connection of 4 Li cells, for a maximum pack voltage of 16.8V.

Most systems identify the AC adapter by using a comparator in either a charge IC, or other circuitry, that looks for an input on the adapter input that is above the battery voltage.  If you have a 4 cell battery and plug in a 15V adapter (This is common for a 3 series config) if the battery is fully discharged, it may start charge as the pack is only at 12V (3V/cell * 4 cells).  As the pack starts charging it will run out of headroom for the internal charger as a 4 cell charges to 16.8, but the adapter only is supplying 15V.  This will cause charge to terminate early.  This can lead to problems in the fuel gauging circuitry in the laptop pack, as it may decide charge was terminated at full, and update the fuel gauge info incorrectly.  It learns a new capacity, but the charge termination was incorrect.

So, just choose an input voltage which is about a volt or higher than your highest battery pack voltage, and all will likely work well. 

[I Shout Loud]

hi Simon, Did you succeed with Eneloop? I am also interested in this topic. I came across this crude method of replacing batteries at youtube. But,  try http://www.duracelldirect.com. They have batteries for all equipments and it is from the manufacturer directly. All the best. 

[Thor-Arne]

Just have to show you guys this in case you missed it: http://www.gooddealchina.com/Batteries/50000mAh-Portable-Power-Station-External-Laptop-Lithium-Battery-c22224.html

A proper test of this one would be nice.