Reviving Thinkpad batteries

[Ash]

I have a few laptops with dead batteries, including Thinkpads X60s, T60 (both 6 cell and 9 cell batteries), X40, X31 and Dell and Compaq too

The Thinkpad battreies are dea and locked out - There is 0v on outside terminals, but few volts on the cells inside. The Dell and Compaq are not locked out but simply power off couple seconds after charger unplugging from laptop

I'd like to buy just the Li Ion cells and replace them inside the batteries

1. Are there any special parameters in the cells i should get, other than Li Ion/voltage/capacity/max discharge rate (based on a guess from computer running time at max load)/dimensions ?

2. How do i unlock the Thinkpad battery controllers after replacing the cells, and if it does not happen automatically, how i reset them to display correct battery level to the OS. Have anyone experience ?

[hellosailor]

Ash-
There are several different "lithium" chemistries and each has slightly different voltage requirements, so you might want to carefully match whatever was in the original packs. But if you get the capacities mismatched, or there's a problem in the charging rates, or you solder any connections and cause overheating in the batteries, lithium cells don't just fail, they tend to catch fire or explode. Here in the US it is often possible to catch a no-name battery pack, or a genuine Lenovo (etc) pack on clearance for 1/4 of this list price. I'd suggest just taking a hard look around and letting some vendor build the pack for you. Computer battery packs, especially the lithium ones, aren't "just" battery packs anymore. (Forgive me if you already have been inside them and know that.) Among other things there are literally full minicomputers integrated as battery management systems, interconnected to every cell in the pack.

--Red

[Ash]

I am familiar with the precautions, im asking for actual parameters and controller unlocking possibility, thanks for the warnings though

[joester]

I've rebuilt a few packs with no fires, but you need to fully understand that should one blow on you, it needs to burn itself out and it's like a blow torch. I happened to have a few metal boxes constructed with very heavy gauge steel that are about the same size as a large toolbox (any larger and you wouldn't want to lug it around the junkyard) that I put my battery packs in for the de-soldering / soldering process. I figured if a fire were to erupt, I only need to pull back and close the lid to contain the fire. I even had a box fan in the window and running to deal with the smoke.

As far as unlocking the boards, I cannot help you. I was successful in rebuilding a T20 pack that was good for 4 hours and an old 600X pack that is still good for about 2 hours.

From one who's been there, the "savings" from a DIY battery project aren't worth the effort. I had a 600X board that was OK, but refused to work after reassembly. I think there might have been a slight difference in the amount of charge between cells and the board locked. The T20 pack worked great for a while, but the T20 sat overnight with the AC power removed until the battery pack was dead. It now refuses to charge.

I would recommend against this project, but it's your money. To me, it's not worth it when for just a few bucks more I get a battery pack that will "plug and play" VS the expense of rebuilding and ending up with a brick. As previously mentioned, you need to see what cells are in the pack and match them up. I used Samsung 18650's.

Joe

[Ash]

Thats one part of the data, so the cells are plain 18650s nothing special. How about the flat prismatic cells as used in X40/60 series ?

Among the commonly available battery manufacturers (in ebay etc), which you'd recommend and which to stay away from ? Are there any known fakes/recalled batches for the recommended manufacturers (if yes how to identify) ?

My precaution is mostly doing the work outdoors, on the concrete (with quick run away in any direction) and far enough away from anything that can get damaged in case of fire, as well as keeping covered connections which i am not working on at the moment (like when working with live mains power)

[RealBlackStuff]

You MUST solder each new cell in parallel to the old cell, BEFORE you remove that old cell!
If you don't do that (or goof), the electronics inside the battery are GUARANTEED to die!

[Ash]

Die as in - beyond the current "lockout" condition ? What is the lowest voltage of existing cells below which the circuit "dies" (as they degrde, those batteries were allready "locked out" for like a year before i got them)

Most batteries i find come without installed tabs. Is the direct heat of a soldering station at 410C directly on the battery (used with sense and im good at soldering, but still batteries are made of pretty hard to solder to metals so it may take some heating to get a connection) enough to damage the batteries ? or it is not possible to do properly and i should not look into anything other than batteries with tabs ?

And, what manufacturers recommended/avoided ?

[joester]

If at all possible, use only batteries with tabs. That tab will conduct heat and take solder much faster and safer than trying to solder directly to the cell. Most manufacturers that use the non-tabbed batteries will spot weld a tab to the case as part of the process to join cells in whatever configuration they need.
As far as manufacturers to avoid, I cannot make any suggestions. I happened to use Samsung.

When soldering, remember that you absolutely DO NOT want the battery to boil. That's why tabbed batteries are preferred for the DIY projects.

Oh, and 410C is a bit higher than I use, but should be OK as long as time of exposure is limited. Try to use a GOOD flux. If you can get some no-clean flux, it's the best way to go.

[twistero]

What is the lowest voltage of existing cells below which the circuit "dies" (as they degrde, those batteries were allready "locked out" for like a year before i got them)

If I understand correctly, the battery control board loses function permanently if the voltage between any pair of battery terminals ever drop to 0V, even for a brief second. I'm not entirely sure of the mechanism, but some people say that some code is contained in volatile RAM onboard, which naturally loses its contents if power is ever disconnected.

This should be separate from the "lockout" you're describing, which sounds like the battery control board cutting off the connection between the cells and external terminals to prevent them from being over-discharged.

So, if you're careful and never let the board kill itself (by disconnecting the existing cells before connecting the new ones or otherwise maintain a voltage between the terminals), there is a chance you could revive them with new cells. But again, I'm not 100% sure on any of the above.

That said, have you tried charging the "dead" batteries overnight? Because I have bought a couple of new old stock batteries that sat on shelves for years, and when they arrived they appear pretty much dead, but after being plugged in for a few hours they would start charging fine. I'm still using one of them.

[hellosailor]

This is part of the mystery of lithium battery controllers. The controller will normally cut off power form the pack if the combined voltage from the pack falls below a certain point, to prevent cell damage. It could well be that if the controller sees any one cell drop below the minimum cell voltage, that it simply shius down in the assumption that the cell has failed and recharging it could be a fire danger.

I'm not saying anyone is using that logic--just that given the way lithium cells can be damaged, and the consequences of damage, it wouldn't surprise me if they were designed to permanently "blow a fuse" under some circumstances.

That would also explain why I've had packs in the past that were perfectly good 100% capacity one month, and stone cold dead the next, just sitting unused on the shelf.

this kind of logic is also used in some car alternators these days. If the car battery voltage drops below 10.5V, even if you recharge the battery the car will not charge it until there has been an ECU or alternator 'reset'. In the logic that if the battery went dead, professional servicing should be required. (Even if it isn't.)

[Ash]

Meanwhile i'm testing out the batteries i have. i begin with the T60 and T61

A couple i insert, there is medium speed flashing of the battery LED on th computer for a while, then it turns into slow flashing and battery begins to charge up (computer does hold on it if i pull the charger after few hours, did not test run time yet)

Most i insert, there is medium speed flashing for a while, then it turns into fast flashing

On the T60 it keeps fast flashing (i tried to leave one for a couple hours and it remained like that, most i only tried for few seconds after it began fast flashing)

On the T61 it alternates between fast and medium flashing. I tried to leave one overnight and it still flashed the same day later, and did not hold the computer when i pull the charger



Is the "fast flashing" a sign of bad battery (needs new cells or is totalled) or it can still rise from the dead if i leave it to charge for long enough ?

Am i doing any harm (mostly to data stored in the controller) by testing the batteries ? (would it be smarter to replace the cells right away before sticking the battery in the computer after all this time - they probably were still fine last time they were seeing a computer couple years ago, and were ominously decaying away in a dark shelf in the IT department after the users upgraded their laptops)

[joester]

I don't believe there data stored in the battery. I think they use SCR's to hold the power until the voltage drops and then the batteries are disconnected.
That being said, I wonder if providing a charge to the batteries individually but simultaneously might induce a "reset" of the monitor circuit.
I know I have a few dead 600 series batteries laying around... Maybe I'll open one up and see if the cells can be charged and the board reset.

I have a Disney Mix-Max that works like that. If the battery is allowed to go flat, the charging circuit will not allow the cell to charge. If I open it up and charge the cell externally, the charging circuit resets and it functions normally again.

Joe

[Ash]

I did just that to an HP ipaq which battery went flat. It displayed "no battery". i charged it up to 3.7v with a voltage divider made of 2 resistors off a 5v power supply overnight, then stuck it back in the ipaq (connectedto its power supply), this time it displayed "0%" and remained like that for few hours but eventually went to 100% and worked fine since

The Thinkpad batteries do have some data stored in them according to stuff i found in google

[hellosailor]

The power manager software does read some data that is stored in the battery pack. If you unintuitively scroll the screen down "below the fold" you will find the power manager knows each battery pack by serial number, maker, and other information that is stored in the pack.

I have no idea what else is stored in the pack, just saying some data definitely is in there.

[joester]

But that data can be written so that power is not required to maintain it. We do that with our products at work. Every device is hard coded with a serial number as a minimum. Some devices have hard coded firmware and can sit on a shelf for months with no battery and no issue.

[hellosailor]

joester, you said "I don't believe there data stored in the battery." and I responded to that.

Whether power is needed to maintain the data is a different issue, but static memory (flash, eeprom, whatever) is so cheap that's a non-issue. Either way the logic can be set to "never use this device again" after one singular event, which could be a low voltage event, or a memory loss. Which in turn could be used as another way to confirm a low voltage event.

When you're trying to reverse engineer the logic some lawyers and accountants have laid on top of what the engineers did...Good luck with that! The lawyers don't care what you spend, they want to make sure you won't sue them over a battery fire. The accountants want to sell more batteries. And if the engineers ever told them "you know, we can put a clock chip in there to make sure the batteries expire..." we'll all be in trouble.

[joester]

I get that.

There was also an edit that got missed in the post. (lesson learned, never respond at work when several people are likely to interrupt)

I intended to write that I didn't believe there was active code written that required maintenance power to retain. I have no intention to override safety protocols and jumper around the electronics that do exist inside the batteries. Anyone doing that is begging for a problem and would be lucky to just lose the battery and laptop in the fire.

I'm now wondering if the same "recharge and restore use" measures can be applied. I do have a few "dead" 600e batteries I can experiment with, and I do have the means to provide a controlled overtemp situation with a small heat gun on the sensor while charging if that can be made to happen.

Clock chip... I believe those do exist in some products. I had a cell phone that worked great until one day when the battery would not hold a charge for more than a day. Cost of the replacement battery was near the cost to replace the phone (not a smart phone).

[Ash]

I cant find the original post i mentioned but here is another one mentiomning the stored data as well

http://forum.thinkpads.com/viewtopic.php?f=2&t=77856