• Please review our updated Terms and Rules here

IBM PS/2 L40SX battery replacement

A

Acorn

Experienced Member
Joined
Aug 29, 2007
Messages
68
Location
Atlanta, GA
I have an L40SX that I bought new when they first came out. The main battery died probably 10 years ago. Just for fun I decided to make a new battery for it.

The original battery is a 10.8V 2.4Ah Ni Cd pack containing 2 groups of 9 cells in parallel. The pack has 3 external contacts "+", "|" and "-".

I carefully opened the pack by sawing along the separation lines on each side. Inside I found a small temp sensor that attaches from the middle contact "|" to the negative contact "-". It's a variable resistor (thermistor) that protects the batteries from overheating while charging. If the impedance falls outside the range from 4K to 30K then the charging circuit shuts down.

Also inside I found some kind of circuit attached from the negative terminals of the 2 groups of 9 cells to the negative terminal on the battery pack. This circuit is encased in heat shrink tubing. I haven't cut into it yet so I have no clue what it is or does. Does anyone know?

On to the "science experiment"...
I bought 2 of these:
AS10.8V1100.jpg


They are 10.8V 1.2Ah NiMH battery packs normally used for paint ball guns. (5" long x 0.7" High x 0.25" Wide). Connecting 2 in parallel gives me exactly the 2.4AH capacity of the original IBM battery but the better non-memory characteristics of NiMH vs Ni Cd. These packs fit fine in the battery compartment of the L40SX but not in the original battery pack shell without some modification. For now, I'll leave the modification for later.

I installed the 2 NiMh packs in the L40SX with a quick tack solder to the backs of the appropriate terminals within the laptop. I also attached the thermistor to the center and negative terminals and physically taped it directly on top of the cells. Everything seems to work well. The laptop charges and discharges normally as far as I can tell. It gets nearly 3 hours of runtime. So far, I have NOT installed the mystery circuit that's inside the heat shrink tubing.

My theory is that this mystery circuit monitors current flow and somehow monitors the charging rate of the batteries. Maybe it senses when the batteries are charged because it sees reverse current flow or something. Who knows?

NiMh batteries seem to be relatively cool while charging until they are nearly fully charged...then they get quite hot! The thermistor protects them until they cool down but then the charging process tries to start again. Obviously there's a piece to the puzzle that missing.

Once I get all of this sorted out then maybe I'll try repackaging the batteries into the old plastic shell.
 
Last edited:
This is all just total guess work. I would bet that the "weird" third terminal is used to shut down the charging process (voltage goes up or current drain drops dramatically?). Some sort of voltage regulator setup would be kind of stupid. Oh, let's keep the current coming and short it out when not needed.

One of my old battery packs did that - was great until it got to 99% then would get very, very hot. I am putting it into the recycle barrel.
 
chuckcmagee said:
This is all just total guess work. I would bet that the "weird" third terminal is used to shut down the charging process (voltage goes up or current drain drops dramatically?). Some sort of voltage regulator setup would be kind of stupid. Oh, let's keep the current coming and short it out when not needed.

One of my old battery packs did that - was great until it got to 99% then would get very, very hot. I am putting it into the recycle barrel.
Click to expand...
No, the third terminal is clearly just connected to the thermistor. It just shuts down the charger when temperature rises above or below a certain point. Nothing inside the battery but the thermistor is attached to this terminal. The charging circuit expects to see a resistance value of 4k to 30k between the "weird" terminal and the negative terminal.
 
More thoughts on the "mystery circuit"

More thoughts on the "mystery circuit"

I was doing some Googling trying to find out more about battery packs and found this:
http://www.pda-soft.de/emate_battery_pack.html

It sounds similar to the L40SX battery pack except it's in an Apple eMate. This battery pack has a thermistor just like the L40SX but also another component IN SERIES with the batteries just like my mystery circuit. According to the article, this circuit "will reduce the charge current to about 100 mA if the cell temperature exceeds 55 degrees celsius."

I'm betting my "mystery cuicuit" does something similar. It was in direct contact with the cells inside the battery pack and it was wired in series.

Hmmm
 
Hmm, ok, guess my "stupid idea" method might really be used then, i.e. when voltage gets to X, shunt the current around the battery. I have noticed those small circuit boards in all the batteries I have taken apart.

Edit: YEP YEP YEP, THAT'S IT. In series, as in cut the current. I knew the shunt idea was stupid and wouldn't work too well. You would need like a transistor or something to keep the battery from "shunting" at the same time.
 
Last edited:
Shoot, guess I could rescue that one that get's super hot. It's clear the only problem is a cute little circuit isn't cutting the current down at the end. I purchased some new replacements already, arrived today. Nah, think I'll leave it. Thing is dangerous. Not worth burning down the house for $60.
 
Mystery circuit

Mystery circuit

Here's what the mystery circuit looks like.
It's about 5" long and runs the full length of the battery pack. The large end on the left is connected to the Negative terminal of the battery pack.
The 2 smaller ends on the right are connected to the negative terminals of each of the 2 groups of 9 cells within the battery pack.
 

Attachments

  • mystery.jpg
    mystery.jpg
    18.8 KB · Views: 1
Well, we know how it works now so not really much of a mystery. Also clear why the pack gets very hot at the end of charging. Energy not going into chemical energy anymore.
 
Also clear that me leaving the NiCad charger on too long is what caused them to expand and break up everything. All the lights are coming on (in my head). They warn you about overcharging. Now we know why.
 
NiCad vs NiMH

NiCad vs NiMH

As a side issue but related to all this, it may not be a great idea to substitute NiMH batteries for NiCad. It makes no difference to the laptop because it will happily use either. The problem is with the charging circuit and the batteries. Up to a point their charging characteristics are similar.

My research is telling me that NiCd's like to be trickle charged and NiMH's do not. Most NiCd chargers are made to agressively charge the battery until the change in voltage with respect to time hits zero or goes negative. Then the charger switches over to a trickle charge.

This point is difficult to measure with NiMH batteries and you want to stop charging once capacity is reached vs continuing at a trickle rate. Overcharging NiHM batteries can severely limit their overall life.

nickel_chargers.gif


Here's more than you ever wanted to know about batteries and chargers:
http://www.mpoweruk.com/chargers.htm

The charger in the L40SX is rated to fully charge the battery pack from a full discharge in 10 hours. That's a pretty slow charge so the rate of charge is fine for NiMH. However, the charger continues to trickle charge after the battery is fully charged which is not good.

I'll have to remember to unplug the charger when the battery gauge hits full!
 
Well, when it hit's 99% anyway. I dug out that battery pack that was getting really hot. It is great until you get to 99% charge. Then it stays there, at 99, for a very long time. Meanwhile, that battery get hotter and hotter. Too hot to even handle if you let it go. Turns out to hold about 42 minutes of charge. It should be good for an backup kind of thing. Is a pain to remember to shut down the charging.
 
epic very yes!

epic very yes!

Entirely seconded. While we're at it, why not post links to other resources, too?

http://www.mcamafia.de/ -- inordinate quantities of ps/2 stuff ; very useful.

http://hpholm.dk/L40SX.html -- HEAPS of manuals and specs ; not quite as useful.

That's all i've got ATM, but google also.

MCA Mafia have useful things about how to modify standard 72-pin SIMMs to work in ps/2-boxen. Very useful if you have heaps of non-parity RAM lying around (like I do :D), and many ps/2-boxen (like I don't :???:) in desperate need of memory. Minimal soldering skill required!

I plan to "pimp out" my L40SX (dammit, everybody has one!) as much as I can, possibly getting a new HD cable for it ; my current one I suspect of being broken.

Anyone know maximum RAM the L40 can handle?

Does anyone else's L40 have a BASIC 1.0 rom? (Really weird, mine is second hand, and I think the first owner did some... uh, modifications. Can this ROM be replaced with something useful? (At least more useful than basic 1.0, haha.)

I have an (probably original) L40 technical reference manual-thing, which is pretty cool.

It is 3:20. In 24-hour time. I am not asleep. I believe it to be time to remedy this situation.
 
Terry Yager said:
I have a 'spare parts' L40SX. I'll have to dig it out and see if the HDD cable has already been cannibalized.

--T
Click to expand...
Is the motherboard good on that "spare parts" L40SX?
I think the on-board RAM on mine is flaky so I may be in the market for a new motherboard.
 
Hdd

Hdd

I just remembered/realised (rememberised?) that HDD not working might also mean a bad controller on the motherboard.

I have tried two hard drives, both known good. The hard drive spins up on boot, but it is inaccessible as far as the BIOS is concerned, which means also inaccessible to any OS. I've tried everything, and it still won't work. This leads me to conclude that the cable or the controller is dead.


Please excuse my australian spelling of "realise". I think most everyone else uses "realize".


EDIT: Also, do the SIMMs need to be paired, or can two different ones be used?
 
Last edited:
Acorn said:
It will take 18MB total. Qty-2 8MB sticks plus 2MB on the motherboard.
Click to expand...

Plus there is even a way to easily modify a halfway obtainable 16Mb EOS SIMM to work in the slots for that 18Mb (and not have to find the rare 8Mb SIMMs that are particular to the model)...
 
You must log in or register to reply here.
Back
Top