Thanks for your replies guys. I took the laptop apart again and did some resoldering , including L7 (it seemed fine anyway) and where the battery connector is soldered to the motherboard - made absolutely no difference. The charging circuit seems to be only passing mA rather than Amps to the battery.
Another symptom is that when the laptop is off, whatever the charge LED is showing when you unplug the AC adapter, the LED remains on. If I briefly disconnect the battery, the LED goes off and stays off until you plug AC back in. Then it does the amber-flash for about 10sec then steady amber. After about a day and a half the LED will go to blinking green, then eventually steady green as it should.
I'm probably wasting my time without ruling out the battery. Where ever the fault is, it's not affecting discharge current - the laptop will run on battery for over 1hr. There's just something somewhere not allowing full charge current to flow into the battery/cells.

This is how I broke my T23. One of my co-workers was the thinkpad tech and figured this out immediately. We were at my house and had shoddy tools. It soldered on just fine. I used flux that I found somewhere (don't ask). It turned out to be mostly water and when we booted the board the power chip went up in smoke. Burnt traces, sparks and everything.

My t23 had all the issues described with the loose inductors. Got a 15w gun from radio shack and some 'high tech' silver solder and tinning stuff. Took out the motherboard and found two inductors just hanging there and tinned/soldered them back. Bingo-- machine is like new! Thanks for the advise!!

Well, I finally got around to trying this out yesterday. After disassembling a T23 that I had sold a couple of months before, that the buyer had returned not working, I found that the rear most inductor was indeed loose on one side, and looked to have been soldered by someone before I ever had it. So, I soldered it back, and re-assembled to find I had not done it right (still had the pulsating fan sound). I think I only got solder on the side of the inductor and didn't make good contact with the pad under it. Now, I'll have to disassemble the thing again and try again.

Is there any way of testing your work without having to completely reassemble everything?

Yes, but it involves some circuit tracing and an ohmmeter.

Look closely at the board and try to follow the circuit trace from each of the inductor pads until it either ends at a through hole (point where it "dead ends" and has a hole in the middle) or reaches another component. Connect the test leads of the ohmmeter between the pad on the motherboard and the other end of the trace. It should read 0 ohms. Do this on both pads (trace the circuit and verify it reads 0) before re-attaching the inductor.

When you have soldered the inductor back on, connect the ohmmeter's leads between the metal tab on the top (or side) of the inductor and the point at which found earlier. It should read 0 ohms. That means that the inductor is attached to the pad. If possible, at the same time you are measuring this path, try lifting up on the inductor or moving it side-to-side.

Thanks Ray, good suggestions.

I was dreading the next attempt at fixing the board, but you've given me confidence that I can check my work without putting everything back together.

Edit: It went much smoother this time! Using it now to write this!

I wanted to 'upgrade' to built in wireless, so I bid on two 'dead' 2647-4nu's on Ebay. I took apart the first one and found an inductor loose in the bottom, so I cleaned it up, re-tinned it and soldered it back on. Now I get a 1-3-3-1 beep code even with a stick of known good memory. So I set it aside and disassembled the other one, sure enough the same problem! After cleaning up the inductor, re-tinning it and re-soldering it I reassembled the second unit enough to try it and it booted to bios so I finished reassembling it completely, put in a 20 GB disk, installed Windows, updated it's bios, controller and wireless card firmware, installed the IBM driver and everything works. I set it aside and went back to the first unit. I've rechecked, and re-soldered all the inductors and still get the beep code. What else can it be? I've kinda given up, for now on this one so I swapped the lid, complete with LCD, wireless card, and processor into my 2647 2tu and it now is a bit faster and has built in wireless that works.
It still baffels me how another problem could develop and still have the inductor fall off.
Regards

It's possible that your first 4NU has a problem with the memory slot(s). If you haven't already, try moving the known good module to the opposite slot and see if that helps. I've run into at least two T23 systems with this type of issue, although both were (and still are) intermittent problems.

The way I see it, both the memory slot failure and loose inductor can be triggered by the same event: a fall or significant physical shock to the system.

My wife has been using my T23 for years now as her primary system thanks to this thread.

I got a T23 around ~2005 or so. I forgot if it was back in 2006 or 2007 that I finally pulled apart the system to re-solder the piece back on.

However, after months/years of working fine, three times now in the past few months the system has done the "fan pulse". Shutting down, pulling the battery, then hooking things back up seemed to wake it back up.

I really suck at soldering, so I dread pulling the system apart again to try and fix the issue again.

I was just directed to this thread from a thread i started about my wonderfull children knocking their t23 off the table..

Let me attempt to find the soldering gun or pencil and fix it.. Lucky that I have had the t23 apart a few times...

Will keep you informed..

Thanks

Here's my most recent experience with a system which was acting up...

T23 (2647-5NU), bought off of eBay 9/07, which sits on my desk at home and is the one which I do all of my web surfing from when I'm in my office. It has been connected to a port replicator for about 7 or 8 months and from time to time is on for several days straight. For the last three or four days, the system would randomly freeze. This mostly happened while in Firefox, right after I started typing or pushing the trackpoint. There were also two times when I came back and before I even touched it, noticed that it was frozen. The screen would stay lit and everything on it appears normal but there's no control over the keyboard or mouse pointer and the clock utility that I keep on top of all the other windows would stop running. What would usually work was holding the power button down for about 5 seconds. Twice I had to resort to pulling the AC adapter and battery out before it turned off. Upon restarting, the system always would boot normally and Firefox would reload all of the tabs that were open. I never saw any BSOD or relevant error messages in the system logs. I tried undocking the system but the problem would still be there.

My initial thought was that I was hitting the system with a static shock but I've been careful to first touch the metal lamp on the desk after sitting down. And that didn't explain the two times when I left the system running but returned and saw it hung. Looking back through my notes, I saw that when I first received the system there was a problem with memtest86+ failing when both slots were occupied. I had since replaced the original modules with 512MB sticks so I didn't think that was the issue. However, because I have a floppy drive connected to the port replicator I decided that the next time the system froze I'd run memtest86+ on the next reboot. So the T23 went back onto the port replicator and in went the boot floppy. When the hang occurred again, I power cycled the system and memtest86+ started running. Everything seemed to be okay but after about two minutes I realized that the progress bar was not moving like it should. I looked at the elapsed timer and it was stuck at 0:30; I had been looking at a frozen screen for at least a minute .

Now I started thinking that I might have another system which has a loose crystal on the motherboard. The symptoms seemed to be very similar in that the T23 would just stop running without any warning. But when I tried to restart the system, I heard the familiar fan surging . Turned the system off and then back on again and it booted normally Seems like I have another project waiting for time to do a complete teardown.....

Transferred the hard drive and DVD burner into this 2647-4U2 (upgraded with a 1.20Ghz CPU but with a slightly pink screen) and I'm back in operation at my desk.

Hello everyone, nice forum do you have here - sorry about bumping up an old thread, but, I thought I shouldn't start a new topic, as my problem somewhat coincide with the main theme. And here it is - I bought second hand faulty T23 few months ago, seller said it was something with motherboard, but that was ok with me, since i needed just tft panel for my old T22 which had some pressure spots on it. Nevertheless, I thought I should give it a try - so i hit the power button, and CPU fan started spinning (steady, without surging), but not much else, no signal showed up on display, neither HDD showed signs of life. I replaced the keyboard with the one from my old T22, and bingo, it booted! However, after i rebooted it some ten minutes later, it went dead, and wouldn’t boot again. Now i searched the Internet for some guidelines, found your forum, and learned that most commom problem with T23‘s motherboards is associated with loose iductors. I dissasembled it, and indeed, there it was, the one in the area underneath the space button had the hairline crack on the solder joint, and i easily took it off. I’m not so versatile in soldering so i took it to friend who works as a computer repairman. Now, he resoldered the inductor, though i think that he had not cleaned the places where inductor was connected to the board, which were lightly coated with black film, but tested it afterwards and said the conduction was good. I assembled the computer again, but it acted just like at the beggining - “Z” button lights up, CPU fan spinning, no display signal, no HDD activity. I should try again the trick with the keyboard (although i feel that was coincidence), but at the moment i don’t have my T22 with me. Any other ideas and advices would be greatly appreciated. Cheers - and sorry for my poor English!

Welcome to the forum.
Start with the standard procedure, take everything out (RAM, hard disk, wifi/modem card, CD/DVD drive).
Take the CPU cooler off, and check that there is thermal paste between CPU and cooler.
If not, the CPU may have died from overheating.
Try to boot. If the motherboard is still alive, it should beep 1-3-3-1.
If not, your mobo is probably dead.

Thanks - i followed your advices, took everything out and tried to boot, unfortunately, no sound came from it...just "Z" light showed up. Anyway, i always can use it for what i first intended - can i replace T22 900 MHz CPU with the 1.13GHz one from T23?
Many thanks again for your time and patience.

No you can't. Different chipset on the T20/T21/T22 from a T23.
T20/1/2 runs at FSB100, T23 runs at FSB133.
The only things that can be exchanged between a T23 and the lower T2x machines: keyboard, speakers, CD cage, PCMCIA cage and lid with LCD (and hard disks and optical drives).

I have fixed several T23 systems with loose inductors and I always take the time to clean off the pads the inductor that have the black film. It is a layer of oxidation which can affect the soldering joint over time. It only requires a few seconds of scraping with a file or screwdriver blade to get rid of the black film and expose the copper.

Just resurrected a busted T23 using a busted T20 for parts.

Reading this thread afterwards, I figured it might be helpful to pipe in with my experience.

Had the pulsing fan issue after replacing the T23's LCD/bezel. I was told by previous owner that "something else was going on besides the crushed screen". After hunting around, I found a bunch of schmeg and corrosion around the S-Video port on the secondary PCB or "sub card". Replaced the sub card with the same from the T20 and the Thinkpad fired right up. Looking good so far...

Tried to install Windows 7 from disc but the unit kept crashing at random points during the install - hard crashes that required battery removal and AC cord unplug. Realized I needed to update the BIOS and BIOS controller, I attempted a Windows XP install. The crashing continued but I was able to make progress through several attempts and came to some conclusion that the CPU was overheating. I installed a 1200/133 PIII from the T20 with fresh thermal paste after removing the previously installed 1133/133 PIII and after several times tightening the heatsink/fan, the system stabilized enough to complete the XP install.

Updated BIOS, updated BIOS controller, installed Lenovo drivers from website and now I can't get this machine to stop!

Literally, after everything I just wrote, the machine is unstoppable! I've run burn-in tests, stressed the CPU, ran streaming video while playing a DVD while on battery, etc., etc., etc.

I love this platform - back in 2005 I was setting up family, friends, and more with T2x models. I'm vindicated again!

Glad I didn't have to solder... now I know about the inductors in case it happens. Some parts available now... PM if interested

The T20/1/2 I/O board does not fit in a T23 without major modification (and v.v.)

I'd be extremely surprised to see a 1.2GHz CPU in a T20.
Besides, T20/1/2 CPUs do not work in a T23 (or v.v).

I assume your donor machine is another T23 with a keyboard that has a T20 template around the power-button.
What is/was the 7-character type/model of the donor? (2647-xxx or 2648-xxx)

good call realblackstuff:

2647-8RU

...and sure enough, it is (was) a T23 with a T20 cover plate. Hadn't thought about this donor T23 in years. It likely has the inductor problem mentioned in this thread because when it went dead around 2007 it was for no good reason other than no POST under any circumstances. I kept it around figuring someday, with some time, I might get it fixed up.

Ahh, the memories are coming back now. Since then, I had a buddy whose T20 keyboard was borked and I replaced it with the dead T23's keyboard - and likely the cover plate.

No wonder everything between the two machines, especially the fan assembly, had exact same FRUs. [forehead slap]

I'll look for loose inductors on the donor system board...

OK, ran into the pulsing fan and crashes with my rebuilt T23 and decided to go back in.

Found the loose inductor on the underside of the system board. I did my best solder job and put the whole kit back together.

Success - The T23 rides again! Thanks again to this thread.

FYI: if you already (re)solder one inductor on the bottom, you may as well redo the second one, even if it still 'feels' OK.
Saves you future heartache.
And also take a look at the same shape/size inductor on the top, left of where the CMOS battery goes.
If that one gets loose, it can also cause pulsing.
(I repaired a T23 today for all 3 inductors).

There are a few good quick hints. The modem cover is larger. You can't find both keyboard screws, without pulling the modem cover. And, the easy one, 2 USB ports.

I just got off Ebay, (15.00 plus 12 shipping) a T23 that didn't work, no mem, no HDD/caddy/cover, no batt/PSU, but has a wireless card AND wireless screen. I just wanted the screen and plastic.


So, as I was pulling parts off, I decided to pull the mobo off the bottom cover. Lo and behold, there was a part sitting on the lower cover. I'll be soldering an inductor soon.

SCORE!!

steve

Just want to thank you people for the advice. My T23 had the pulsing fan and, sure enough, 3 out of 4 solder joints on the two inductors had fractured. After nervously wielding my 20w soldering iron and a few false starts, I now have a working T23 again. Thank goodness, as I use it as a printer server for my network. They are great little machines, aren't they?

Kind regards,
Doug
T23 WinXP / Fedora 14
T61 Fedora 14
Assorted desktop boxes Fedora 14 / One Win98 (blech)

Ok, I've got 2 more coming off of Ebay. Both were less than $25 each, including shipping.

One of them was disassembled, with a picture of the inductor fallen off. It comes with a DVD burner, as well as a DVD drive.

The other one has a wireless screen, and the guy referenced this thread, talking about the pulsing fan. He pulled the HDD, and kept the AC adapter.

Both have batteries, neither have AC adapters or HDDs. Good score for $45.

steve

After much deliberation on how to go about the job, it turned out very ugly. The inductor was about .050" above it's normal position, and feared it would contact the housing. Fortunately it cleared the case. So I connected the keyboard to do a simple post, and NO BEEPS. Then I looked down at the speakers laying on the desk. After replacing the speakers the 1-3-3-1 confirmed a basic post. Stuck in a HDD and it booted fine, wrapping it up as I type.

One more T23 saved by THINKPADS.COM

Just a follow up on my exp. I used a selectable 40/20w iron, and used 20w only. There was no black residue on the board pad, but after removing the inductor (which I didn't want to do) notice the inductor feet were charred w/ black oxide. Used a nail file/emery board to get a bare copper finish. Bought a roll 62/36/2 silver-bearing solder .015" diam, and found just the amount that would "sweat" on the tip to be too much (the reason it is .05" too high). A little crowded on the side broken in this page http://www.ziva-vatra.com/index.php?aid=18&id=SGFyZHdhcmU= but managed not to do any damage. Had that page b/m for years, yet found it strange no on brought it up on this thread.

LOL <Been there, done that>