TP 755CX - issues with mwave, sound and modem

[stamasd]

I did a bit of experimenting today with some batteries. Still no progress, to sum it up.

I connected a 8-cell NiMH pack to the internal battery connectors (fully charged, reads 11.04V). I also connected a 10kohm resistor between "therm" and ground to simulate a thermistor.
The laptop does not recognize this combination as a valid battery.

I checked, with the above configuration connected, the "send" terminal. It reads 5.1V on my multimeter. A check with an oscilloscope confirms that it's a pure DC component, with no signal present. Connecting a 4.7kohm resistor to ground brings the voltage on the pin to 2.7V, no change in the status of the battery indicator. Connecting to ground via a 100ohm resistor brings the voltage down to about 0.2V, still no change in battery status indicator.

I think this indicates that the "send" pin is a TTL input (probably pulled up by a 5kohm resistor), awaiting for a certain signal sequence from the battery. But without a working battery there's nothing I can do to start figuring what the signal is supposed to be.

[RealBlackStuff]

Some Dell BIOSes can be updated without a battery using an undocumented switch: /forceit
Example: C:\Documents and settings\xxx\desktop>D620_A10.EXE /forceit
Wouldn't cost anything to check if Lenovo also has a similar option (although I doubt it).

Or read this ancient forum thread: http://forum.thinkpads.com/viewtopic.php?t=6577

[stamasd]

Some Dell BIOSes can be updated without a battery using an undocumented switch: /forceit
Example: C:\Documents and settings\xxx\desktop>D620_A10.EXE /forceit
Wouldn't cost anything to check if Lenovo also has a similar option (although I doubt it).

Or read this ancient forum thread: http://forum.thinkpads.com/viewtopic.php?t=6577

Nope, no such option. The BIOS flasher is in fact a GUI application that takes you through a README, asks you to read a license, then lets you choose to enable the flashing; then you have to reboot, the flasher starts again and if you have a charged battery, starts the flashing. If you don't have a charged battery it won't let you do it. It does not take any parameters on the command line.

Also the method used for the TP 600/770 series does not work for the older 755 because the flasher is different. There is no flash2.exe on the system upgrade disk for the 755. And the byte editing method doesn't work either because updtflsh.exe for the 755 is different from the one for 600/770. I already tried.
Using phlash.exe is not an option either. Phlash requires 2 files, a .fl1 and a .fl3 file, one of which is the bios file and the other a machine description. The bios upgrade disk for the 755 only has a .fl1 file, not a .fl3 file.

[RealBlackStuff]

H132156 - 9545 LI-ION BATTERY INFORMATION.


SYMPTOM:
The following Tip is applicable to 755CX and other 755 models
using a Li-Ion (Lithium Ion) main battery FRU P/N29H9232.

The system:
1. fails to power up using the battery.
2. will power up using the AC Adapter.
3. will fail to charge the battery during a normal charge cycle
with the system off.
4. will not charge at all while it is powered on.

PROBLEM ISOLATION AIDS:
1. The battery fuel icon shows the battery as a dashed line
instead of a solid out line.
2. The charge indicator will blink during the charge cycle when
the system is turned off.

CAUSE:
The Li-Ion battery was completely discharged for an extended
period of time due to use or new battery shelf life.
Follow the recovery procedure in FIX section to correct this
condition and obtain full battery potential.

FIX:
The battery must be charged for a minimum of eight (8) hours.
This can be done in the system while it is off or with the
external battery charger.

To ensure this does not occur again, recharge the main battery
when the system indicates a charge is needed. The Li-Ion battery
can be recharged at any time during the normal discharge cycle
and obtain a full charge - they do not have to be completely
dis-charged before recharging.

Once the battery is charged it can be stored for extended
periods of time (up to a year) without incurring this condition.

This condition will not effect the life of the battery.

[goldeneagle]

I just skimmed through this thread (trying to get NetBSD working on a vintage Sun at the same time, with a 755C as my console).

I actually had problems with mwave as well. In both OS/2 Warp 4, WFW 3.11, and I think NT 4, I couldn't get midi to use the wavetable sounds. It's profoundly annoying, and followed all the directions. Wave files work, sometimes anyways. I had a 760ED that I actually swapped the mwave board out with a 760XL that had an FM board (forgot if it was crystal, Cirrus logic or Analog devices). That worked like a charm, but I don't think that's an option for our beloved 755 series machines.

While the concept was cool, mwave was clearly the worst mistake of IBM since Microchannel (after struggling with PS/2 problems for weeks).

I'll keep everyone updated if I get a working battery for the 755CE/X; I rebuilt one for my 755C, but I might've messed up the circuit board, though (after testing certain pins with a multimeter per the HMM). I actually used new Sanyo cells and spot welded them. I got lucky and had a battery that worked on both the 760 and 360 series (even held a charge for over an hour), so I update the BIOS on all those machines while I had the chance.

[stamasd]

Quick update.

I tried recovering one of the batteries I have with the quick travel charger; I left it in the charger overnight (9+ hours) and while the light on the charger was flashing as if it was actually charging the battery, in the end the battery failed to take any charge and was at the end as dead as it was at the beginning. Not surprising, though disappointing. I will try the other batteries I have but don't have much hope. If only I had or could find somewhere a NiMH battery - all the ones I have are lithium.

[goldeneagle]

Quick update.

I tried recovering one of the batteries I have with the quick travel charger; I left it in the charger overnight (9+ hours) and while the light on the charger was flashing as if it was actually charging the battery, in the end the battery failed to take any charge and was at the end as dead as it was at the beginning. Not surprising, though disappointing. I will try the other batteries I have but don't have much hope. If only I had or could find somewhere a NiMH battery - all the ones I have are lithium.

I'll have to let you know how that goes. I have some NiMH batteries for the 755CE that I'd like to rebuild, and I believe these should work with the 755CX. The 760 series Lithium Ion batteries aren't compatible with the 755CE/CX, especially the way the battery casing is keyed (to prevent accidental insertion). It's just a matter of me opening up a pack, disconnecting the control circuitry, rebuilding the pack, and reassembling, and also finding the time to do it (I like to set up the spot welder outside on a work bench, and it's not pleasant doing so when it's below freezing for 2 days straight).

[stamasd]

Hmm, I beg to differ on the 760 batteries. Back in the day when I had a 760, I used its battery in the 755CX and it worked well. It was Li-ion too. A 755 battery won't work in a 760 though because it's thicker and won't allow the keyboard to be seated correctly.

I do my spot welding inside, it's not like it releases toxic fumes when you weld a thin strip of nickel to a battery. But I don't have any 17500 cells, and the prices on those are insane lately.

[RealBlackStuff]

Maybe this helps to get some battery power: http://makezine.com/2016/11/29/quick-ti ... ack-cells/

But I still think you should make a battery cable like I described before...

[stamasd]

Maybe this helps to get some battery power: http://makezine.com/2016/11/29/quick-ti ... ack-cells/

But I still think you should make a battery cable like I described before...

I don't have a working battery for any of that generation of laptops, so I would have nothing to connect at the other end of the cable. I only have working batteries for modern laptops, and those use a completely different type of signals to communicate with the laptop. See my experiments earlier in the thread; I still don't know what the "send" signal is.

Replacing cells is an option, however:
1. I don't have the 17500-type cells that those battery packs use (I have plenty of 18650 cells that are used in the majority of more recent packs but those are too big to fit); 17500 cells are hard to find and expensive
2. the cell harness inside the battery is very complex (more so than any of the other battery packs I've rebuilt before) and I doubt I'm skilled enough to remove the old dead cells without destroying it; if that happens I'd be unable to rebuild the battery even if I had the required new cells.

I'm trying to find a NiMH battery, even a dead one as I suspect it would be easier to rebuild than a Li-ion one; I haven't had any luck finding one though.

[RealBlackStuff]

Download this PDF: BH_US_11_Miller_Battery_Firmware_Public_WP.pdf
It might have some useful info.

Also have a look at the description of this interesting kit: http://www.ebay.com/itm/112237499665

More info: http://frantisek.rysanek.sweb.cz/battery.html

And do a Google search with: battery send signal site:smartbattery.org
to get more PDFs with relevant info.

The END.

[stamasd]

Download this PDF: BH_US_11_Miller_Battery_Firmware_Public_WP.pdf
It might have some useful info.

Also have a look at the description of this interesting kit: http://www.ebay.com/itm/112237499665

More info: http://frantisek.rysanek.sweb.cz/battery.html

And do a Google search with: battery send signal site:smartbattery.org
to get more PDFs with relevant info.

The END.

While all of those are interesting (yes I read all that material), they only apply to SMB/I2C batteries and newer, which use the above 2-wire signals to communicate with the laptop. There is a separate clock line, and a data line. The TP755 batteries use (in the case of the NiMH battery, which I am reducing the case to as the simplest model) only 1 wire, and I think it's unidirectional, from the battery to the laptop. These batteries precede by a few years all of the SMB/I2C/1-wire protocols. It was probably a proprietary in-house IBM protocol, which I don't know the details of.

While the above device is ingenious, it's unlikely to help as it needs 2 wires to communicate with the battery. And the battery has only one.

[stamasd]

Well, while I'm still looking for a solution to get the BIOS flashed, I installed DOS 6.22 and Windows 3.1 on this laptop. Mwave works fine with both DOS and Win3.1. It just makes me sad I can't have it working in Win98.

[RealBlackStuff]

Last suggestion, then I'll stop.
Just connect the + and - from any other charged battery to the + and - in the laptop, using some wires.
Don't connect the charger yet.
Now try to boot.
If that works, switch off, connect also the charger and try to boot again.
If that works, update your BIOS right after that.
Good Luck!

[stamasd]

I tried that already with a NiMH battery pack (8 NiMH AA cells in series; fully charged gives almost 12V). It was not recognized by the laptop as a battery, and it wouldn't boot off it if AC adapter was disconnected.

[stamasd]

Quick update, I managed to find a NiMH battery for TP755. It's not dead but pretty low, reads about 6V. Won't take a charge in the Thinkpad or in the quick charger. I tried an external power supply, at 12V it will only take 40mA which is I'm sure only going through the protection circuit, none of it to the cells themselves.

I will next try to freeze it overnight and see if it will take charge tomorrow.

Worst case if that doesn't work I may open it up and try to replace the cells, or use the PCB + connector with an external NiMH battery pack.

Goddammit, I only need a battery that will read 100% charged for a few minutes.

[RealBlackStuff]

Don't want to read the thread again, but have you ever checked the fuse(s) in/near the mobo's battery circuit?

[stamasd]

Don't want to read the thread again, but have you ever checked the fuse(s) in/near the mobo's battery circuit?

If you're referring to the laptop's motherboard, no I haven't. But with no good battery I don't think looking for fuses would be the next step to do. If you mean the battery's protection circuit, there isn't anything resembling a fuse on it.

If I had a good battery and the laptop didn't recognize it, then yes I would be troubleshooting the laptop itself further.

[stamasd]

Little progress, but somewhat encouraging.

Freezing the battery didn't do anything. It started at approx 6V, was frozen overnight wrapped in paper towels and a ziplock bag; reheated to room temperature this morning, was reading 2.9V. So no voltage boost from the freezing-thawing process. I wasn't really expecting it anyway.

Since then I have hooked the battery to a constant current source. I have pumped in it a constant 100mA for the past 8h. Initially the battery required 13.6V to achieve this current, after 8h it has dropped to 12.2V. This is good, shows that the battery's internal resistance is decreasing. The unloaded voltage of the battery (with the charging circuit off) is about 8.5V. Moreover, when inserted into a Thinkpad the "fuel gauge" reads it as 26% full. The Thinkpad will not boot off it though. Nor will it try to charge it (this is bad).

Now, with the 100mA for 8h the battery received roughly the equivalent of 0.8Ah (minus losses). The battery is rated for 3.5Ah, so 0.8Ah would be approximately 1/4 of that, so the numbers make a bit of sense.

I have connected the battery back to the current source and increased the current to 150mA (which now requires about 13.05V to achieve). Will leave it like this overnight and see what progress I've made tomorrow.

[stamasd]

After 12h at 150mA, the battery now reads 9.2V in open-circuit. I tried to see what % the Thinkpad sees it charged to, but for some reason the laptop doesn't want to boot when the battery is inserted. With both AC and battery in, it powers up, shuts down immediately, powers up again, shuts down etc. With no AC plugged in (just the battery) it will not power up at all.

I will let the battery rest for a few hours and try again. If it still doesn't work I will probably discharge the battery partially (down to about 5-6V) with a resistor pack. 100ohm/1W should give me 80-90mA discharge current which should be enough. Then I will try to charge it again, and again discharge/charge a few times.

This will take some time.

(edit) with the 100-ohm resistor connected, the battery only gives about 50mA and its voltage decreases to 5.6V; this means high internal resistance, estimated 60-80ohms.

(edit2) After 6h of discharge the voltage dropped to 4.4V. I have restarted charging the battery at 0.35A (=0.1C) with a voltage of 13.5V. Internal resistance is down to 40 ohm, good but still a long way to go to <1 ohm...
(prior to restarting the charging cycle I have also shocked the battery 4 times with 32V/1.33A, 1-second pulses)

(edit3) After 4h at 0.35A, the battery's internal resistance has decreased to 0.9/0.077=11.6 ohms.

(edit4) after an additional 10h at 0.15A, battery now is at 10V (open circuit), 8V with 100ohm resistor load; internal resistance is 12.5ohm. Will let discharge again to 4V.

[RealBlackStuff]

Did you see this?
http://www.thinkwiki.org/wiki/BIOS_Upgr ... ad_battery

Don't know if it applies to old(er) machines.

[goldeneagle]

Just found this old thread; forget I replied at one point. I'm going to order 12 x 17500 to try and rebuild a 760 Li-Ion pack. I'll let you know if I'm successful. So far, I've successfully rebuilt 3 x 750 packs, and unsuccessfully a 755CE NiMH (the amount of battery leakage is astounding, so I'll try again when the vinegar and distilled water dries off the flex cable/PCB I'll let air out tonight and tomorrow morning.

[goldeneagle]

Okay, sounds like it comes down to flashing somehow the latest BIOS... Lemme see if I can find a way to force that. I have found some posts about hex-editing the flasher, and a few others about forcing a flash manually.

Good luck on that. I already tried investigating that route and found it won't work on the 755C/CE (noted both since both have two different system program disks).

Amazingly, I did get sound working under Windows 95, but not without wasting about an hour of my time and more reboots than I can count. I'll probably break it installing the wavetable samples (did that for OS/2 and miraculously got it working again, but without wavetable midi). I actually use an MS-DOS vm under Microsoft Virtual PC and the use of a VHD to copy various IBM driver files, extract to a virtual floppy, and then xcopy the contents over to a directory back on the VHD, then after closing the virtual machine, attach the VHD via Computer Management/Disk management, and then copy files from there onto a CF card and then install onto my target machine. In this case, I combined all 3 mwave diskettes into one (you can do the same under OS/2, as the disk labels are contained in a file, so long as you xcopy the directory structure from your floppy image).

And for the record, wavetable MIDI does NOT work under either OS/2 or Windows 95. Proof that mwave is complete garbage, but also garbage that we have to live with, since it's darn near impossible to find a PCMCIA audio card these days.

None of the Ni-Mh I have even lights up the battery charging light on the status bar but the Li-ions do. I think it would be a relatively simple matter of cutting open the plastic case and replacing the internal cells

I have opened a Li-ion pack (85G8520), it has 17500-type cells inside (12 of them, 4P+3S)welded to a custom grid made of thin nickel strips. It would be quite difficult to replace them. Removing the old cells requires cutting them from the grid, which will destroy it, and you'd need to fabricate your own method for connecting the replacement cells. There's very little extra space inside the casing, round wires of an appropriate diameter won't fir so you need to make flat connectors. And that custom harness for 4P+3S would be quite complicated especially since one of the 4P groups is split in 2+2 (the cells inside are arranged in 2 rows of 6 cells).

I'll let you know how replacement works. I'm supposed to get 14 x 17500 cells (order 2 spares just in case), and they arrive this Tuesday. More than likely, I'll carefully remove the existing nickel strip, flatten them out, and just respot-weld to the battery. If that doesn't work, I have nickel strips I can still use. It's been a decade since I've rebuilt a Li-Ion battery (professionally, for a battery rebuilding firm I worked for), but as my other posts illustrate, I've already successfully rebuilt 4 NiMH batteries (including a 755CE battery that gave me about 5 hours of runtime each time I let it run down to or past 0%).

Another option, use this quick charger to revive old/dead batteries: http://www.ebay.com/itm/IBM-ThinkPad-Tr ... 2070332153

https://groups.google.com/forum/#!searc ... YUOyUeetUJ

Good luck. If the cells are truly dead, and you have something like a half a volt between the 14 cells in the battery, you're kinda screwed. I got lucky and have a single 760 Li-Ion battery that holds a charge (around an hour from 100% to machine shut off). That allowed me to FLASH my 760 BIOS before I decided to rebuild my own battery packs. But... the quick charger is AWESOME for reviving newly rebuilt batteries, as the machines usually won't charge them. I used one to charge up my rebuilt pack, and then finished it off in my 755CE when voltage was around 10.5 volts or so.

[stamasd]

Haven't visited this thread for a while...
I was eventually unsuccessful in any of my attempts. I couldn't revive any of the batteries I have, and have given up trying to find a working battery online as all of the ones I found were dead.
I have opened up one of the battery packs and considered if I can replace the cells, eventually I decided that it's beyond what I can do.
So... no sound for me.

The only other avenue left for me to try is to locate the BIOS chip on the motherboard, unsolder it and try to flash it outside the computer with a EEPROM programmer.