brassicGamer wrote:

matze79 wrote:

Any Progress ?

Waiting for parts. 😀 Plus I keep getting distracted by other projects (the list gets longer every day).

Same here, waiting for parts and doing other stuff in the meantime. 😵

Power board was unsaveable unfortunately 🙁

Looks like the electrolyte leakage caused more significant damage than the eye could detect. It was worth a go. I will sell the good parts in the hope that someone else will be able to extend the life of their model.

Good luck with restoration. After recapping and battery change wish you will be able to run the system, POST diagnostic card will be helpful in this case. Did you check power supply already apart the system?

It's back on.

I just purchased another T1600 on eBay for less money than I paid for the last one. I will have to drive 60 miles to pick it up though. Hopefully with my broken one and this 'new' one (also not booting) I will end up with one working laptop.

Hi, how is this restoration going? I ask because I myself just bought a similar T1600/40 with two battery packs and no power adapter.
I got myself some 6.3mm x 3.0mm barrel connectors and wired up a 12V power supply, the power LED lit! So did the left & right battery pack indicators.
The battery pack indicators also shut off if I ejected either battery.
However, when I would press and hold the power button, all the LEDs would momentarily flash, the drives would kick, and then the laptop would die immediately.
I left the power supply connected for a few hours and I guess the batteries got enough charge that the laptop would power up (briefly).
First message: CMOS error, battery dead, press any key. I did this, same error came up again. Did it again, it tested up to 640K RAM, CMOS error, press any key. I pressed that key. Then it came to the HDD selection (so many heads, so many tracks, so many sectors, total size 41.3 MB, press F1 to exit / F5 to save / F10 to apply or something like that). But before I could press anything, the computer died (presumably, so had the batteries).
I tried the power button again, same response as when I had started: all LEDs flashed momentarily, drives kicked, laptop died.
So I left it to charge up a few minutes and then tried again.
This time it fired up the screen again (CMOS error, press any key), but died in seconds.
Anyway, the way it is looking is it needs some life in the battery pack in order to power on - and won't run off of that power connection alone.
Which is kind of a shame, because I distinctly remember running my original T1200 strictly off the power connection for hours at a time & no problems.
So I'm going to pop this open & find a replacement CMOS battery and see what kind of leaky battery juice is inside needing wiping up.
But I'm pleased to report, so far this is looking promising! It might have a faulty power supply or need capacitors, but it seems to be running for almost a minute at a time, and I've only just started messing around with it.

Incidentally: here is the power supply I got, a 12V & 15A unit for LED lights.
https://www.banggood.com/AC110V220V-to-DC12V- … -p-1457071.html

So I took apart the first battery (Toshiba PA8703U).
The case is glued, the way I did it was with a razor knife (box cutter).
Starting very carefully at the front edge, this is the seam that first goes into the battery port on the laptop. I just pressed the blade tip into the seam and cut into it slightly.
The battery case is separated in the middle so I did half the seam, then the other half. I kept going until the blade finally had cut all the way in (I only let it cut in about 1/4"). This is from the middle out towards the two battery contacts.
Then starting at each contact, I did the same thing going out to the back end (the back end is the side that has the eject handle. Both sides, many careful slices with the razor, progressively deeper each time.
Lastly I went along the final edge, inwards from the corners towards the eject handle from both sides, same thing. Progressive cuts with the razor through the glue.
Then for the final step: I pried open at the battery contacts with a small flat blade screwdriver and popped open the case along the front edge until the entire front edge was disengaged.
Then I pried from the contacts back along the sides toward the back end. At this point the procedure is obvious, just keep prying up the edges until the whole top case pops off. As you can see I was able to get into mine without hardly any gouge marks at all. If you gouge: you are pushing Too Hard and you're going to Hurt Yourself! Don't do that!

There are two cell packs, one each side. Interesting square cells. The entire battery is rated at 2,000 mAh and 7.2V.
To get the cells out: I carefully pried up a little bit to get the battery contacts out of their grooves, they aren't glued in there, mine slid out easily.
The battery cells come up with the contact. The grooves are angled slightly so you can't just pry straight up.

The battery contacts are spot-welded to the end cells, and the cells are spot-welded together with similar pieces of metal.
I don't know if these are lithium or NiMH or NiCad or what chemistry they are.
Note: as I was prying the first edge I reached in there deep & made a spark inside the case. Short? Treat these carefully.
From the corrosion you can see these must be pretty much dead but I just pulled them out of the laptop which had been charging them all night.
I mean, I doubt they hold enough charge to run the computer but they're not exactly dead & could be a fire hazard, definitely a chemical hazard.

So there's a black thing wired across the left bank of cells, maybe it's a charge controller. Also note the eject handle popped off, it consists of the plastic handle, a metal hinge rod, and a spring. The spring fits between those two wires onto that little cylinder-shaped boss on the bottom case half. One wire goes to the negative terminal contact of the battery case and the shorter wire goes to that center contact. I have an inkling this is how the charge controller on the computer balances the two halves of the battery.

All six cells are in series so they should be 1.2V each, each bank is therefore 3.6V and the whole battery is 7.2V.
I checked each cell with my voltmeter, they are all coming back 1.394V to 1.397V - pretty consistent!
But there's no way this battery has any useful life left...
Incidentally I tried an 18650 cell into the case, it fits snugly, in fact I might be able to tuck three 18650 cells into each side.
But 18650 cells are lithium chemistry and possibly incompatible with the charge controller.

So I tried turning on the computer with just the power supply and neither battery pack, it exhibits the same behavior. It turns on for a few seconds at most, then shuts itself off.
So definitely it needs servicing on the power supply, remove & replace CMOS and sustain batteries, probably re-capping, who knows.

My 1000-1100-1200's behave the same with a dead battery, and from what I read that's expected behavior, the machine won't run without a battery. Which is kind of a head-scratcher, but that's how Toshiba designed them.

I *think* yours is probably the same, and you'd need batteries, or to figure out how to wire the power *around* the batteries, electrically speaking, which is what some people do with the earlier models.

They are NiCd batteries, and from what I can see online so is the T1600, you want to measure them and see if there's a replacement size that matches. With the ones I have I'm lucky, they're Sub-C size batteries that are still available. I'm not sure if those rectangular packs you have are the batteries or covers for the batteries.

Grunchy wrote on 2019-10-24, 23:32:

Hi, how is this restoration going? I ask because I myself just bought a similar T1600/40 with two battery packs and no power adapter.

I somehow missed your reply. Impressive investigative work with the batteries - I haven't got as far yet, as other projects took over. One such project was a successful restoration of a T1000, which I fully documented on my blog. I have been away from the VOGONS for too long, it seems! All my stuff is in storage at the moment due to some difficulties with accommodation (i.e. I don't have any), but the T1600 WILL live one day...

Merovign wrote on 2019-10-25, 17:50:

My 1000-1100-1200's behave the same with a dead battery, and from what I read that's expected behavior, the machine won't run without a battery. Which is kind of a head-scratcher, but that's how Toshiba designed them.

I *think* yours is probably the same, and you'd need batteries, or to figure out how to wire the power *around* the batteries, electrically speaking, which is what some people do with the earlier models.

They are NiCd batteries, and from what I can see online so is the T1600, you want to measure them and see if there's a replacement size that matches. With the ones I have I'm lucky, they're Sub-C size batteries that are still available. I'm not sure if those rectangular packs you have are the batteries or covers for the batteries.

Yes, the T1000 uses sub-C and I bought a bunch of extras when I got mine working, so hopefully they will also work with the T1600 - thanks for the tip!

I'm restoring a T1600 at the moment that had lots of leaking caps a blown tantalum and, unfortunately, a few blow MOSFETs as well.
I managed to find replacements for all the components but one as it is blown up pretty badly.

This might be a longshot as you stated your projects are in storage but here it goes:
Could you perhaps take a photograph of transistor Q29 on the PSU board and/or post what the chip markings on it are? With the DC barrel jack and powerswtich facing away from you it is located in the top left part of the bottom right quarter on the bottom of the PCB.

FuST wrote on 2020-10-19, 07:00:

This might be a longshot as you stated your projects are in storage but here it goes:

Yes, they are still in storage, sadly, BUT I am planning to go there in the next couple of weeks to grab some stuff so I will let you know the component details ASAP (unless someone else gets there before me of course).

brassicGamer wrote on 2020-10-19, 11:08:

FuST wrote on 2020-10-19, 07:00:

This might be a longshot as you stated your projects are in storage but here it goes:

Yes, they are still in storage, sadly, BUT I am planning to go there in the next couple of weeks to grab some stuff so I will let you know the component details ASAP (unless someone else gets there before me of course).

Hope you both manage to restore to working order. Will be waiting for updates 😀

That'd be awesome! Thanks in advance

I have finally retrieved the item. No good news. But first, I checked out Q29 and it has no markings on it, just a small depression in the middle. So I wired it up to my component tester.

Seemed easier than desoldering and resoldering the component (I do not have the skills or the equipment to deal with SMD components. The following was observed:

So that's not very helpful. I'm not sure what else I can tell you, but the laptop is still in pieces so I could take another look.

Anyway, this laptop had the precise same issues as the last one. It was completely dead and had the corroded standby battery, which had leached through the wires to the power board. Damage appeared to be very limited though (appeared being the operative word). Again, the electrolytic capacitors (except the small capacity ones) had leaked their fluid onto the board, spreading down the surface to one of the ICs. After removing all of them and giving the board a vinegar bath, it appeared much better. I tested each pad for each capacitor to check continuity and ensure I hadn't destroyed any with my soldering (this time round I was more patient and had better equipment). Satisfied all was well, I fitted the new caps and then tested the laptop. Nothing but the blinking red "DC In" light (which implies a supply issue). The supply is a stable 12V so the problem is elsewhere.

I'm not sure what to do next. Short of reverse engineering the entire power board and tracing component-by-component I don't know what else I can do. Although I have just thought of one thing while I was typing the above - although 12V is being output by the power brick, I've just realised that there will obviously be a bunch of large caps in there that will need replacing so I'm gonna try that next.

Anyone else got any ideas?

EDIT: thought I'd add that I did add in temporary replacements for the standby and main batteries while testing. 12V was being received at the battery terminals at the front of the laptop. I have got the power supply open (they hid the screws under melted plastic, which I had to drill out to get to the screws) and it looks super clean. I guess I'll remove and measure the caps anyway, but I'm thinking it's not the PSU.

That's not very helpful indeed but I really appreciate you taking the time. With combined effort we can firgure this neat machine out.

In the mean time I have done a lot of research on the machine and powersupply. Some helpful links:

US patent application for the PSU logic: https://patentimages.storage.googleapis.com/2 … 9/US5383140.pdf
US patent for the battery charge logic: https://patentimages.storage.googleapis.com/f … c/US5300874.pdf
US patent for the battery-powered computer system: https://patentimages.storage.googleapis.com/7 … c/US5553294.pdf

These documents have diagrams of the general layout of the PSU, it's location and connections relative to other hardware and a flowchart hinting at what the internal software is doing.

The whole PSU board is actually quite simple, it's just a load of buck- boost- and buck/boost converters controlled by 2 MCUs
This is what my measurements and research told me (might have missed something due to the board being damaged):
IC12 is a 4-bit MCU, this is the charge controller. It is always on when the brick is plugged in to the 12V jack and controls the rate of charge for the 2 main batteries (battery-L and battery-R) and the RAM backup battery (sub-battery).
IC13 is a 8-bit MCU, this controls the switchmode part of the PSU board. It is supplied regulated 5V DC by IC2, which is switched on by Q30 which is in turn swiched by IC12.
IC5 and IC11 are UPA1600 N-Channel power MOSFET arrays, acting as buffers to drive the push/pull transistors which drive the through-hole power MOSFETS.
IC6 is a hex inverter acting as a buffer for various things.
IC7, IC8 are quad op-amps used for sensing the current for the entire machine, together with the RM01, RM02 and RM03 resistor networks.
IC10 is a dual op-amp, I haven't found what this is used for exactly.
IC3 is a custom Toshiba chip, probably a gate array of some sort. It buffers and amplifies the 20MHz clock signal for IC13.

I'm pretty far along with creating a schematic for this thing but the last part is always the hardest. On my to-do list is:
- Figure out what Q29 is (probably the switching transistor of a buck/boost converter)
- Figure out what kind of diode "9" is. I have it drawn as a shottky but that doesn't make any sense to me.
- Figure out what IC15 is. It's part of the sub-battery charge/discharge circuit but I can't find anything on it.
- Figure out what pin 17 on PJ2 (bus connector) is.
- Map out the sub-battery charge/discharge circuit.

Currently I have about 24 components which still need to be mapped out and then I have to reorder the schematic a bit as it's a mess.

I bought another T1600 from ebay yesterday which should arrive the end of this week or beginning of next week, I'm hoping I can use that to compare some measurements.
I'm lucky enough to have a hot-air soldering station so I'll probably lift Q29 off the board and check with my component tester to verify it's a transistor but that is really the only thing that makes sense to me.

In the mean time I can at least share the partial schematic so it might benefit others:

EDIT: It might be that the collector and emitter of all 2SC3325 NPN transistors should be swapped, that's something I need to verify as well.
As for the power brick, you could try using a standard ATX powersupply capabale of supplying 2.5A and connect that to the barreljack to test.
I'd probably use a current limiting resistor in series so nothing explodes in a dramatic way when there is a short somewhere.

EDIT2: Based on your measurement I would guess that Q29 is in fact a NPN transistor but your component tester is being thrown off by either C6 or C7 (or both) being charged up. Pin 1 on your tester is connected to pin 1 of Q29, which commonly is the transistors base. Pin 2 is connected to what is commonly the collector which means pin 3 would be the emitter.
Following that logic and representing a transistor as a pair of diodes it would match a NPN perfectly as described here: https://learn.sparkfun.com/tutorials/transist … nd-construction

EDIT3: I made an error in the pinout of PJ2, please ignore it for now.

I've finally traced all components on the board and fixed my miscount in the pinout of PJ2.
There might ofcourse be some things that aren't quite right yet but this is the best I could do.
As a bonus I renamed some nets and organised the schematic a bit, it should be much more readable now.

Q29 and diode "9" are still unknowns but my second T1600 is scheduled to be delivered tomorrow so I'm hoping I can make some progress on those in the next days.

Version 0.9 of my scematic is attached, I'm pretty confident I can finish it within the next couple of days and release a version 1.0.

FuST, this is incredible work! Honestly, one of the most niche and impressive efforts of reverse-engineering I have seen on these forums - great job. I'm looking forward to diving into the schematic and testing my board. Hopefully we will solve its mysteries.

brassicGamer wrote on 2020-12-06, 21:13:

FuST, this is incredible work! Honestly, one of the most niche and impressive efforts of reverse-engineering I have seen on these forums - great job. I'm looking forward to diving into the schematic and testing my board. Hopefully we will solve its mysteries.

Thanks a lot and good luck with testing.

I've created a new revision as I have finally figured out the last 2 mystery components.
Some minor mistakes have also been corrected and I moved the sensing voltage deviders to their respective rail. This should make it a more readable.

Using this schematic I managed to fix my second T1600, it had micro-cracks in the traces controlling Q4 and Q7, causing the 5V and 12V rails to ramp up too high. This made the MCU shut down the unit to prevent damage to the computer.
Fixing these traces with some enameled copper wire was the solution. The traces looked fine by eye, even under 10x magnification so a multimiter is required to verify electrical connectivity.

Now on to fixing the first machine, which might be a bridge too far as the PCB is severly damaged by all the corrosion.
Might have to try to make a replacement PSU, though I'm not very confident I've got the required knowledge to do so, especially for the battery charging circuitry.

I have updated the schematic slightly as I noticed some errors while fixing the first machine's PSU board, V1.1 is included as an attachment.

On fixing the other PSU board: I succeeded.
As I couldn't find anything wrong using the equipment I had I finally decided to buy a digital oscilloscope as an early christmas gift to myself.

I would really advise getting or hiring/borrowing a storage oscilloscope (doesn't have to be digital) as the PSU only tries to power everything up for about 100ms before shutting down.
You also have to wait for the 5VB (power to IC13) shutting off before trying again as it seems to hold on to the error state and refuse to retry. I soldered an LED to the output of IC2 for a visual indication.

Using that I found the -9V not switching. Apparently the MOSFET driving it (Q5) had blown again but replacing it didn't fix the issue.
Tracing back the circuit I found diode 8 to be defective. Luckily this type (BAS16) is still available.
After replacing that and resoldering some bodge wires the thing sprung to life!

So, after months of measuring, verifying and ordering a lot of new components and de- and resoldering 90% of the board I now have a working T1600 and a working T1600/40.
Now on to fixing the JVC harddrive that came with the T1600/40, though that might come after some cosmetics on the machines. I might also just end up using an XT-IDE or similar solution.

I am documenting the entire process and will release it to my YouTube channel when I'm done.
I also made a vieo describing how the PSU board actually works which I will release somewhere in the following weeks.

FuST wrote on 2020-12-20, 14:43:

I have updated the schematic slightly as I noticed some errors while fixing the first machine's PSU board, V1.1 is included as a […]
Show full quote

I have updated the schematic slightly as I noticed some errors while fixing the first machine's PSU board, V1.1 is included as an attachment.

I am documenting the entire process and will release it to my YouTube channel when I'm done.
I also made a vieo describing how the PSU board actually works which I will release somewhere in the following weeks.

T1600 PSU Schematic (v1.1).pdf

Wow, that's some amazing work. Thank you for reverse engineering that board for us all.
What where the most common components to fail in your experience? Might help me repairing my two T1600's

You mentioned looking for an XT-IDE solution, I have a Compact Flash adapter up for sale that fits in the expansion slot with XTIDE rom on board. It can also be configured to control the internal IDE controller. https://www.benl.ebay.be/itm/264877993192
If you're interested I'd be happy to send you one for free as a thank you for all the hard work.

Now next up the T1200 and T3100sx. 😉
(They appear to have similar power board issues, recapped, no signs of corrosion, check all vias, no sign of life. Power LED blinks red when trying to power on)