Service Forum

[Wojtek]

Hello there,

First of all I'd like to thank you for making this information public, and for writing the software

I however do have a problem with making it work. Every time (and god knows I've tried sooo many times) I received the Eeprom not available error and a 0 byte dump files gets written.

I've used the simple circuit diagram, 2 resistors and 2 zener diodes.
The laptop I have is a Thinkpad T42 with the amtel chip located underneath the mousepad.

I've redone the circuit twice now. First time I've used wires in between every connection, and when I saw this was going nowhere I redone it and attached components directly.

I don't have a picture of the first one (which was more aesthetically pleasing) as I've recycled some wires from it, but here's some pictures of my current setup.

I know the circuit looks a bit messy, but everything follows the way it should.
Before starting, I modified the circuit in your pdf to make it easier to follow + color coded it to match my wires.



I've ran the itester app with the circuit attached and both values change to 0/1 every 5 seconds. Unfortunately I don't have a voltmeter to test the voltage as suggested in other posts by you.

I must note that if I turn on the laptop with everything attached (ground, sda, scl) it turns on but nothing happends. No screen, no beeps no nothing. If I unplug the circuit, it boots fine. Therefore I'm assuming the connection works fine.

Btw, I've tried different locations for the ground cable. I've attached the cable to the Amtel chip's pin (gently pushed it between the pcb and chip leg) or attached it to some random screw on the laptop.

Also, please pardon the poor soldering. I'm an amateur. First time I've used an iron That's why I did not solder anything to the Amtel chip. I fear that I drop too much solder onto it and mess things up. That's why I purchased that little black thing. (chip socket) and connected SDA and SCL to the connectors on the top.

At first I though maybe my zeners are the wrong way. But after double checking it seems to be okay.

What could I be doing wrong here?

Any help greatly appreciated. Thank You.

[victor]

Hi

I doubt itester showed all ok, because the pins assigned to the DB9 are wrong. The diagram shows the COM port DB9 (male), so you have to replace the wires quite in oposition with the actual setup

Good luck
_________________
Victor Voinea
ALLservice HQ, Romania.

[Wojtek]

Victor, thank you for your reply.

I don't understand how the DB9 can be badly wired.
I've soldered the corresponding numbers exactly as on your diagram.
If you look at the second picture of my DB9, you will notice a faint 9 (upside down, looks like a six) printed on the plastic above pin 9.
I've mislabeled the pins on my picture. My bad. 7 and 8 are reversed. (was 2am, tired ) First DB9 picture is properly numbered.

All pins were numbered on my db9 connector so I used the same numbers on my DB9 as on your diagram.
5 = white/black cabke
4 = white cable
7 = yellow cable
8 = red cable

Are you saying I should reverse them so 1,2,7,8 are soldered?


Also, I'm using a 10ft female/female DB9 cable to connect my computer to my circuit. If I only have that cable connected, without the circuit, itester works as well. Both values change from 0/1 every 5 secs. (same behavior with circuit attached)

Under safe mode however they both change to 1's but then only the second value changes. First one stays the same all the time. (same behavior with circuit attached) I'm thinking its because safe mode didnt load the required drivers for com1 to operate.

Which one should be the good behavior for itester with no circuit attached?

[victor]

You soldered the coresponding numbers to a male connector, not a female DB9, which is different.

No matter, what type of cable you're using, female/female or male/female , the fact is that the pins array is mirrored in your setup. So is wrong.

My advice, get a DB9 female and see the pin numbers. Or better, solder to it, this will eliminate any confusion and wasted time


Itester must show 1-1, 0-0, 1-1,...
_________________
Victor Voinea
ALLservice HQ, Romania.

[Wojtek]

Victor, as stated in my previous post, I got a Female/Female cable.
One end is connected to the computer and the second in a male DB9 adapter.
I must say that the numbering on the Female DB9 *matches* the numbering of the Male DB9 connector. (when plugged in, Pin 5 aligns with pin 5, pin 9 with pin9, and so on)

Nevertheless, I bought another male DB9 adapter and will redo the circuit this evening, this time mirroring the DB9 connections.

I also bought a small voltmeter to test the voltages. But I'm not sure to which position to turn on the meter, and also which points in the circuit to test with the test leads.

Here's a picture of my unit, could you answer the above two questions.


Thanks again for your time. I appreciate it.

-Wojtek

[victor]

Yes, well, turn it left at "20" V- .

Black probe goes to the ground, red probe goes to the SDA or SCL.
When itester shows 1:
SCL = +5.1V
+3.3V <= SDA <= 5.1V

-----------
itester showing 0"

SCL = SDA = -0.6V

Good luck!
_________________
Victor Voinea
ALLservice HQ, Romania.

[Wojtek]

I've redone the circuit and mirrored the DB9 connection.
I then tested with the voltmeter and got the following results:

===========
Black lead = GND
Red lead = SDA

itester 0 = +3.00V
itester 1 = -0.80V
===========

===========
Black lead = GND
Red lead = SCL

itester 0 = +2.97V
itester 1 = -0.67V
===========

If I check the "invert" box in itester, the positive/negative signs are inversed.
But the numbers themselfs don't change. I don't get exactly +3.3 and -0.6


This time I've built the circuit using some long test clips I bought (it takes up my whole floor ). I figured it would be more flexible so I could play around without having to resolder every time. Picture at the bottom.

I've tried reversing the zener diodes but I then got values of +0.75V and -4.60V so they are in the good direction. Reversing the resistors had no effect.

I’ve attached a double switch so I could have SDA and SCL attached to the EPROM at all times and only switch ON the connection when needed.
(The swtich does not make any difference on the voltages read by the voltmeter. Same values with or without)



What did I do wrong this time?
Why my voltages are reversed and the numbers 0.2-0.3V difference?

Thanks again Victor.

-Wojtek

[Wojtek]

Update:

I've tried the circuit on 4 different computer:

PC 1, 2 and 3 get the same voltage results:
(I've only tried the reader program on PC1. Not bothering on pc2 and pc3 because the same voltages were obtained and it didnt work on PC1)

===========
Black lead = GND
Red lead = SDA

itester 0 = +3.00V
itester 1 = -0.80V
===========

===========
Black lead = GND
Red lead = SCL

itester 0 = +2.97V
itester 1 = -0.67V
===========

Just after those computers are powered on, before windows start, all report the positive values. +3.00V for SDA and +2.97V for SCL

I also tried on an old old TP 560.
It does not have windows on it and I don't habt any drives to install it,
but when I was in the BIOS, my voltmeter reported the negative values. -0.80V for SDA and -0.67V for SCL. So I assume if I could run itester, I'd get the same values for positive as above.

So I'm therefore assuming there's something wrong with my circuit for me to not get +3.30V and -0.60V on both SDA and SCL.

Victor, could you please take a look at the last pictures I posted above and tell whats wrong this time.

Thanks again
-Wojtek

[victor]

Too strange..

Remove the diodes and check the voltage without them. When itester shows 1, all must be +8..+12, when shows 0 all must be negative -8...-12.
BTW, I hope you do not invert the signals in itester! invert box must remain unchecked.

If you can't get those from above, the serial port is f.. up

[EDIT]

A fake ground can give this weirdo result as well. Are you sure that cable is good? Why don't you connect the DB9 female directly to the COM1, to eliminate any suspicion. And the 3.3V could also indicate a 3.3V zener, but I repeat, if the ground is fake, all the signals are screwed
_________________
Victor Voinea
ALLservice HQ, Romania.

[Wojtek]

Here are some voltage results from the DB9 connection directly, no circuit attached. (Itester's invert box NOT checked)
PC com1 -> 10ft female/female cable -> male adapter
( i don't have a female adaptor to plug directly into PC com1)

======================

Black lead = GND
Red lead = CTS

itester 0 = +10.08V
itester 1 = -10.34V

======================

Black lead = GND
Red lead = RTS

itester 0 = +10.08V
itester 1 = -10.34V

======================

Black lead = GND
Red lead = DTR

itester 0 = -0.98V
itester 1 = -21.40V

======================

Black lead = DTR
Red lead = CTS

itester 0 = +10.98V
itester 1 = +10.98V

======================

Black lead = DTR
Red lead = RTS

itester 0 = +10.98V
itester 1 = +10.98V

======================

The 10V result seems to be allright. It's in between the 8V and 12V you said in your last reply. Although the +/- is reversed.

Do those numbers tell you anything else?

[Wojtek]

I think I found the problem.
That 10ft female/female serial cable seems to be a crossover cable.

Label on cable's bag says "DB9 Null Modem Cable F/F-10FT"

According to my research on google it appears "null" means crossover...
Last time I take advice from a store employee... Said null was normal cable.

On top of that that store is closed until Monday so tomorrow I can't buy a female adapter to continue this project.

Which pins are crossed?
If I reverse the connection maybe I'll reverse the crossing and make it work as a straight cable?

[Wojtek]

It worked!

That crossover serial cable was indeed the problem.
I cut one end and created a female adapter out of it

Connected that to the pc and my circuit, checked the volatges and this is what I got:

Both SDA and SCL
itester 0 = -0.78V
itester 1 = +4.80V

At that point I was a bit disappointed as I still didn't get the -0.6 and +3.3, but I decided to give it a try anyways.

And it worked!

Password recovered successfully.

Once again, thank you victor for your help here.
Hopefully this thread will help out others who may have a similar problem.

If you're using a DB9 extension, it may be a crossover cable. No good.
I'll know for next time

[victor]

Glad to see that you've done it.

But I am still suspicious about the inverting signals, cause a crossover modem cable can't just do that, normally it will exchange pins 2-3 signals, but of course is not good. Null-modem, yes that will do that all the time.

Anyway, I think I mentioned many times to not use null modem cables.

[amor]

hello. i'm new here.

i've a t40 tp and i'm just brocken the atmel chip out of the MB(MainBoard).

my question is, does the methods (descripted before) work on a 24rf08-chip, which one is not in the MB anymore??

thx

[Wojtek]

Victor, I think that cable I had was inverted in more ways then only pin 1/2

When I was connected to it I tried many different combinations.
And for example, if I reversed the DTR and CTS cables, on SCL I *did* obtain -0.56V and +3.36V (+3.36 is exactly what my Amtel chip was reporting). But SDA reported some weird values like -0.80V and +5.04. I tried that way and the reader still was reporting eeprom not found.

That's as close as I was able to get to the -0.6 and +3.3. I then gave up with the cable, cut out the connector and connected directly into the pc. And it worked right away on the first try Even if the voltages were a bit off.

As for the zeners, I'm almost 100% sure they were 5.1V
I went to the store 3 times, and each time bought 2 of them, asking a different salesman to give them to me (they were behind their counter). Everytime I asked them to make sure if they were in fact 5.1V zeners, and they showed me the label on the box to confirm.

So unless they wrote the wrong label on the box, it seems they were 5.1V

Thankfully all worked fine in the end and I managed to get the password.
Thanks again for this great info!


Amor, from what I understand, you could use the simple circuit, but you'll need to build the battery box as well if the chip is outside the laptop.