How to determine the accurate model of MCU

[daisizhou]

Hallo every one.

I am repairing a motherboard,Unfortunately, the three main MCU models on the motherboard have been removed,Through the remaining traces, I vaguely see the INTEL logo
U13, I saw the XX53 Intel 80 logo
Additionally, I measured the VCC and GND pin definitions of three MCUs(The other definitions of pins are incorrect, I borrowed the definition from the 8088 package)
I would greatly appreciate it if you could tell me the exact model number. Thank you

[wasedadoc]

8086 ?

[daisizhou]

According to the pin definition of 8086, it seems to be inaccurate

[Kim Christensen]

Not sure if this is the right manual for this. No detailed schematics but it might help you somehow.

Additionally, I measured the VCC and GND pin definitions of three MCUs(The other definitions of pins are incorrect, I borrowed the definition from the 8088 package)

Did you measure these using diode check (beeper)? If so, they may not be 100% correct. Measure on ohms range and make sure each pin measures less than 1 to VCC or GND.

[fzabkar]

The Keyboard/Display PCB assembly (A1A2A1) contains à 16-position keyboard, five 7-segment LED displays and associated keyboard and display circuitry. It is controlled by a dedicated peripheral on the System Controller PCB via a ribbon cable and transition connector Jl.

I'm guessing U13 could be a PIA, eg something like an 8255A (unless there is an additional controller on the keyboard PCB), or an 8042, etc.

Intel databooks of that era:

https://mirrorservice.org/sites/www.bitsavers.org/components/intel/_dataBooks/

I would determine which pins connect to the LEDs. They will probably be driven from an 8-bit port, or perhaps two 4-bit ports.

The keyboard's 4 rows and 4 columns would be similarly connected, I would think.

[daisizhou]

I repeated the measurements and confirmed that the resistance from VCC and GND to MCU pins is less than 1 ohm

[daisizhou]

Give me some time to disassemble the keyboard and test the definition of the chip U13 connected to the keyboard. The key layout is indeed 4 rows and 4 columns

[daisizhou]

I have drawn the schematic diagram of U13 pin, but it is not complete. There is also one ULN2013A chip that has not been drawn yet

[fzabkar]

U13 could be an 8279 / 8279-5.

https://mirrorservice.org/sites/www.bitsavers.org/components/intel/_dataBooks/1987_Intel_Microprocessor_and_Peripheral_Handbook_Volume_2.pdf (page 352)

http://www.elektronikjk.pl/elementy_czynne/IC/8279.pdf

"The Intel 8279 is a general purpose programmable keyboard and display I/O interface device designed for use with Intel microprocessors."

Where does the crystal go? Does it connect to one of the MCUs, or does it go to a clock generator?

What is the Signetics IC at U8? Could U10 be an I/O expander, something like an 8243?

Where do the DIP switches go? I expect they would go to an 8-bit MCU port.

Where do the EPROM data and address buses go?

The manual states that there are two 8-bit ports (LED and Function) on the Module Bus at J12. I expect that each would be buffered by an octal buffer/latch/transceiver, and these would connect to 8-bit ports on a controller, or perhaps they would connect to the data bus of the master CPU.

[daisizhou]

I drew all the leads of the U13 chip pins.Unfortunately, part of the U13 leads are connected to U12, U10, and U9 have also been erased from their marks，From a functional perspective, U13 is only used for keyboard control and LED display.
8279 seems that the pin definition and GND definition of VCC are not consistent with the actual circuit.
U8 model is 74LS154N。DIP switch, I am looking for a circuit

In order to see clearly, I took a clearer picture again

https://i.postimg.cc/ydNGRrBV/P30423-071816.jpg

[fzabkar]

I would concentrate on U15, U16, U21 and U22 ... and the crystal.

Edit:

Could pin #1 of U13 be shorted to ground via a fault condition? The unused row inputs are pulled high.

Edit 2:

Could U12 be an 8155H or 8156H 2Kbit SRAM with I/O ports and timer?

https://pdf.dzsc.com/815/8155H.pdf

Edit 3:

The PROG pin on the 8243 I/O expander is a clock input, so it would match an 8279 at U13.

https://www.cpu-galaxy.at/cpu/ram%20rom%20eprom/other_intel_chips/other_intel-Dateien/8243_Datasheet.pdf

Using the 8279 as a reference, pin #9 of U9 would be IRQ or INTR. U9 would be the main CPU (master).

[daisizhou]

Insert pins 1, 2, 5, and 6 of U13 into the sketch I drew

I updated some pins for U12, but the measurement was difficult.

I am still updating the route, hoping to find clues

[daisizhou]

I have updated some of the circuits and found that the crystal is connected to U19, which seems to be a "switch" function.
It's also the first time I've encountered such a crystal oscillator circuit
I will continue to draw the crystal oscillation circuit, which may take some time

[daisizhou]

https://i.postimg.cc/mDQGHfj0/P30423-212525.jpg

I used a magnifying glass to observe the remaining traces of U10 and found?? 827? Or?? 62? number
I am sending pictures hoping to provide clues

[fzabkar]

It looks like I was wrong about the 8243 (I forgot that the chip was upside down). That said, I can't find any Intel chip of that period which matches the numbers you have identified.

As for the clock circuit, Google offers several different examples:

https://www.google.com/search?q=crystal+clock+"74LS04"&tbm=isch

Do pins 10 and 11 of U13 connect to U9 (pins 32 and 31, respectively), and does pin 10 of U13 connect to the EEPROM OE* pins?

https://deramp.com/downloads/mfe_archive/050-Component%20Specifications/Hitachi/Hitachi%20IC%20Memory%20Data%20Book.pdf (page 319)

If I were reverse engineering this circuit, I would start with the known components and work backwards. For example, we know the pinout of the EPROMs, so we would trace the EPROM pins back to the main CPU. We also know the pinout of the 74LS154 4:16 decoder, so we could trace the inputs back to the same CPU. The latter would probably be connected to the address bus. I expect that at least some of the outputs of the 74LS154 would the CEs for the EPROMs.


BTW, I found this site:

https://www.ebme.co.uk/forums/ubbthreads.php/ubb/printthread/Board/30/main/11064/type/thread

[daisizhou]

I am referring to the three DIP40 pin chips produced by INTEL（U13、U12、U9）,However, this U10, also known as 8243, does not seem to be produced by Intel

Next, I will focus on drawing the crystal oscillator circuit and the EEPROM input circuit,74LS154 to MCU partial circuit

[fzabkar]

FYI, I have discovered that all the EPROM dumps have a checksum byte at offset 0xFF0. This is computed by summing the EPROM contents in the range 0 - 0xFEF. I don't know how to validate the last 15 bytes, though.

[daisizhou]

The A0-A11 and O0-O7 of the six EEPROM chips are all short circuited together. I tried to draw the connections between U1 and U2, and the others are the same connections.
Unfortunately, the U7 has also been polished out of its model chip.
I have taken a video of normal use and hope it can be helpful
https://youtu.be/Blup_MmSBbY

"EPROM dump has a checksum byte at offset 0xFF0"This may be a zero position.
I am still working hard to find the crystal oscillator circuit

[Kim Christensen]

The A0-A11 and O0-O7 of the six EEPROM chips are all short circuited together.

Hopefully you mean A0 on one EEPROM goes to A0 on all the other EEPROMs; A1 on one EEPROM goes to A1 on all the other EEPROMs; A2 on one EEPROM goes to A2 on all the other EEPROMs etc...

So, now find out which of the 40 pin ICs are also connected to A8-A11 of the EEPROMs. That will most likely be the CPU...
It is highly likely that A0-A7 and D0-D7 are multiplexed onto the same pins on the CPU via something like a 74LS374, so A0-A7 of the EEPROMs may not be connected directly to the CPU.

[fzabkar]

It is highly likely that A0-A7 and D0-D7 are multiplexed onto the same pins on the CPU via something like a 74LS374, so A0-A7 of the EEPROMs may not be connected directly to the CPU.

That appears to be the function of U7. I have tried to trace the address and data pins (on the underside of the PCB), but some disappear into thruholes.

This is my guess at the pinout of U7.

Code: [Select]

     1    20  Vcc
A7   2    19  A6
O7   3    18  O6
O5   4    17  O4
A5   5    16  A4
A3   6    15  A2
O3   7    14  O2
O1   8    13  O0
A1   9    12  A0
GND 10    11

[Kim Christensen]

This is my guess at the pinout of U7:

That looks a lot like a 74LS374 or 74LS373...
If so, pin 1 would be tied low, and pin 11 would be connected to the ALE (Address Latch Enable) pin of the CPU.

[daisizhou]

yes.What I mean is that the A0 pin of U1 EEPROM is directly connected to the A0 pin of U2 EEPROM,Also directly connected to A0 of U3, U4, U5, and U6,The remaining A1-A7 connections are the same.In the end, they all connected to U7 (The model has been removed).
The same goes for O0-O7,Finally, they are also connected to U7.
U1-U6 EEPROM CE pins are independently connected to U8 (74LS154N).The OE/VPP pins of U1 EEPROM are connected to the 32 pins of U9

Please refer to the file I drew for details

thanks

[daisizhou]

https://i.postimg.cc/02Yj6MzL/P30424-210538.jpg
https://i.postimg.cc/1zt3YvvQ/P30424-210557.jpg


I took clear pictures, hoping it would be helpful

Yes, drawing is very difficult, many wiring runs through repeatedly, and some are even lost in the circuit board

thanks

[Psi]

I have not read the entire thread but if you put some isopropyl alcohol on the chip it may make it easer to read.

Another option, if you can find a camera without any optical filter on it, take 3 photos one illumined by white light, one with UV light and one with IR light. Try a few angles of the light as well.
But using all methods combined you can often figure out the text.
Of course this depends if there is some information left there to see. if it's totally gone then this is probably not worth the time.

[fzabkar]

This is my guess at the pinout of U7:

That looks a lot like a 74LS374 or 74LS373...
If so, pin 1 would be tied low, and pin 11 would be connected to the ALE (Address Latch Enable) pin of the CPU.

I wonder why the manufacturer chose to scrub the markings off this jelly-bean chip?