EEPROM Data Failure

[tjankovs]

The PE+ troubleshooting guide just says to press the reset...but I am beyond that.  Pretty sure I have hardware issue going on here.  The crazy thing is the old EEPROM works when the machine is on for a while/warmed up.

TJ

[Stayouttadabunker]

...The crazy thing is the old EEPROM works when the machine is on for a while/warmed up...

That sounds like a possible cold solder joint issue or a defective chip.
Run a solder gun on the pins from underbeath just enough to reflow.

An 24C04 EEPROM lasts long but doesn't last forever.
Its lifetime expectancy is about a million writes, and that the chip gets written to at about every 100 games played.
I doubt your game has been played that much
It sure sounds more like a defective chip problem more than anything else.
That thing is well...old as dirt? lol   
I'd cut the legs off above the surface of the circuit board, unsolder the pins and solder in a new chip.
UPDATE
I see you've put in socket!
Good job!

Check out this sheet...you'll see the specs on it...>>>

[tjankovs]

Good idea... a cold solder joint would match the symptoms.  I'll explore this more as I have run into a dead end on the replacement EEPROM.  I have not had any luck tracing/finding the where the SDA and SCL lines of the EEPROM goto the main microcontroller.  The micro has no I2C port so did they just bit-bang it? 

The Microchip part does not seem to work in the machine.  I do have a socket now which is nice.  I did find the datasheet on the old XICOR part, only 100k cycles!  I'll attach that plus the microchip part I tried.  They look compatible to me.

TJ

[tjankovs]

Today I got the scope setup.  Trace 1 is SDA and trace 2 SCL.  I tacked on some wires on the processor board just ahead of the 100k series resistor.  The plots show the data to and from the EEPROM when the error is present and a reset is performed.  FYI...holding the reset does nothing, the data transfer occurs on the release of the button.  Also interesting, while in normal play cycling the jackpot reset writes and reads from EEPROM.

The first plot is when the machine is cold and will not reset.  I see attempted data write but no burst read.  Next, I put a hairdryer near the motherboard for a few seconds.  I pressed the reset and it worked!  The second scope shot shows that data and you can see the burst read of the data.

All solder joints for the clock and data on the motherboard were reflowed so I can rule that out.
The 5v looks fine too.

Next up...order the exact EEPROM.

TJ

[lookes]

While you have the sillyscope out, check the 5V at the EEPROM for cleanliness. Maybe replace C1 with a .1uF 50V MLCC type capacitor. Also, I see corrosion in the motherboard socket pins quite often. Make sure they are clean.

[tjankovs]

I received my 'new' XIRCOR X2404P EEPROM and installed it in the socket.  After a single press of the reset to clear the EEPROM error the game came right up!  Every cold boot since then has been successful.  Luckily all the stats from CMOS RAM persisted and should be happily stored in the new EEPROM part.  I am still not sure why the Microchip part did not work but the exact replacement works fine.  Seems like the original part had reached it's lifespan and started having a difficult time reading/writing at cool temperatures.

I'll call this fixed!  Thanks for all the tips and suggestions.  My machine is 100% now...time to win that Royal Flush.


TJ

[Tilt]

Glad you got it running again .  The original EEPROM you received may have been defective?  I've used other EEPROMs in them other than XICOR and it worked fine.