MC68HC11E1 works correctly only when heated

[mitslo76]

Hi Guys,

While repairing TTI TG1040 RF generator, and I have found very strange fault. The instrument in general was dead, meaning screen was off only back light, but some times worked with some errors on LCD screen (not error codes simply some characters ware bad) the condition that never lasted for more than 5 min and then eventually all display fade out returning to "brick" state. I was able to regularly repeat condition but without any change. I noticed that instrument is actually working and that problem is display and  keyboard. Based on those observations I was able to narrow down problem to control module with  MC68HC11 MCU that works with external 32K flesh and 8K RAM. On first I thought that problem is with flash chip and SW (since the display had errors and behaviour was not stable) but unfortunately all was ok there. As last resort I have tried to re-solder board in order to avoid eventual soldering problems. Curiously enough I have found that there are no soldering problems, but Instrument works fine if I heat-up MC68HC11 with soldering iron. It seams that it has some internal problem which is resolved when heated. I will probably change the MCU and I expect instrument to return to working condition, buy while I am waiting for new chip to come I was wondering if someone had or at least heard for similar problem.  I know that flash chips sometimes can be repaired with heat, but this particular MCU has no internal ROM/Flash...

[Rerouter]

in my head it still sounds like a solder joint or lifted pad, but it is possible a bond wire has snapped inside the package,

[mitslo76]

Unfortunately it is not a soldering joint. I have tried to push the chip back and no change, only work when heated. I did small experiment  punting  300ohm resistor on top of PLCC52 chip package  fixed with electrical tape. When I heat resistor with current after some minutes  instrument starts to work perfectly and work as charm until I switch off current from resistor, then some minutes later it went crazy...  So definitely it is internal bond or some internal transistor problem... 

[nctnico]

Sounds like the chip is bad internally. I have seen similar failures and confirmed them by replacing the chip.

[PA0PBZ]

I did small experiment  punting  300ohm resistor on top of PLCC52 chip package  fixed with electrical tape. When I heat resistor with current after some minutes  instrument starts to work perfectly and work as charm until I switch off current from resistor

Don't switch off the current -> Fixed! 

Does the oscillator stop working? I could be that the Xtal or caps gone bad, however it is not very likely.

[olsenn]

Ah, the good old Motorola 68HC11... the first micro I learned to program in assembly. Those were the days.

[Hardcorefs]

Could be due to SCR latch up of one or more  Cmos input structures being damaged.

for the cost of a new one on Ebay , I would just replace it.......

[lapm]

Ah, the good old Motorola 68HC11... the first micro I learned to program in assembly. Those were the days.

  Me too... Still remember the serial upload system they had to easy prototypeing...

[free_electron]

hold it ! plcc52 in a socket ? check your socket ! if it is a surface mounted socket resolder all the pins.
i've seen a number of times pins pop-off pads. not all plcc's are mechanically identical if there is ever so slightly additional stress sidewards then pins may pop off the pad ....

[Clint]

Yes I have seen sockets with issues. just replace the socket then continuity check all pins with next components inline.

[mitslo76]

Hi Guys,

Thanks for all replays. I already order new chip and will see what happens. Meanwhile will try to re-solder PLCC52 again.

BR