Mitchell Gant
Active member
I've recently got a Stargate with a CR2032 lithium battery conversion on the PCB. Battery was flat as machine had been non working for several years from previous owner.
I replaced battery, reset factory defaults, powered off/on and all checked out OK.
Several days later machine reported bad settings. Checked new CR2032 and it read just over 1 volt. Not good for a new battery.
With battery removed I checked resistance from battery +ve to -ve. Very low reading of about 500 ohms showed why battery had drained so quickly.
CMOS RAM power is fed from two diodes D9 and D10, one from battery, one from game +5V. Desoldered one end of each diode and each checked out OK.
Removed CMOS RAM thinking that would be the most obvious fault, but no, still got a low resistance reading.
Not much left in circuit now, so lifted one end of decoupling capacitor next to CMOS RAM (either C3 or C85 on schematics). That read about 1.5K ohms, so bad cap was the problem. I guess the heat from the iron had changed the actual resistance as it had gone up from the 500R I initially saw.
Replaced capacitor, tested RAM Vcc to 0V resistance which was now in the megohm range, so problem solved.
If it had been any other of the decoupling caps across the PCB 5V, then unless it went very low resistance it would not have caused any noticeable fault, but across CMOS RAM battery it obviously drained it quite quickly!
I replaced battery, reset factory defaults, powered off/on and all checked out OK.
Several days later machine reported bad settings. Checked new CR2032 and it read just over 1 volt. Not good for a new battery.
With battery removed I checked resistance from battery +ve to -ve. Very low reading of about 500 ohms showed why battery had drained so quickly.
CMOS RAM power is fed from two diodes D9 and D10, one from battery, one from game +5V. Desoldered one end of each diode and each checked out OK.
Removed CMOS RAM thinking that would be the most obvious fault, but no, still got a low resistance reading.
Not much left in circuit now, so lifted one end of decoupling capacitor next to CMOS RAM (either C3 or C85 on schematics). That read about 1.5K ohms, so bad cap was the problem. I guess the heat from the iron had changed the actual resistance as it had gone up from the 500R I initially saw.
Replaced capacitor, tested RAM Vcc to 0V resistance which was now in the megohm range, so problem solved.
If it had been any other of the decoupling caps across the PCB 5V, then unless it went very low resistance it would not have caused any noticeable fault, but across CMOS RAM battery it obviously drained it quite quickly!
