Neighbor's 96 Mustang's battery had been draining if not driven 4-7 days over the past 18 months. I read about others having the same problem. This thread had some good tracing info.
https://www.stangnet.com/mustang-forums/threads/battery-drain-how-to-i-trace.814333/
His car has stock radio and Chapman alarm.
Circumstance:
The battery was dead with car in street so I charged it and pulled into the driveway. The next day the remote wouldn't open the doors - the Chapman alarm's LED indicator was out - the electric locks wouldn't open/close with key off but would lock (not unlock) with key on [locks are tied through the alarm]. There was also a "burned" smell inside the car, that wasn't there the day before.
Initial test/info:
I checked current draw with ammeter in series between disconnected battery cable (either pos or neg is ok) and battery post. This can be hard to check... I have a higher-end Radio Shack meter that will measure 10a (unfused) - neighbor had expensive Fluke that blows a fuse if over 2a, which it did twice when the alarm auto activated/siren chirped. Neighbor read 1.2a a while back - and I saw (before fuse blew) current ramping ~0.05, 0.19, 0.40a.. I wouldn't think 0.40a would drain battery in 4 days, but when I left my dome light on (about 1.2a), my Cavalier would drain after 2-3 days. I estimated 1.0a draw might drain battery in 4 days.
Sidetracked - fix locks, alarm, etc:
Started checking fuses and found #8 inside car blew. This feeds the radio, courtesy light, clock, instrument cluster, and alarm. This is hot w/key off with minimum current needed to power alarm, and store radio stations and clock time.
Found the cause:
Next step was to check current draw after replacing the fuse. When I hooked up the battery, I got a good electric arc/snap (can be normal if charging up capacitors somewhere, but shouldn't be continuous) but then noticed smoke coming from the dashboard over the instrument cluster. Pulled out the cluster and saw a trace on back was burned... no sign of cluster back hitting metal behind it.
Took cluster apart and saw a burned component... a varistor (varies resistance depending on the voltage across it - used for voltage surge protection). There was one on the speedo gauge circuit card and another on the tach (tach varistor looked okay). Resistance of speedo's varistor was 17 ohms (would pull about 2/3 amp) - for tach it was 72K in one direction and infinite in the other. I got new part from DigiKey (arrived in 2 days) for 85 cents - original PN is E3SF-AA TDK, replacement is S10 14K AUTO. which I found on this thread about Ford transmissions (about 1/3 down page).http://www.feoa.net/threads/second-engine-typical-auto-transmission-behavior.83128/page-2
The new varistor measured infinite resistance in both directions on a 1 megohm scale. Replaced it and all works okay now. After fixed, current draw at battery shows 0.08 - 0.10 amp with alarm plugged in, and steady 0.08a with alarm box pulled.
Neighbor had good idea of checking battery voltage (I didn't think of this) - he said the dash gauge would show 8 volts after 4 days when the car wouldn't crank. Now (after fix) voltage has been between 12.88 (after driving/charging) and 12.33 (after sitting 24 hrs) with engine off. I'll repost in a few days to confirm this, but the part was bad, and I'd expect the battery voltage to drop at least 1v in 24 hrs if there was a drain.
Summary:
Got lucky because it would have taken a while to find what was pulling current (pull fuses to see what circuit is pulling excessive current, then remove each component supplied by that fuse to find what's responsible) - was also lucky that the car didn't burn up (that 10a fuse did its job). I didn't get photo of burned trace before repair on back of cluster, but this was a thin copper trace that was acting like a fuse. I soldered on a 30 gauge wire-wrap wire to fix it.
Conclusion:
Ford probably got a marginal run of these varistors and they're failing - maybe not enough to blow fuses, but neighbor's had been borderline for 1.5 years (he'd replaced the cluster years ago after the speedo quit, so maybe same failed back then). I suspect the varistor resistance was lower while it was smoking (it had to pull 10 amps to blow the fuse, and burn the trace on back of cluster... the cluster trace was probably still conducting but broke when I touched it).
Note on repair:
I have a solder iron that removes solder. I didn't have to pull the small PC board from back of speedo (good because I hate pulling needles from gauges) - t his saved a lot of time. Don't use more than 30w solder iron.
https://www.stangnet.com/mustang-forums/threads/battery-drain-how-to-i-trace.814333/
His car has stock radio and Chapman alarm.
Circumstance:
The battery was dead with car in street so I charged it and pulled into the driveway. The next day the remote wouldn't open the doors - the Chapman alarm's LED indicator was out - the electric locks wouldn't open/close with key off but would lock (not unlock) with key on [locks are tied through the alarm]. There was also a "burned" smell inside the car, that wasn't there the day before.
Initial test/info:
I checked current draw with ammeter in series between disconnected battery cable (either pos or neg is ok) and battery post. This can be hard to check... I have a higher-end Radio Shack meter that will measure 10a (unfused) - neighbor had expensive Fluke that blows a fuse if over 2a, which it did twice when the alarm auto activated/siren chirped. Neighbor read 1.2a a while back - and I saw (before fuse blew) current ramping ~0.05, 0.19, 0.40a.. I wouldn't think 0.40a would drain battery in 4 days, but when I left my dome light on (about 1.2a), my Cavalier would drain after 2-3 days. I estimated 1.0a draw might drain battery in 4 days.
Sidetracked - fix locks, alarm, etc:
Started checking fuses and found #8 inside car blew. This feeds the radio, courtesy light, clock, instrument cluster, and alarm. This is hot w/key off with minimum current needed to power alarm, and store radio stations and clock time.
Found the cause:
Next step was to check current draw after replacing the fuse. When I hooked up the battery, I got a good electric arc/snap (can be normal if charging up capacitors somewhere, but shouldn't be continuous) but then noticed smoke coming from the dashboard over the instrument cluster. Pulled out the cluster and saw a trace on back was burned... no sign of cluster back hitting metal behind it.
Took cluster apart and saw a burned component... a varistor (varies resistance depending on the voltage across it - used for voltage surge protection). There was one on the speedo gauge circuit card and another on the tach (tach varistor looked okay). Resistance of speedo's varistor was 17 ohms (would pull about 2/3 amp) - for tach it was 72K in one direction and infinite in the other. I got new part from DigiKey (arrived in 2 days) for 85 cents - original PN is E3SF-AA TDK, replacement is S10 14K AUTO. which I found on this thread about Ford transmissions (about 1/3 down page).http://www.feoa.net/threads/second-engine-typical-auto-transmission-behavior.83128/page-2
The new varistor measured infinite resistance in both directions on a 1 megohm scale. Replaced it and all works okay now. After fixed, current draw at battery shows 0.08 - 0.10 amp with alarm plugged in, and steady 0.08a with alarm box pulled.
Neighbor had good idea of checking battery voltage (I didn't think of this) - he said the dash gauge would show 8 volts after 4 days when the car wouldn't crank. Now (after fix) voltage has been between 12.88 (after driving/charging) and 12.33 (after sitting 24 hrs) with engine off. I'll repost in a few days to confirm this, but the part was bad, and I'd expect the battery voltage to drop at least 1v in 24 hrs if there was a drain.
Summary:
Got lucky because it would have taken a while to find what was pulling current (pull fuses to see what circuit is pulling excessive current, then remove each component supplied by that fuse to find what's responsible) - was also lucky that the car didn't burn up (that 10a fuse did its job). I didn't get photo of burned trace before repair on back of cluster, but this was a thin copper trace that was acting like a fuse. I soldered on a 30 gauge wire-wrap wire to fix it.
Conclusion:
Ford probably got a marginal run of these varistors and they're failing - maybe not enough to blow fuses, but neighbor's had been borderline for 1.5 years (he'd replaced the cluster years ago after the speedo quit, so maybe same failed back then). I suspect the varistor resistance was lower while it was smoking (it had to pull 10 amps to blow the fuse, and burn the trace on back of cluster... the cluster trace was probably still conducting but broke when I touched it).
Note on repair:
I have a solder iron that removes solder. I didn't have to pull the small PC board from back of speedo (good because I hate pulling needles from gauges) - t his saved a lot of time. Don't use more than 30w solder iron.