Turning key kills power

I haven't figured out why my starboard engine dies after a while but now the port is acting up.

Background (which may or may not hold a clue): When I went to winterize in 2019 it was running fine until, after shutting it off for five minutes I couldn't restart it. Seemed to be a spark issue but didn't get it running again. Not touched since until now. Started up. Was running fine. I was trying to recreate same conditions, i.e. run for a while, let sit, try to restart. After running it for a while and shutting off had trouble getting it to turn over, but would after several attempts, then wouldn't anymore. Discovered that the bolt on the battery (of which you usually on see the threads and used to connect using a wingnut) where you attach the cable had separated from the battery. How or why don't know, was no physical force on it, there was some evidence it had gotten hot but no wholesale melting of the lead. Put in another battery.

Issue: would not turn over. discovered that there was power (12V) to key, but when you turn the key, no power, at the key. I don't mean voltage drop like it was trying to turn over, no power. No noticed sound (e.g. no clicking), meter drops to zero volts (on wire that supplies power to ignition key) when the key is turned on (to on, not start).
Figured I would try disconnecting wires at various to see if I could localize the problem. Tried neutral safety switch and wire to coil, no difference. Then disconnected plug to harness at engine. That killed power to key. Cleaned up plug and reconnected. Engine cranked. (Late in day at that point, packed it in)

Question: could cleaning up the plug have solved the problem or is it likely to be intermittent and it was just a coincidence?

Second question: can anyone explain how turning the key on would kill the power? Turning the key on should just allow the power to continue, right? So even if there was a bad connection or ground somewhere downstream there should still be power at the key, right?

Thank you.

I once had some similar symptoms caused by an intermittent failing battery switch. Thing was, in the "failed" state it still would pass enough current that you could read 12v downstream. But put a load on it (even turning on the ignition which turned on an electric fuel pump), the the voltage after the switch would drop to near zero. And if one turned the switch off and on a few times it might work fine for a day or two. Finally jumped around it with automotive jumper cables and that solved the issue until I could get a replacement.

Thank you for the reply. I cleaned (in 2019) all connections, including the battery switch (which all looked fine before cleaning). More importantly, it was only the port engine. The starboard engine (connected to the same battery switch) works fine (for a while, but that is another thread). Plus other stuff (whose power comes through the switch), e.g. blower and bilge pump work. So don't think it is the switch, but if it happens again will bypass it to be sure.

KAG said:

...

Question: could cleaning up the plug have solved the problem or is it likely to be intermittent and it was just a coincidence?

Second question: can anyone explain how turning the key on would kill the power? Turning the key on should just allow the power to continue, right? So even if there was a bad connection or ground somewhere downstream there should still be power at the key, right?

Thank you.

Click to expand...

if the issue was in the main harness interconnect and you resolved it, the problem may have been solved.

what you were likely seeing is a voltage drop caused by a bad (high resistance) connection...with the KEY=OFF, no current is flowing thru the BAT lead going to the SWITCH so you would likely measure the full battery voltage....now you turn the key, so KEY=RUN....now the BAT lead is connected to the ignition circuit and there is likely some current draw....and that draw was enough so that the voltage drop across the 'bad connection', consumed the battery voltage so the meter read ZERO....

When dealing with suspect 'bad connections', i've found it very beneficial to use an old fashioned test light (with an 1156 or 1157 bulb)...they draw over an amp of current so any suspect connection will keep the bulb from illuminating....the other concept that helps is that you can have the same issues on the ground or return side....in other words, the RED (+) side of the circuit can be just fine but the BLK(-) has the 'bad connection'...

Thank you, makes sense. Hopefully it was just some corrosion in the connection.
Electricity was never my strong suit. I was thinking if there is voltage there is power there, without realizing it could be a minimal amount of power, i.e. just enough for the meter to read. Using a test light rather than (or in addition to) a meter is a good idea.