Turning key to "run" causes all power loss.

Those are the two options I know of where 8 gage wires were used.
Here's one version showing the feeds going through a gromment on the firewall.
View attachment 1716301141

The heavy duty wiring versions with rear defrost grid often used the alternator side of the ammeter as a junction post. So another visual clue there.

[URL]https://www.forabodiesonly.com...-rear-window-defroster.558390/post-1974625739[/URL]



With no current flowing there will be no voltage drop.

To do a voltage drop dest we need to turn something on with a known (or measureable) current.
The difference in voltage from the power source (or any point in the flow path) to another point in the flow path will only occur if there is resistance to the flow.

No flow, no resistance.

So if the engine is off, our power source is the battery, and the battery feed has an ammeter on it.

Then if we want to measure voltage drop from the battery to the main splice (standard wiring), we can turn on the headlights - that's about 10-12 amps with the marker lights included. A glance at the ammeter will confirm that.
Then measure the voltage drop from the battery positive to the main splice. The wire from the main splice to the alternator Batt terminal has no current flowing in it so we can use that as an extension of our test probe.

View attachment 1716301154

Anyway the main point is that any voltage drop test must be accompanied by a known or reasonably estimated amount of current flow.
V = I x R

Use this method (voltage drops) to look for resistance that can not be measured be a typical handheld resistance meter. If you can measure it on meter, or see damage visually, then you don't need to do this sort of investigating
I mostly agree, and I hope you don't take this as a criticism. On the contrary, I greatly appreciate your help. It wasn't until I made my original post and received the first round of suggestions that I remembered how to do a proper voltage drop test. :thankyou:

The voltage drop test assumes that resistance increases at the point of failure while under load. The magnitude of the load will proportionally increase the resistance. This is a critical assumption in this test framework. I could, for example, take a 1/8 Watt 220k Ohm resistor and add it in series with my tester probes. I would measure a significant voltage drop, but the current would be miniscule. If I had a megger, I could likely identify the source of failure just like the current-on voltage drop test, which you point out in your last statement (most meters can't do load-based resistance measurements); certainly my Fluke 111 can't. I only mention this in case somebody many months from now searches this forum and comes across this thread.

Regarding your attached photo, while I don't have the grommet on my firewall, it is the connector on the red wire that's touching your thumb that's failed.