Join Date: Mar 2002
Location: Austin, TX, USA
Re: Purple-Yellow Starter Wires
You are correct about the connection of the purple and yellow wires. The yellow wire provides 12V to breaker point ignitions (bypassing the resistor wire) for an additional spark boost during crank. Non breaker point ignitions (HEI and other electronic) don't need the yellow wire - with the exception of some 1967 vehicles which did not keep the IGN feed hot during crank.
It sounds like you may be describing a "heat soak" problem. The theory is that header heat causes the solendoid to build up high internal resistance preventing it from working properly.
I am a big fan of the "voltage drop" method of finding electrical problems. For example, during crank, the voltage on the battery (+) terminal should be nearly the same as the voltage on the purple wire at the solenoid. And, if the voltage is exactly the same, a voltage reading between those two points (while attempting to crank) should be zero. In reality, you're always going to see a little voltage between those two points, but it should be very small (less than 0.1V). [Actually, the voltage you see between points will vary depending on the load]. If you find the voltage drop is much higher, begin moving your measurement points around the circuit under test until you isolate the component or components where the voltage drop resides.
Starter heat soak theory would indicate that very little voltage drop exists between the battery (+) and purple wire at the starter. Instead the resistance would live within the solenoid coil or between the solenoid's power wire and the motor (accross the contactor terminals internal to the solenoid). To my knowledge no one here has posted any results of such a test and usually jump straight on to heat sheilds or external contactor/relay setups. However it may actually be that the ignition switch itself or the firewall bulkhead connector is heating up while the car is running causing the resistance of a bad connection to increase. The external solenoid / relay work around would eleviate the symptoms but not necessarily remove the root cause. As a result, I personally would use the voltage drop method to make certain the crank issue is in fact a starter heat soak related problem before I spent too much time replacing parts or modifying the crank circuit. It should be a pretty quick and easy thing to check when the starter is acting up. Just realize the voltage drop readings only have meaning when you activate the circuit under test (cranking in this case). The only issue would be avoiding a burn while you try to test the circuit on a hot motor.
68 Coupe, 350 w/ Edelbrock Performer RPM heads, cam, intake, 700R4, Dave's small body HEI