Alternator Bench Testing of Diodes

[CycloSteve]

I am in the midst of an upgrade on my son's 1.8T B6 150A alternator - going the route of the Valeo B8 180A version with the B7 Brush/Regulator.

I received a used/low mile alternator from a wrecker/eBay, and testing the diodes shows that the "B post" to ground flows as it should (one way only)...but when testing individual diodes in the rectifier one side of the alternator is passing current both ways (six out of the 12 diodes). I then tested the one I pulled out of the car, and all of the diodes pass current both ways, but the overall "B post" test is also fine. Both alternators test fine at the local NAPA, and the car had no issues with charging or discharging. In the midst of a T-Belt and hose change, so I can't do the AC test at the battery until that is done (waiting on multiple parts from ECS and FCP).

Any thoughts on why, or is this just how Valeo alternators work? The last thing I want to do is have to replace the alternator shortly after install, or to damage the car with AC bleeding into the system.

[kapilamuni]

Anyway, there shouldn't be a flow from B+ to ground. That will drain the battery within a short period.

[Kevin C]

Testing the diodes on the bench with them still connected is not a trivial task unless you have a tool specifically designed to test them. For sure you can test the output diodes for a short but the other ends are all still hooked up through the windings. There are testers that can do this, they take advantage of the inductance of the windings to allow the diodes to be tested in place.

As you noted, the battery terminal should block signifigant current one way and allow it the other. and of course it should block flow to ground ( mentioned in the previous post). Most times its just a polarity of the meter and the diodes are not in backwards.

I would suggest that the best test is a bench test with a scope to check for ripple.

Info on typical voltage regulator for our charging system:
https://www.bosch-semiconductors.com...tronics/cr719/

Nice writeup on diode testing.
https://www.fluke.com/en-us/learn/bl...to-test-diodes

Testing Diodes that are still connected:

https://pscorp.ph/wp-content/uploads...ench-Tests.pdf

[CycloSteve]

Good articles Kevin, thanks.

I have three DVM, and two have the Diode testing capability. The newest tester says everything is GTG, so I am going with that coupled with the load test at NAPA coming back perfect. Will install and test in the car once the T-Belt service is done.