I changed my oil recently, and my careful scientific observation led me to discover that it was black and a bit runny. Not clear golden like new oil, but neither tarry sludge like a haven't-serviced-it-for-two-years oil. That's essentially my current limit for evaluating the oil condition.
Having read through plenty forum threads on the Longlife oil and the even longer service intervals recommended by my car's manufacturer, I've gathered that the service interval in the car manual probably isn't absolute truth. Every engine seems to end up in slightly different condition, and of course driving habits and environment vary greatly - to me it makes sense to have an empirical guideline on when to change oil.
Blackstone Labs offer the type of services that I'm looking for, but I'm wanting something I can do at home, and for cheap (at least over repeated tests). A few examples I've thought of:
- Filter to measure mass of foreign particles
- Spectroscopy to measure change of base compounds
- Opacity to also measure change of base compounds
- Phase separation to test for foreign liquids
Though I'm no automotive engineer, so these are all uninformed guesses. Is there such a way to empirically deduce the condition of oil, and thus when to change it?