Here's the scenario ... the engine started hesitating/stumbling on the freeway and especially under load, but only after a hot soak. The car then acted as though it was driving through deep sand, but only intermittently. I replaced all the critical ignition components (plugs, wires, ignition module) and the MAP sensor, all to no effect. I also found and replaced a faulty injector that was causing a fuel trim problem (I have a real-time diagnostic monitor), but none of this stopped the intermittent dragging under load after a hot soak.

The fuel trim indicates the engine is running rich when it stumbles, but the computer is able to compensate so it doesn't throw a code 45.

I hooked up a vacuum gauge and this read fine at idle (18.5" that quickly drops to 10" and just as quickly recovers to 18.5" when I goose the throttle). However, the problem doesn't show up at idle. I took it for a road test and found the vacuum was normally 14 to 18", but dropped to around 2" when it started misbehaving. However, it would quickly recover to 18" when I took my foot off the pedal. I checked all the vacuum lines with a MityVac and couldn't find any leaks, but I did find that with the canister purge solenoid open it's impossible to retain any sort of vacuum. I traced the cause of this to the charcoal canister itself and replaced it thinking it was faulty, but the new one acts the same. I don't understand how this system can ever have worked unless the air flow is limited by the bore of the line so I capped the throttle side off at the valve. For good measure I also capped the cruise control line. With these two lines capped the engine doesn't misbehave as badly as before and the vacuum now drops to only 6" and still recovers to 18" when I take my foot off. Again, it idles perfectly.

It's difficult to diagnose when this only happens under load after a hot soak. What is likely to be causing an intermittent vacuum loss after hot soak? Any ideas? Also how does the EVAP system not normally cause a huge vacuum leak?

Thanks for reading this.

 

Tape a fuel pressure gauge to the windshield & drive it. Sometimes the regulator will leak & then stop. Check the vacuum hose for gas inside.
A sticky egr valve will cause a rough idle.

 

Thanks for your reply.

I thought about the fuel pressure. The hose on my gauge isn't long enough to reach the windshield so I'd need to rig an extension.

In the driveway, the fuel pressure seems fine. Actually it's 1 psi over spec (48 psi KOEO) and the regulator works fine to drop the pressure at idle.

It idles like a kitten so it's hard to imagine the EGR valve is sticking.

 

With the vacuum lines to the canister purge solenoid and cruise control disconnected and plugged I took the car on a 40 mile drive today. It only stumbled once, at the very end.

Manifold vacuum was about 15-18" on the freeway and as much as 22" coasting downhill with the throttle closed. When I open the throttle the vacuum drops and the EGR and IAC counts both go up. I didn't record the exact EGR and IAC numbers but they're roughly proportional to throttle angle. Going up a moderate incline with a throttle angle of 10% vacuum drops to a steady 15". At a throttle angle of 15% vacuum is about 10".

The stumble occurred when going up a short steep hill near home. Manifold vacuum dropped to about 4" even though the throttle wasn't open much more than 15%. Maybe 20%, but I wasn't looking.

Thus, the stumble appears to be coincident with very low engine vacuum, 4" or less.

Unfortunately, I've never driven with a vacuum gauge hooked up when the engine was behaving well so I don't know what these numbers ought to be. Do these numbers sound reasonable?