Cjshutts

You are probably tired of BS suggestions.

PaulW

I have a rebuilt TC, put in a rebuilt valve body that didn't change a thing, put in a new speed sensor, rebuilt the tranny, have swapped TCU codes with no difference, changed out TCU's (have two, a 98 and 97, no difference), changed out ECU's, no difference.

But the TCU gets data from the ECU, which is the throttle position sensor and idle switch, I can tell you that this problem happens when the idle switch is closed, so it's now time to try setting the switch.

ttuling

If that said Coasting when you in fact had your foot lightly on the gas to go 45 mph on flat ground (with hunting), that would worth investigating. If it switched back and forth that would be even more suspicious. It would be interesting in either case to see what the TCM shows for throttle position, engine load, whatever.

But if you sometimes get hunting under more than very light throttle, you probably won't see any confusion with regard to coasting. And again, if abruptly taking your foot off the gas still allows a drop out and thump, what engine data could possibly mislead the TCM into actually commanding as much clutch unlock as eventually happens to cause the drop out? That it reads the engine as no longer making any power at all, so it may as well unlock the clutch? I don't think that's what the TCM is asking for, but you could confirm by seeing whether N218 current goes to zero (as it would when you step on the brakes) or just to some middling value when the dropout occurs.

Basically, after your discussion a while back about how the lockup clutch works, I won't be surprised if VAG-COM shows no TCM input except engine rpm and no output except N218 current changing during hunting, and it's all just a result of it looking for the elusive desired converter clutch slip rpm.

Tom

ttuling

Any inappropriateness there, like it cutting off the fuel before your foot is all the way off the gas?

Like Tozo, I'm willing to bet $5 that it's something not commonly recognized as failing or wearing out and thus not routinely changed during a valve body or torque converter rebuild. Or maybe the late 1999 valve body design change compensates for something else like a bit of leakage or stickiness in the converter clutch control valves. (Tozo's crucial swap was between 1997 and 2002 valve bodies.)

Sorry to have been talking so long without bringing it to a conclusion. I remember getting excited about engine sensors causing falsely low load readings when rpm falls, but somebody's MAF sensor swap made only a few days' difference. (Wonder what it means that it made any difference at all.) On the other hand, early forum discussions about this stuff weren't even sure it involved the torque converter so there's been some progress. Though there's not enough communication speed for VAG-COM to log everything from the TCM simultaneously, it could still be interesting to see what if anything is changing during hunting. Or maybe you'll find the switch was saying idle when your foot was on the gas, and that will do it.

The people you met once at ZF aren't there any more, or aren't interested in troubleshooting the problem? (The answer is probably already written down somewhere in Germany.)

Tom

tozoM8

It would be nice to move the TPS before the throttle so it would lock the TC before it opens the throttle. The 32V engine is more responsive, than the 40V because of the throttle cable instead of the throttle by wire. I think a post face lift car wouldn't hunt with a 97' VB because of the slower engine response.
Just thinking out loud.

Angus Beef

but have you trouble-shooted the wiring for the G182 and G195 sensors? There is a procedure in bentley I beleive...