loss of power

[jvandyke]

Hi all. We broke the 924s out of hibernation (briefly) Having done much work this winter and more to come, recent issue:
Start cold, fires right up, idles fine, take off after about 1/4 mile she stumbles, looses all power, drops to idle, you can clutch in, and feather the gas and it sorta responds, after about 30 seconds it clears and off you go, then runs great until it cools down again, repeats.
We have played with coolant bleeding (some serious weirdness going on here).
Suspects:
air in coolant freaking out temp sensor
bleed coolant
DME relay; swapped in another spare, no effect
bleed coolant
smacked DME during event, no effect
bleed coolant
O2 sensor wiring examined
bleed coolant

age old question, fuel or spark, since it's only cold and extremely repeatable i don't think it's something like fuel pump or filter
We drove it 25 miles last Friday and it was fine other than wanting to dump coolant when shut down (erratic coolant levels is another issue)
I run scenarios like "head gasket failure that opens up when just starting to warm up, hydro fouls a cylinder, then clears and heat closes it up and off you go" that kind of thinking costs me lots and lots of time and trouble.
I going to start running sensor checks on the temp sensor, TPS, O2 sensor and so on since I really don't want to make more work for myself. This car is killing me. 2 years and it's time for my kid to enjoy it for a bit.
Compression test may be advisable too.
I want to be systematic and intelligent and diagnose well...for once.
_________________

[Pipemaster]

Well, fuel and spark would be something you should check. If it stays running fine at idle, let it warm up at idle then check everything. Leakdown tests, compression tests, etc. do all of the tests when the problem presents itself

[jhowell371]

Check the fuel injection temp sender in the head. It acts like a cold start valve, enriching the mixture as needed for a cold engine. Maybe yours is going wacky and flooding the engine or causing a lean condition. Any black smoke out the back during that 30 seconds of clearing out? It may be leaning out but I doubt that.

[jvandyke]

DME temp sensor is on the list for testing. No black smoke noticed. It's a cheap enough part (for a change) so can swap but will run the test first. I'm also going to look hard at the TPS since that is also implicated in such conditions.....as is the AFM......

[jhowell371]

Went thru something similar with the wife's VW Cabriolet. It has a thermo-time switch for the enriching valve, I think the 944 ECU/DME compensates thru extra fueling thru the injectors. Swapping a spare or borrowed one might be an easy check. Too bad you're not closer

[Jfrahm]

A bad knock sensor can kill power and mileage.

[jvandyke]

no knock sensor on an '83 motor (although i think there's a plug for one on our harness, not connected, there's an extra 2 pin EV1 sitting there, that's what i figured it was for, a non-existent knock sensor, could be wrong)
to muck things up, we're running an '83 944 engine in our '87 924s (we figured that out the hard way, by ordering parts that didn't fit and THEN checking the engine number, yep, she's had a bit of a jaded past, poor girl, she's looking good now though
would I be correct in saying on our Porsche the "enrichening valve" (which I'm familiar with on earlier Bosch J-tronic systems) is the ICV (idle control valve) which is connected and newish, it too could be the culprit, not sure if Clarks has a test for that guy or not, will go look, or is it just the DME temp sensor that tell the DME to richen things up until warm and the ICV does something else? Anyone remember carburetors..............

[Pipemaster]

Originally Posted by jvandyke

no knock sensor on an '83 motor (although i think there's a plug for one on our harness, not connected, there's an extra 2 pin EV1 sitting there, that's what i figured it was for, a non-existent knock sensor, could be wrong)
to muck things up, we're running an '83 944 engine in our '87 924s (we figured that out the hard way, by ordering parts that didn't fit and THEN checking the engine number, yep, she's had a bit of a jaded past, poor girl, she's looking good now though
would I be correct in saying on our Porsche the "enrichening valve" (which I'm familiar with on earlier Bosch J-tronic systems) is the ICV (idle control valve) which is connected and newish, it too could be the culprit, not sure if Clarks has a test for that guy or not, will go look

It is not the ICV. I replaced mine before i sold my engine. I drove without a working ICV for about 4 months as a daily. It didnt affect anything except for THE IDLE or lackthereof.
But if your dead set on testing it just disconnect it. If theres no change in how the engine runs then its broken. If there is then its not the problem

[jvandyke]

Okay, round of testing:
Temp sensor cold: in spec
Temp sensor hot: in spec
TPS; off idle: good
WOT: pedal would not activate. It grounded against a stop screw under the pedal before it reached WOT, manually moving throttle it tripped the switch
Ran the stop screw a few turns further down (to allow pedal to travel down more) and WOT signal.
Put on fuel pressure gauge; shoots to 29 instantly, hold steady during idle but drops down to 10 on shut down and bleed down pretty quickly BUT we found a pin hole in the fuel hose where it transitions from the filter over to the hard line, that could explain the rapid bleed off
Also running original higher pressure expansion cap, no bubbles in coolant, fans kicked in when temp reached no drama now here's the deal:
During testing, the idle got rough (while still quite cold) just like we were experiencing on the road, she didn't want to rev, fuel pressure was still 29, idle would often fluctuate, up to 2000 then down as I tickled the throttle (by hand in engine bay) then I noticed the throttle piece that the return spring is attached to wasn't retracting fully, sometimes it would pull it back all the way, (and the idle would drop to normal) sometimes not, easy fix; tighten spring a few coils, revs and return to idle were completely stablized. Problem found?? a loose spring? could it be that easy?

[Pipemaster]

Originally Posted by jvandyke

Okay, round of testing:
Temp sensor cold: in spec
Temp sensor hot: in spec
TPS; off idle: good
WOT: pedal would not activate. It grounded against a stop screw under the pedal before it reached WOT, manually moving throttle it tripped the switch
Ran the stop screw a few turns further down (to allow pedal to travel down more) and WOT signal.
Put on fuel pressure gauge; shoots to 29 instantly, hold steady during idle but drops down to 10 on shut down and bleed down pretty quickly BUT we found a pin hole in the fuel hose where it transitions from the filter over to the hard line, that could explain the rapid bleed off
Also running original higher pressure expansion cap, no bubbles in coolant, fans kicked in when temp reached no drama now here's the deal:
During testing, the idle got rough (while still quite cold) just like we were experiencing on the road, she didn't want to rev, fuel pressure was still 29, idle would often fluctuate, up to 2000 then down as I tickled the throttle (by hand in engine bay) then I noticed the throttle piece that the return spring is attached to wasn't retracting fully, sometimes it would pull it back all the way, (and the idle would drop to normal) sometimes not, easy fix; tighten spring a few coils, revs and return to idle were completely stablized. Problem found?? a loose spring? could it be that easy?

I believe that could just be part of your problem. You said it still struggled to rev up. It could be a vacuum leak at this point. But, have you checked the distributor cap for moisture or anything like that? Id really make sure and cross off the easy things first

[jvandyke]

It revved find and idled good after I tightened the spring so I want to road test it but not with a leaky gas line I know that ends in.....
been there