odd problem, need some help

yes, it relates to a boxster, but its the same 5.2.2 dme as early 996s and I figured you guys might have some ideas. here goes: So i have a 99 boxster with the 5.2.2 dme. the car has a 3.2l box s engine with a 987 airbox and maf sampling tube. fairly standard swap. has headers, 996 cable throttle body, and underdrives. I had softronic do the engine tuning. So for approximately 30 minutes after key on, the engine runs fine, revs to 7300rpm redline with no issues. After 30 minutes though the car will run fine but it hits what feels like a rev limiter at 6200rpm. The car will not rev through it, wot or not. It doesn't feel like a miss, just a hard limit. If you pull over and restart, you get another 30 minutes before this reoccurs. I asked softronic about this and they said the dme programming is not the issue. i have checked the maf wiring and obviously the maf is plugged in. There doesn't seem to be much else they are interested in suggesting. Any things to check or input would be great. Thanks very much

a passenger using a scan tool looking at data while it wont rev could help .....crank sensor ? it does not seem to be fuel pump issue ......i suspect the tune after checking what you can

Total guess: you have a component (some engine sensor monitored by the DME) that has an intermittent failure whose failure mode is triggered by heat soak. As you drive along, and things start to warm up, all is good, but then things get too warm, and the component fails, sending a signal to the DME, which then drops into a closed loop mode which has a RPM limiter. I haven't used the Durametric, but on other cars, the tuning software I've used lets you monitor the sensors individually and see what they are doing. In your case, debugging could end up being a pain in the neck since failure seems to happen after 30 min, so you could end up needing lots of testing cycles. Again, total guess, but I've seen this kind of thing before, where static testing and even running testing checks out, but heat causes the failure. They're a pain in the neck to diagnose, but usually easy to fix once you find them (swap out the sensor).

Quotes source posts: 14547853, 14547875

I was thinking crank sensor also. I have a spare I can try, but would also think this would throw a code of some sort. I was thinking along these lines as well, but if its a heatsoak issue, why would it be cured by a restart, as the temp would not have been reduced at all. I should also note that I have no reported or pending codes from the DME. Car is a manual transmission. Thanks for the help thus far.

the no cel makes me suspect the tune

I tend to agree, but Softronic is convinced they are not the problem, which isn't so helpful. I have a durametric to datalog, so I can try that.

Quotes source post: 14547942

You could have an intermittent failure that triggers the limp mode. The sensor may not continue to fail, imagine a switch that goes off then back on. The part where it is off may trigger the DME, but then it is back working, so when you restart the car, the DME scans the sensors and thinks everything is fine, so you get to go again until you get the intermittent failure pops up again. Sure, maybe not super likely, but I wouldn't discount the possibility.

Its definitely a possibility. The engine has the following sensors: Maf which includes intake temp Coolant temp oil level-irrelevant knock sensors crank sensor cam sensors Thats really it. I will go over all of them and try to datalog.

throttle position sensor ? and look at o2 sensors when acting up.....see if you are running lean

Flash back to stock and see if it continues.

Quotes source posts: 14548282, 14548330

good call, with try to log all that stuff is durametric is up to it. not an option as the base program is for a 2.5L engine, not the 3.2 thats currently in there. thanks all

Quotes source post: 14548351

Flash the 1999 3.4l 996 program

My bet would be the throttle position sensor .Not sure how Bosch does it but on a lot of cars a restart resets tps .

Boxter and 996tt share the same throttle body, my tt exp similar issues. Remove the throttle body clean the butterfly, pop off the black plastic cover and you will see the brush that travels across the board much like a slot car on a track. You will notice extensive carbon build up. Get some tb cleaner and wipe this area down. Reinstall and see where you are at. *also a dirty air filter can make a car exhibit these symptoms. You running a oiled filter by chance?

Quotes source posts: 14549710, 14549725

Good to know, thanks. My throttle body is cable operated so no tracks or anything. Car has a brand new paper filter. Thanks for the input guys, gonna go through the basics now.

Whoa-a Cable throttle body has to have a throttle position sensor .I think Atrox was describing how to clean it .That is a basic .

Quotes source post: 14549819

Yes it has a tps, but there is nothing to clean on the sensor. he is referring to the electronic throttle body. I replaced the tps with a spare and cleaned the cps. will drive it in a bit and see.

Well i swapped tps and thought i had it with an intermittent crank position sensor. no dice. cleaned all sensor connections. maf reporting normally. Here is a video. 1 min and a restart later and it revs clean You Tube

Did you log the MAF voltage and compare between good and bad runs? The MAF values in kg/hr could be substituted values. Only the voltage is the raw measurements in Durametric. I still suspect your MAF and tune.

Can you compare throttle position voltage or % at wot after restart to a failed occurrence ? A marginal wire connection or bad ground could act like this .The big clue is that a restart clears the condition -whatever is wrong gets reset . Someone who knows the dme might have more insight .That is a good question for Softronics .

One more thought.......have you checked fuel pressure and flow ? This is a stretch but if the fuel pump starts pumping air it would diminish performance -also a faulty pressure regulator .I recently replaced my pump on the volume test you could see a combination of air and fuel -new pump all fuel .

Quotes source posts: 14554542, 14555136, 14555164

I tend to agree that its some maf to dme glitch. i did not test raw voltage. what sort of result would indicate a problem? Throttle position was the same before and after. no change there. The fuel pressure regulator is mechanical so i dont believe that is an issue. fuel pump flowing air is interesting, but it seems odd that it would happen precisely after 30 min and would always happen at the same rpm regardless of load. you would think at easy throttle the fuel demands would be much lower. The repeatability of this makes me go back to my dme glitch assertion. Thanks guys for all the great suggestions

Actually with return fuel system the pump is constant -the only variable is the return .I may be way off but I think you could have a fuel delivery issue -starting with the fuel filter and working back .

Quotes source post: 14555185

I have seen cases where the MAF signal has drops out (goes to zero intermittently). To see that, you have to log with the highest sampling rate with only a couple of variables selected such as rpm and MAF voltage (so not to degrade the resolution).

Quotes source post: 14555543

thats true, but the mechanical fuel pressure regulator would control that on a vac signal, which I checked. Fuel filter has maybe 2k miles on it. I hear where you are coming from, but fuel filters don't reset after a restart. If the fuel pump was flowing air, why wouldn't it do so for the first 30 min? What about variocam? would a bad actuator cause something like this without a code? Cam deviation at idle when hot is like 1 and 2.3 or so.

^^^ as said the sampling rate of the monitor is key....some sensors need a lab scope on them to really really read them for a glitch [my snap on scope has a glitch mode just for this]. i suggested looking at 02 sensors/injector on time to rule out fuel when the event occurs .

Maf voltage approx 3.9 at 6200rpm, no fluctuation or change when issue occurs voltage above 13.6 at the obd port tps remains steady when this occurs 02 sensor readings were inconclusive Occurs in neutral at a stop as well as in gear Occurs at damn near, if not exactly, 30 minutes after start ignition timing does not change. at part throttle its about 30 degrees, closer to WOT its more like 20 degrees ECT was between 196 to 208 degrees rpm reading does not disappear during the issue Maf, TPS, and CPS replaced with new or known working maybe its just me but I'm going with Occams razor here....gotta be the tune/dme

Agreed.​

TPS steady -is the voltage the same between ok running and the event ? I assume you have all the evap stuff connected ? Will the Durametric report fuel tank pressure ? I still think you have a fuel delivery issue .Here is why :The repeatability of the failure .The fuel pump runs at one speed -how the car is driven has no effect .For the DME to fail at the same time interval is dubious -

Quotes source post: 14560607

i wanted him to monitor the 02 sensors to check for a fuel problem [lean] he should have seen the ecm trying to get fuel in the motor by increasing injector time on,backing off the timing etc AND a cell but i think we both agree a couple more hours checking the fuel system is called for before yanking the ecm out .maybe a fuel pressure tester in line while road testing could help but it will not show volume,remember these cars pop a cel for a loose gas cap so i think the ecm should see a tank venting problem [key word should].

Quotes source posts: 14560607, 14560675

I can only read throttle angle and that did not change. Evap stuff is all connected. Fuel delivery. Pump is driven 100%. Mechanical fuel pressure regulator. Injectors controlled by ECU. As far as I'm concerned, the only way fuel is not being delivered is if the ECU tells the injectors not to open. The DME is not failing. The DME is being told by the tune to do something at the 30 min mark. Thats my guess anyway. Fuel pressure is mechanically regulated though. I'm reading narrowband 02 sensor values through an admittedly slow-refresh-rate durametric. Its not the ideal scenario, I agree. Fwiw, the car feels/sounds like some of these cars on launch control: You Tube

sup

curious-did you fix it ?

No, i took it to my last track event where it ran fine (sessions were under 30 min so no probs there) and niw its in storage. my indy so far has not made any progress and ultimately we have to call softronic. im 100% convinced this is some sort of checksum error in the dme programming

thanks