george996

Please take a look at the saw tooth like pattern of the voltages. The peaks and valleys are not rounded as expected. I am trying to diagnose a rough idle and am wondering if any conclusions can be made from the graph.

Regarding the rough idle, it's a long story - I have already checked all the normal things. The previous owner had the DME chipped and I had it put back to stock since it was causing pinging. At the same time I rebuilt the top end so I am struggling finding the root cause and am questioning the DME reprogramming.

1997 993 non-turbo

some addition notes;

runs great above idle
it ran great before doing the work mentioned above including idling
no fault codes
new MAF
new o2 sensors (all 4)
new lifters (checked clearances also)
used Porsche cam timing tools to set timing
175 psi compression on all six cylinders
leakage equal on all six cylinders
new wires, cap, rotor and plugs
checked dual dist, belt
cleaned icv
had fuel injectors cleaned and tested (swapped injectors from B1 to B2, no difference in adaptation)
mixture adaptation TRA1 - 0.09; TRA2 -0.04; FRA1 0.99; FRA2 0.99
checked fuel pressure
checked throttle position sensor
checked temperature sensors
checked crank position sensor with oscilloscope
checked injector coils with oscilloscope
checked ignition coils
double checked plug wire routing
checked carbon canister valve
checked SAI check valve
checked DME relay
the one thing i didn't do was a thorough vacuum test. I felt that with negative adaptation values there is no vacuum leak but it is on my list.

thank you

pp000830

Dirty idle valve or throttle body?

geolab

for info:
1- are the cams original profile ?
2- did you use the floating bolts in picture with the timing tools or just M6 bolts ?

george996

yes, the cams are the original profile

yes, i was able to use 2 M6 bolts on each cam housing. However, i pulled out the engine a second time to double check the cam timing and found that bank 1 was about 1 crankshaft degree retarded behind bank 2. Bank 2 was right on with TDC. I had to advance the crank by 1 degree to install the M6 bolts on bank 1.

Also, when the engine was out the second time i used a 964 rocker to check the timing and there was about 1.1 mm lift at overlap.

I convinced myself that the 1 crank degree difference was insignificant and did not readjust. What is your opinion, is this significant?

asmperformance

The pre cat O2 values look normal. The sharp changes are because of the speed of your log. The post cat voltages are stuck at around .7v . Was this engine fully warmed and in closed loop? What were the fuel trim values?

george996

Thanks for the reply. I will try to take readings at the sensor connection with an oscilloscope to confirm.

Yes, the engine was warm and in closed loop.

TRA1 is - 0.09; TRA2 is -0.04; FRA1 is 0.99; FRA2 is 0.99. By the way, i checked fuel pressure since all adaptation values are reducing fuel and it was correct.

Trim values are about the same as they were before the rebuild. They are also about the same with the new o2 sensors and when i swapped fuel injectors.

geolab

Perfect, half a degree on cam, is negligeable

Steve Weiner-Rennsport Systems

Good advice about cleaning the throttle assembly.

DMF or RS LMF?

Did you replace the rubber sleeves and injector manifolds when the top end was rebuilt?

geolab

This chart is very good
post o2 sensors perfect on new Cat reads between 800mV and 900mV
above 500mV on yours , is very low o2 content, so still very good

.pre-sensors to me are in adaptation mode, and not in the more (heart beat graph) Integrator mode

george996

throttle body was cleaned during rebuild.

It has a DMF and i checked it (rotation back and forth). Also, i had no idle issue before the rebuild with this DMF.

I did not replace the injector manifolds or rubber sleeves but i did inspect them for cracks, etc. Also, i am assuming there are no vacuum leaks because the TRA adaptation values are negative. Do you think this is a valid assumption?

george996

can you please explain more about adaptation mode and integrator mode? What are you seeing in the chart that makes you think that?

I also noticed that FR1 and FR2 mostly fluctuates from 0.96 to 0.99 and rarely 1.0 or above. I thought that FR is supposed to fluctuate around 1.0 and if it cannot then TRA will adapt. Unfortunately i do not remember what FR was doing before having the DME reprogamed back to stock.

some additional information - The idle is smoother when the engine very hot. It is worse about 5 minutes after cold start.

Steve Weiner-Rennsport Systems

I would NOT assume its OK. There are specific tests for the DMF in the factory workshop manuals.

Based on experience, I always replace these items during the rebuild since they deteriorate from heat over time and will leak air. Don't assume there are no vacuum leaks based strictly on short & long term fuel trims; I'd recommend smoke testing to be absolutely sure.

george996

thanks Steve.

I used the factory info for the 996 to check the DMF, attached.

I agree that the only relying on the fuel trim values is not 100% conclusive about having a vacuum leak and will get to it. Actually, i hope i find a leak.

george996

just an update - tonight i tested for vacuum leaks by pressurizing the intake manifold to 2 psig and used an ultrasonic leak detetor i borrowed from work. there were no audible leaks and it held pressure for a minute or so.

I'd like your comment about the bank 1 chain tensioner. It doesn't have as much dampening as bank 2 and i think that is corect becasue it has a larger oil escape passage. Even so, the dampening feels very weak compared to the other one. I was wondering if at idle the cam overruns when when the valves are closing since the valve springs could push the cam in the direction of rotation faster than the chain is pulling it. Greater dampeining would counteract this, if this was actualy occuring. I am questioning this becasue during warmup i think i hear an ocasional rattle but i am not sure if the noise is the chain loosening and then slaping the guide or sometihng in the distributor or alternator.

Steve Weiner-Rennsport Systems

I use a stethoscope to listen at individual points on the engine to troubleshoot various components. You can also use a long screwdriver held against your ear for effect, too. In this case, you can listen to each chain box cover to see which one is noisier or abnormal.

Left side tensioner failures are more common than right side ones due to the CW rotation of the engine.