Spark Angle a big fat zero on the freeway

[mhuppertz]

AZVET suggested that I check my knock sensor count for the stumbling issue, it says 9999, and I just checked live data while going down the freeway and found the spark angle fluctuating all over then it just went to 0 and stayed there until I parked and revved it in neutral.

I guess if I can find the knock sensor count it will tell me something, but if not is it probably the knock sensor? O2 sensor?

I have a new one of each to install just in case...

 

 

Edited May 28, 2018 by mhuppertz (see edit history)

[2seater]

The name "spark angle" is sort of a misnomer. It is the spark retard ordered by the ECM in response to knock. I don't know of a reason a standard naturally aspirated engine should ever indicate spark retard, although it is not impossible. Under ordinary driving, it should never indicate spark retard, that will really cause drivability issues. It is not unusual to get some knock counts, ED17, especially on startup or under some high stress situations, but it takes a lot of knock counts to get the ECM to retard the spark, in other words, some knock counts are expected. The number should not stick in one place and stay there. You can usually trigger the knock sensor to indicate knock counts by rapping sharply on the rear engine lift bracket while idling. You may even be able to get the engine to stumble. If the knock counts don't change after doing this, then I would suspect the knock sensor or connection first, or a flaw in the EPROM. If the knock sensor is tightened excessively, it may become overly sensitive.

 

The ED08 reading is about right for warm idle,  and will likely jump up a few degrees when idling in gear. While driving, the spark timing will vary all over the place and is normal, although it should stay somewhat constant under constant speed and load. You need to watch that ED16 while driving in that zone that is giving you fits. I think Daniel is on the right track.

 

I don't know what the conditions were for the O2 sensor readings. The sensor must warm up before it will become active and a cold start or prolonged slow idle will not keep the sensor hot enough. It is best to watch it while driving for a bit. If the cross counts stay at zero, it's either the sensor or connection. A good active sensor will likely be in the teens for cross counts while cruising, essentially more is generally better.

Edited May 28, 2018 by 2seater (see edit history)

[mhuppertz]

On 5/27/2018 at 6:35 PM, mhuppertz said:

Thank you so much for the info. The O2 sensor was fully warm and I was cruising on the freeway. The knock sensor is another issue. Maybe I will try a tank of premium to see if there is any difference. It is almost undrivable at times on the freeway because the power just cuts out when it tries to upshift at part throttle. I'm guessing it goes open loop at full throttle because it runs smoothly floored...

 

[padgett]

O2 cross counts should be over 6 when cruising. Old PA is the knock counts. There are usually a few on cold start up and may get some on sudden hard accel but no more than 3 digits. Yours is maxed out.  Note: anything that causes vibration in the block from a bad bearing to a bad harmonic balancer to just rapping with a screwdriver can increase the counts. I'd find what is causing that first (knock sensor may be bad).

 

Stark advance is relative to the initial setting (about 8 deg BTDC AFAIR) but should be in teens or 20s in a lean cruise. Is relative to manifold pressure. It helps to be able to look at multiple values.

 

ps Were you cruising over 50 mph ? You are in fourth gear but not lockup (TCC)

 

Edited May 29, 2018 by padgett (see edit history)

[2seater]

3 hours ago, mhuppertz said:

 

Multiple values at the same time are helpful as Padgett mentioned. In the absence of a scanner, you could try a snapshot of the readings when it acts up. It may show where something is amiss, but the update rate isn't that fast and may miss something, but worth a try. Interesting that it seems normal at full throttle. The knock sensor is still active under those conditions and should indicate if you are seeing high knock counts, which will trigger the spark retard.

 

The O2 should certainly be active at cruise and moderate throttle. As far as I know, the O2 will still give actual output voltage under full throttle but is ignored by the ECM and reverts to the base programming values. I do not know if a dead O2 sensor will give the odd behavior. I would think at the least, mileage would be pretty bad. It is never a bad idea to install a new O2, especially if the history of the present one is unknown.

 

I don't have as much experience with the touchscreen, but as pointed out, it looks like not in lockup and in open loop, which would not allow lockup?

Edited May 29, 2018 by 2seater (see edit history)

[2seater]

 I know the knock sensor behavior has been our focus, but just out of curiosity, are the fuel integrator and block learn, ED19 & 20, stuck on 128 by chance? If I read the touchscreen indicator correctly, it was in open loop, warmed up and cruising. Once beyond a certain TPS voltage value, the ECM goes into power enrichment, which may be masking some other issue. What coolant temperature is indicated?

[mhuppertz]

The coolant temp in the gauges section reads normal to high, but the coolant temp in the diags reads very low, like 110.

Is there a separate temp sensor for the ECU?

[2seater]

I do not know if there is a difference between the 89 and 90 regarding temperature sensors, but the '90 only uses one. I do not know if the '89 has selectable F* and C* for the display, but the '90 displays in C* only, which would be about 230*F. Perhaps someone with more experience with the touch screen could enlighten. Any luck on the rest of the readings  or snapshot?

[mhuppertz]

On 5/27/2018 at 6:35 PM, mhuppertz said:

 

Yea, I snapped to the fact that it was probably C after I posted last time.

Interestingly, premium fuel make the car run noticeably better, but it didn't sure the cut out issue. 

Today I ran it up over 100 and when I let off the throttle just a little the engine cut all power. The only way I could get the power back was to floor it or let out most of the way.

Bamboozling...

 

 

[2seater]

It is curious. Likely not fuel delivery since it apparently has no issue with speed or acceleration. We really need actual sensor readings when it acts up, like a snapshot of everything. Short of that, watching each individual ECM data output to see if it suddenly changes in the trouble zone. If it isn't a sensor acting up, then the ECM or Prom are suspect. Without more data, it's just guessing.

[padgett]

Part of the issue is that the different screens were at different times and only the Old PA3 showed 4th gear and closed loop. That 9999 reading is definitely something wrong.

[2seater]

Without data, we are lost. Just for reference (although my chip is not stock), I checked a few readings while driving to my cottage. This is at 65 mph on relatively flat highway:

spark advance: 40*-45*, O2 cross counts: 8-24, Integrator: 124-132, Block learn: 119. I would consider all of that normal. Block learn is a little low so I must have calculated the 24# injectors slightly low, but well within tolerance for the ECM. Everything else boringly normal as well.