PhLaXuS

I've done a bit of searching here and found a lot of good info on Closed Loop, but some posts contradict others...

Engine Info:
---------------
·1991 305 TBI
·Stock for the most part
·New Bosch O2 ~6 months ago
·New Knock Sensor ~6 months ago
·8746 ECM w/ANTT PROM (I believe)

Basically, my car runs best until it goes Closed Loop. Whether I've let it sit overnight, for an hour, a few minutes, etc...it always runs best when first started. I key on and the engine is very smooth and has much more power than normal. Typically (due to my neighborhood's layout), the ECM goes into Closed Loop while sitting at a red light. Out of nowhere, the engine gets very lopey and loses a noticable amount of power when I take off from the light. The coolant temps stay right around 180* when up to temp on all except for the hottest of days. The engine gets ~16MPG w/2.73 rear gears. The only code I get is 43 if the coolant temps gets >230*. I get this code about once a month during the summer. I've not gone into it really deeply, but all the scanning I've seen looks OK to me, but I'm not all that experienced in reading it.

So...if anyone has any guesses on the above issue, feel free to reply, but what I really need is a list of the sensors used in Open Loop ONLY. I suspect possibly the MAP sensor, but I don't have known good logs to compare mine to.

I've attached a WinALDL log as a .BMP file. Rename it to "Log.rar" and extract with WinRAR if you feel like taking a look. Turn off all word-wrapping and it should fill out quite nicely as I edited the WinALDL header to make it easier to read. The log is my drive in to work. I make stops at two STOP signs and two red lights before getting on the turnpike for the bulk of the log. Note that this was before installing the manual fan switch, so when I'm stuck at a stop, the engine gets a bit warmer than it does now.

Vader

You don't need a scanner to check the MAP.

"I don't need a scanner to test the MAP."

You can simply use a voltmeter to test the MAP voltage.

" I can use a voltmeter to test the MAP voltage. "

You need to focus on the cause, and not the symptom.

I should focus on the cause, not the symptom."


The data presented by the scanner will only be the scanner's interpretation of the ECM's interpretation of what the wiring harness is translating of what signal the MAP is generating. That's a lot of translation. It's like reading the seventh edit of a book. You may miss the author's meaning entirely.

Test the MAP directly, and cut out all the signal conditioning. With a vacuum pump and voltmeter, you should get something like this:



Lacking a vacuum pump, you can use a vacuum gauge to measure idle vacuum and check thw output voltage.

You can also do the same thing with the TPS, CTS, IAT/MAT, and O˛ sensors, although the TPS, CTS, and IAT/MAT are active at startup as well as in Closed Loop.

The one sensor that becomes a major part of operation in Closed Loop Mode is the oxygen sensor. Prior to achieving Closed Loop, the ECM effectively ignores the O˛ except to monitor crosscount rate to determine when it is active enough to support Closed Loop Mode.

If the O˛ is weak, or if the sensor is getting an artificially lean gas mixture due to leaks, you'll get really fat mixtures, poor mileage, and poor performance.

If the sensor is old or contaminated, it may be producing a weak signal. You can remove and test the sensor output by connecting a voltmeter to the sensor, then placinng the sensor tip in the flame of a propane torch. A propane flame is nearly devoid of oxygen, and should allow the sensor to produce a signal of almost 1.00 VDC. If the sensor does not produce at least an 850mV signal, then fall off to below at least 250mV within a few seconds of the sensor being removed from the flame, it is either too slow and/or weak.

Additionally, if the A.I.R diverter/port valve array is not shutting off manifold air completely once the ECM hits closed loop, you'll get an artificially lean O˛ indication. You can test that by simply removing an AIR hose from a check valve and testing for air pressure from the hose while the engine is running in Closed Loop Mode.

A sticking EGR can do similar things, although is usually accompanied by poor idle. However, that should affect both open and closed loop operation, and shouldn't simply appear as soon as the ECM achieves Closed Loop Mode.

However, a sticking EGR solenoid can cause poor running just like a stuck/leaking EGR, except when at closed throttle. Since the EGR vacuum source is ported, the EGR will close at closed throttle regardless of the solenoid being stuck. Testing the solenoid with a vacuum pump/gauge is also very easy. Again, that should occur in both Open and Closed Loop Modes, but if the EGR solenoid is sluggish or partially sticking, the symptoms may not occur until the ECM hits Closed Loop and starts to duty cycle the EGR solenoid for the first time. Those are FUN to diagnose (NOT!).

Check those items, and report the results, please.

I'll check those things and report back."


You may go about your business...

PhLaXuS

I love the matter-of-fact method in which you post.

I do have a hand-operated vacuum pump with gauge and a DMM so I will definitely give this a test when time permits. I'm not the type to go replacing everything hoping I correct the fault...I just never knew the proper values for really any of the sensors. With WinALDL, I was able to corroborate a few of the sensor data as being accurate for the conditions, but I understand that it is MUCH better to test directly.

Earlier today I found the following info posted by rgarcia63, but it is stated to only be accurate for up to the 1990 model year. Is this data good for a 1991 or do you have that?

The CTS (and MAT) sensor resistance can be compared to factory specs as follows:
All 1981-1990
13.5K ohm @ 20°F
7.5K ohm @ 40°F
3.4K ohm @ 70°F
1.8K ohm @ 100°F
480 ohm @ 160°F
185 ohm @ 210°F

Also, I have tested the TPS (and replaced it some time ago) as the former one was defective. As stated above, I'm running a fairly new O2, but I've heard nothing good about Bosch. I have a propane torch here and will give it a test as well when time permits.

My testing has shown that the EGR system is operating correctly unless there is some intermittent problem which only shows up while in motion. I removed the AIR system and blocked the manifold holes with brass plugs.

Vader

AlkyIROC

I had/have a similar problem with my daily driver truck. Runs great when cold but as soon as it goes into closed loop, it has a wierd flat spot. At idle there was a slight intermittent miss which felt like an ignition miss. Under power there was no problems. Accellerating up to speed was great but as soon as you started to coast to maintain a speed, the engine would have a flat spot as you gave it a little bit of gas to maintain the speed. If you pushed the pedal more, it would accellerate with no problems. Trying to maintain a speed, you could feel a surging.

I changed the O2 sensor, ECM coolant sensor and IAC. I reinstalled the factory FPR and ECM chip (calpac). Nothing improved. Finally I bought 2 new injectors. I've been wanting to change them for years. My truck is now very drivable again. There's still a very slight hesitation in the morning when I go to work but driving around town all weekend, I never noticed any hesition. The occasional miss at idle is also gone.

Was this the fault? I don't know. Because there's still a slight hesitation when I go to work in the morning, I'd say no but the new injectors sure made a big improvement.

PhLaXuS

My problem appears similar:

·Rough idle (as explained above)
·Accelerating seems good
·Maintaining high speed is OK, but the engine seems to "pulse" a little if I'm cruising at 35MPH.

Along with all the sensor tests I will perform, is there a method to test the TB injectors other than a simple cc/min reading? This would not test spray pattern. I could try a timing light aimed at the injectors -- I'd expect it to illuminate the pattern fairly well, but aside from these...any other test ideas?

AlkyIROC

Just bringing this back from the dead to find out if there's been any improvement.

Today it was snowing and very cold. I was struggling to get home today stuck in traffic and the engine was acting up again the same as it was before. Anything under 35 mph was a constant surge

I replaced the MAP sensor when I got home and reinstalled the factory air cleaner with the stove pipe to help keep the fuel from freezing. I'll find out tomorrow if it helps.

Vader

If you don't mind staring at an annoying light for a couple of days commuting, disconnect the vacuum line to the EGR and plug it (so it doesn't leak). See if that doesn't make a difference. You can clear the code and light easily later.

AlkyIROC

OK. I never got a chance to try it going to work but using a different MAP sensor had no change.

My AIR pump is long gone and the holes are plugged in the header tubes. I can reach into the EGR and move it freely so it's not seized. My EGR valve in the intake is controlled by an electrically controlled switch. In the old days it was a temperature switch in the manifold that controlled how much vaccum was sent to the EGR. I pulled the 4 wire connector off it before coming home. No fault light and there's still a hesitation/surge when at a cruising speed under 35 mph. Not as bad as it was before I changed the injectors but still enough to be annoying.

My cruise control also hasn't worked in years. I tryed a different controller on it during the summer and it still doesn't work. It's also all computer controlled. No vacuum ball etc. I checked to see if the controller was receiving the signals from the switches and it was so I suspect it's not getting a speed signal from the ECM.

This is all on a 91 454SS truck but the GM TBI system is pretty much all the same depending what year you look at.

AlkyIROC

Pulled back from the dead again for one more update

I think I found the problem in my truck. I was checking the O2 sensor. It's hard to get at because it's on the top of a merge box and the truck is lowered. Anyway, I noticed that it wasn't pointing straight in. Somehow I must have screwed it in slightly sideways, crossthreading it. I had to retap the threads, real fun, then I put the same O2 sensor back in.

No hesitations of any kind since then. The real test will be when it hits -20c again. My guess is that it either had a poor ground, which I doubt, or that it might have been sucking air past the sensor creating a leaner mixture which would have caused the stumble but I never had any fault codes.

At least I'm not fighting that under 35 mph, cruise rpm or off idle stumble any more. I hope it's finally fixed for good.

Vader

Seems like you may have discovered the air leak/dilution at the O˛. We'll cross our fingers.

That is a strange one though.

PhLaXuS

Well...it's been a while, but I finally got to testing some components today (...and hey!, only nine weeks late, huh?).

Here's the MAP info:

125 Ohms from sensor ground to battery negative terminal. +4.92V reference voltage at sensor.

4.0" = 4.26V
6.0" = 3.89V
9.0" = 3.32V
12.0" = 2.75V
14.5" = 2.28V
17.5" = 1.75V
21.0" = 1.09V
23.5" = 0.64V
26.5" = 0.10V
28.0" = 38mV

The sensor was holding vacuum indefinitely, but no matter how quickly I pumped, I couldn't get more than 28" vacuum. It's probably a limitation of the pump itself. I'm operating off of memory here, but last summer when I was troubleshooting an EGR problem, I believe at idle I had about 17".

These MAP numbers look OK at low manifold vacuums, but appear to get progressively worse at higher vacuum. Anyone think this is a problem which is resulting in my low mileage? The surging problem I have is at low RPMs and higher loads -- such as cruising at 35MPH.

I also checked the CTS & MAT sensors:

CTS:
3k @ 70* F
225 Ohms @ 180* F
172 Ohms @ 220* F

These numbers are always a bit low. They DO resemble the readings for a 1990 style sensor. Does anyone know if the 1990 and the 1991 CTS sensors are physically interchangeable? Maybe someone in the past purchased the incorrect one?

I tapped into the line feeding from the sensor back to the ECM and monitored voltage at the sensor "output" feeding to the ECM when the key was on:

3.75V @ 70* F
2V @ 120* F
1V @ 180* F
0.735V @ 220* F

I'm guessing that there is another resistor in the ECM completing a simple divider network. From these numbers, it's probably a 1k.

MAT:
These readings seem to track the CTS fairly closely, except that when I first measured the MAT, it read something really low such as 140 Ohms when it was first pulled form the vehicle. I touched the tip (before measuring) and it was about body temperature. I let it sit for a minute and then the readings looked normal (~3k).

And a quick question on the injectors themselves. I pulled the injector harnesses and measured the resistance across each injector. I initally got 1.5 Ohms/each. I measured right after letting the coolant reach 220*. I measured over a minute or so as the injectors cooled off a bit and the resistance on each rose to 1.7 Ohms. I believe this is correct for the 8746 TBI ECM, but would just like to verify this.

There was also a small film of fuel around the throttle body at the base of the pressure regulator as though the base gasket may have been leaking. It was by no means a puddle of fuel -- just a small film. Could this be overspray from the injectors or is the gasket leaking?

I didn't check the TPS this time around, but when I replaced it maybe a year ago, it was operating properly. Also, any time I datalogged, the TPS % looked about right. Also, the MAT looked OK in past datalogs, so it's either flaky or my meter probes didn't make good contact in my measurements above. At this point, I'm discounting the MAT. I will retest again next time I'm under the hood.

Vader

The MAP numbers are right in the ball park. FWIW, If you could make 29" Hg with a hand vacuum pump, you'd be doing really well.

The raw CTS numbers also look close enough. I'm not sure what the voltage divider uses for signal conditioning in the ECM I/O, but a 1K makes sense.

The initial low MAT seems reasonable, since heat soak may have made it quite warm. The final numbers you show appear to be within reason.

Altogether, the readings make sense, given that you are measuring the reference voltage a little low. That could be due to a slight meter calibration issue or a VR in the ECM that's "losing it", or a marginal ECM reference ground. Theoretically, if you increase the reference voltage by 2% (or calibrate the meter 2% higher) the readings would be very close to the theoretical tables. Also, we know the MAP and CTS are not linear, so some variation is expected.

The fuel you are detecting around the FPR is probably leakage (TBI 220 unit?). There would have to be some serious reversion to get injector tip fuel back up to the pod cover. Leakage could be the regulator diaphragm/gasket itself, of the injector pod cover gasket.

Speaking of the FPR, have you tested the standing and running fuel pressures yet?

It may still be good to temporarily cap the EGR vacuum line in case the solenoid is leaking internally. Then again, with a handheld vacuum pump, you should be able to test that by seeing if the solenoid holds vacuum without leaking through.