kbaer28

I have an 89 305 TBI. It has been in storage for a few years and I decided to get it out and try to get it running. I drained as much gas as I could and added fresh gas. Replaced the fuel filter, cap, and rotor. It fired right up. But even at idle I could hear a slight stumble or miss. When I hook up a timing like the timing retards about 2 degrees when I hear this stumble. I hooked up the scan tool. CTS was reading -30. Replaced it. Took it out for a drive and as soon as its under any load it has a massive stumble. Almost dies but catches itself under any type of load. When in park, I can easily rev it up with no hesitation but still has a slight miss or stumble but doesn't sound too bad.

Initial thought was fuel pump. Hooked up FP gauge. 12 psi all through the rpm range. Did not try fp under load. Checked coil, didn't seem right, replaced it. Didn't fix it. Timing is good but jumps around a few degrees when I hear the miss. Car is stock except for K&N air filter, headers, no cat, no air pump. Car ran fine after these mods years ago.

I have looked all over looking for vacuum leaks, sprayed carb cleaner all over, didn't find anything bad. I am running out of things to check and do not want to start throwing parts at it. Anyone have any ideas?

RFmaster

MAP sensors can go bad due to moisture accumulation, but it is a long shot. If you have a scan tool you should able to see MAP voltage (should be around 1.2 to 1.5 Volts at idle) and increase as load increases. At WOT MAP should read around 4.5 Volts.

GL

//RF

kbaer28

I checked the MAP sensor volts. Around 1.3 at idle. In park, revved up to 4500, floated around 2.5 volts. Should the volts move with the car off like the TPS does?

RFmaster

No exactly - MAP sensor is used to measure intake manifold vacuum levels. At idle intake manifold vacuum is high (stock LO3, LO5) , 20 to 18 in-Hg, which corresponds to low engine load. Conversely, when you put it in gear try to accelerate engine vacuum drops which should correspond to higher MAP voltage. With engine off, ignition off you should see atmospheric pressure (i.e. no vacuum) inside intake manifold and MAP should read 4.3 to 4.6 Vdc. It looks like MAP is reading idle vacuum level just about right, but does it report correct output voltage when engine under load???? Under acceleration ECM delivers fuel mixture based on MAP readings and I would check that to rule MAP out. This is all assuming that Fuel Pump provides adequate pressure and volume of fuel under load conditions.

//RF

kbaer28

When I am driving it, the MAP voltage increases up into the 2-3 range. This is under hard acceleration for a brief amount of time up to 20 MPH or so. That would mean it is working correctly right?

There also does not seem to be any dead spots in TPS. Even when it stumbles bad, the TPS voltage seems to hold constant.

I will check FP with it in gear and brake held to see what comes up. Something tells me it might be the fuel pump but I need to confirm it before I tear it out.

Any other ideas?

kbaer28

FP under load is good. With my gauge I am not able to drive it, but even when I put it in gear and hold the brake the gauge does not fluctuate at all. Steady 12 PSI.

I am running out of ideas..

RFmaster

Do the the engine OFF, ignition ON MAP sensor test. It must read 4.3 to 4.7 Vdc. If it reads low, i.e 2 to 3 volt range this will explain stumble....

//RF

kbaer28

It reads 4.5 volts.

I tested my EGR. Hooked up vac gauge inline with EGR and vac source, revved motor up until it registered vacuum, maintained that rpm, unhooked from EGR, and there was no rpm change. From what I read, this is incorrect. It should increase rpm and move, correct? It does not move at all when I unhook the vac source. And it is very difficult to move it manually. Would a malfunctioning EGR cause my massive, undriveable stumble and backfire shortly after idle?

RFmaster

OK MAP is ruled out...

On EGR...
IRC, depending on calibration, EGR gets activated by ECM when it sees vehicle speed above 2 to 3 mph and TPS above idle threshold. Carbon buildup may preclude EGR from delivering correct amount of exhaust into intake. This would make intake mixture slightly richer, but by not much. You should see a idle stumble when EGR gets manually activated.

Follow-up questions:

At this time there is no stumble while you are revving engine in park?
Does timing still retards when you hear stumble?
KS - connected?
Are there any resonances in the exhaust? Does it get loud at certain RPM??
How much fuel do you have in the tank? If less 1/4 tank there maybe be fuel slosh during acceleration.

//RF

kbaer28

If I push very hard on the EGR it lowers the rpm slightly.

When revved in park it sounds pretty good. Revs up to 4500 easily, doesn't sound bad. At idle you can still hear a slight miss or hiccup every few seconds but it isn't consistent. Sometimes a few seconds apart, sometimes 10 or more seconds apart.

If you drive it on the road and idle around you are fine, but whenever you give it any gas, any at all, it coughs on itself and sounds like its going to die, but catches itself, tries to rev again and coughs again. i have not tried to go over 35 MPH because of stumble, it is undriveable. After it stumbles it will randomly backfire, sometimes a little, sometimes a lot.

The timing does still retard when you hear the little hiccup at idle.

Where is the knock sensor on these? That is one thing I have not troubleshot yet.

Exhaust does not change volume.

Fuel tank is full.

Thanks for your help and keep throwing suggestions my way!

RFmaster

So it would appear that EGR valve does work.

The stumbling part and backfiring may indicate the following:

Lean mixture (fuel pressure somehow dropping under load)
Timing is retarded too much.

When you run up engine from idle to 4000 RPM there is no load on it. Under these conditions ECM will dial in about 30 to 40 deg of SA timing (2500 to 4000 RPM and 30 kPa ). If engine is under load it will drop to about 15 to 28 deg at WOT (90 to 100 kPa), provided that base timing is set correctly and have EST plug connected. Now if somehow base timing was set incorrectly all of this timing will be way off causing poorly running engine.

At idle LO3 will bounce around 18 deg BTDC +/- couple of degrees with EST plug connected and ECM code 43 cleared. Bring engine RPM to about 2000 RPM while observing timing - it should advance to about 28 to 30 deg. If you have a dial - back timing light it makes this job easier. If does not advance - check set EST connector!!!

Did you clear Code 43 after you set base timing??? Base timing on LO3 0 to 4 deg depending on the calibration.

Another possibility - balancer outer ring has slip - rare, but happens. This will give a false timing indication.

You have not touched ignition control module?? No need.

Knock sensor (KS) and ESC module together retard timing in case of pre-ignition. But loud exhaust, resonating headers, broken engine mounts have been known to cause false knock and timing retard.

//RF

kbaer28

I have not cleared any of the codes. I will clear them first thing in the morning. Will the codes mess with the ecm?

I checked the base timing with the est unplugged, its set at 0 degrees. When I hook the est back up the timing does advance and advances more when revved up. I am not sure how many degrees each peak and valley is, but I know that the first valley is 0 degrees and that is where it is set at.

The strange thing is that it ran fine before it was put in storage. That's what makes me leary that it is a timing issue. The timing has never been messed with prior.

When I watch the injectors they seem to be creating a nice cone shape, but I am not familiar with what the cone should exactly be like. I am thinking since it sat to long maybe there is a partial plug in an injector? Again, they do make a nice cone, but maybe it isn't nice enough. I am not familiar with these injectors, is there a screen or any rebuilding them or do you have to buy new?

RFmaster

Injectors may have gummed-up from non use (two years is a long time in storage), but I have seen JY injectors come back to life very quickly. While on this subject Withhunter performance services injectors for a reasonable fee. You'll also need injector pod gasket kit - dctrumpet has a very nice kit. There are coupe Youtube videos - fuel cone should consist of fine fuel mist, no large drops.

To clear codes just disconnect negative battery terminal for about 10 seconds and reconnect with engine off.

At this time I suspect timing, but I will be the first to admit that I hate guessing. I would find a dial back timing light - for example Equus 5568 which allows dialing amount of timing (advanced or retard) - while looking at single timing mark. Makes life a lot easier. Good luck - do not be afraid asking questions.....

//RF

kbaer28

So you think my base timing is still off? Or do you suspect what controls the advance of the timing higher in the rpm's is malfunctioning?

kbaer28

What about the distributor itself?

RFmaster

This is a gray area of troubleshooting since the engine will start and run Ok until you put some load onto it. This why it leads me to believe either there is an issue with the base timing (or something in ignition system) or drop of fuel pressure under load. However, being hundreds or thousand miles away precludes visual troubleshooting. Sometimes a quick look can reveal things that out of place - contributing to your engine performance issues.
1)
I would take a critical look at the balancer - make sure that outside ring has not slipped. This happens after decades of service - rubber simply deteriorates.
2)
Inside distributor take a look at reluctor coil and magnets on the rotor. If you have high humidity at your local reluctor coils windings will oxidize and break down overtime. However, I have seen magnets fall-off from reluctor stator causing ignition issues like you are having. It is a real PITA to get there on third gens! Pull a cap and rotor and take a close look at the ruluctor - if it is rusty looking, or if there are bits found under the hat chances are something went bad.... Take a look at it to begin with....
Over the years we have seen people with strange engine behavior that was tracked down to a faulty dizzy.

//RF

kbaer28

I checked out the dizzy, incredibly rusty, thought for sure that was the problem. Replaced, timed, same exact issue. Absolutely no change. I will take a look at the balancer next.

kbaer28

Got a dial back timing light. With the base timing at 0 degrees, then EST plugged back in, dialed to 30 degrees, revved to 2000, the timing is right back at zero. So I think timing can be eliminated? I messed with the base timing way advanced, some retarded, and everything in between with no success, it just ran worse, but the same stumble when on the gas under load and backfiring.

Thoughts?

RFmaster

This thing is turning into a real back scratcher!!!

Back firing leads me to believe to look into the following things:
lean mixture?
wrong ignition timing - no longer considered.
stuck valve ?

The car sat for "few years" - during this time one or more valves were in the open condition and springs may have compressed fully. During this time of no ops rust may have formed on the valve stem and now valve can not close! This may explain a stumble - one or more intake valves are not fully closed. Run a compression check. If any intake valve is hung-up you'll get 10 to 15 PSI vs 135 to 150 PSI for a good cylinder!!!

Worth a check....

//RF

kbaer28

Hmm.. Could be, I hope not thought.

If a valve was hanging I feel it wouldn't run so well without a load. It's a completely different animal when a load is put on it. And the stumble is so consistent when you accelerate. Its odd. Even when I advanced and retarded the timing the stumble behaved in the same manor at the same time when accelerating hard.

I will check the compression just to rule it out. Might not get to it for a few days. How do you disable the ignition on these when checking compression?

Any other ideas?

RFmaster

Disconnect white plug with large gauge pink wire from ignition coil. This will remove IGN voltage supply from coil and ICM rendering ignition system inoperative.

1) Going down memory lane - I have heard anecdotal stories of fuel lines being pinched off when trany gear selector was moved from park into drive.

2) Just as sanity check - check alternator output voltage while engine in park, and in drive. A healthy alternator should be providing 13.5 to 13.8Volts at idle.

freaky

try unhooking the battery for a few min. computer could have a glitch .
you can take it for a ride with the est wire unhooked an see if it still acts up .that wire takes the computer advance out of the picture . if it runs ok that way you got a computer problem .

kbaer28

Unhooked computer, no change. Unhooked est, same stumble. Alternator output is 14.3 under load and not under load. Will check compression later today.

RFmaster

Thanks for update kbaer. Disconnecting EST disables timing advance by ECM and will set Code 42 - EST. It will make engine run very lazy - fixed timing.

Did some digging over the weekend, but it will have to wait until compression test rules out any internal issues.

//RF

kbaer28

Checked compression, all cylinders 175 psi. Tried new wires and plugs. Same stumble and backfire. Same random miss at idle. If you can get in the upper rpms, 3500ish plus, it smooths out some, but as soon as it shifts the stumble comes back, bad. It seems like it gets a bit worse the warmer the engine gets.

Ideas?

RFmaster

Ideas??? Hmmm...

Do you have a WinALDL and ALDL cable? I guess we have to look and see what ECM sees while stumble is taking place. I suspect that you may have gummed up injectors (from two years in storage). If injectors are not delivering enough fuel ECM will try to compensate by adding fuel (BLM values will be right at the limit 158 to 160 counts). All along O2 values will be well below 300 mV indicating lean condition. Some data logging and analysis.

For all other considerations you have new:

Distributor
spark plug wires
plugs
ignition coil
CTS is new - OEM or aftermarket (make , model)
TBI base gasket
EGR was checked
compression is good
base timing is set.
No codes flashing when ALDL A & B are jumper-ed except for 12?

//RF

kbaer28

I do not have WinALDL but I do have a Genisys evo scan tool. I have done some datalogging with it but when I look at the block learn (is that what you mean by BLM?) it didn't log the data. I will hook it up again and pay close attention to those values. I am way more familiar with OBD2 cars but am trying to learn about the TBI. What exactly does the BLM mean?

I keep thinking its an injector issue too. But they sure seem to be coning right.

The base gasket is not new. Other than that, everything else you listed has been replaced. The CTS is a Napa sensor and according to the datastream on the scan tool it is reading correctly.

I will check the O2 values and BLM tonight, hopefully it uncovers something!

RFmaster

@kbaer

Yes BLM = block learn memory. In OBDII lingo it is refered to as long term fuel trim (LTFT). In GM OBDI BLM base count starts at 128 counts (it is and 8 bit number 0 to 254). BLM of 128 counts is a center of adjustment range. Based on O2 readings ECM will reduce fuel (BLM 108 to 127) if rich or add fuel (BLM: 129 to 150) if lean exhaust is detected. BLM changes are driven by integrator which is an instantaneous fuel trim (STFT) value - thus it takes time for BLM value to change.

EVO is a very nice diagnostic tool - glad that you have a good diagnostic tool to work with! Lets take investigative steps. You may need to log at minimum 5 to 10 minutes of driving while it is misbehaving since these early OBDI systems have very slow data rates (160 baud). Just be very patient with it - ease into the engine RPM area where it is having a problem. A long straightway or a hill may do the trick. Capture the data logs - import them into Excel and start looking for anomalies. You can post your log here as well (use plian text or csv file format).

//RF

Ronny

iTrader: (1)

You say the TPS holds steady in #5.

Did you hook up multimeter with engine off ign on and slowly move throttle to see if there is a spot is does not show consistency in v change?

kbaer28

I will try to get some logging done tonight and get it uploaded.

Ronny - I have not used a multimeter, but I have watched it on the datastream on the scan tool and have not seen any dead spots.

Ronny

iTrader: (1)

It is just a controlled environment in that you can repeat the TPS 0-100% say time and time again and concentrate on the meter readout for any anomalies. You need to solder up a jumper.

RFmaster

Actually Ronny brings up a valid point. The ALDL data stream is a slow sampling of actual data - when you hit a dead spot in the TPS the ECM may not be providing snap shot in its stream packet. IIRC tps uses a flat style connector with large size pins. Just pull harness plug off and measure resistance between pins A & B or B & C - resistance should vary smoothly (1k to 6kOhm or in reverse) The key is a smooth variation in resistance!

//RF

kbaer28

I have never uploaded the logs from this scan tool before. Couldn't figure it out, not certain I have the proper software on the scan tool to allow it. Hopefully someone smarter than I can pull some information off of it.

The TPS really seems to be responding the way it should. I can hold the throttle steady at 40 or 50 mph, TPS holds steady according to the scan tool, and the revs bounce all around. Even with the slow sampling rate you can see the rpms bounce around when the throttle is steady. It was sputtering and backfiring some when cold but the stumble isn't as severe. As soon as it gets warm, 170 or so degrees, thats when the stumble is very strong when under a load and the backfiring is very, very consistent.

RFmaster

Copy data streams use USB stick to transfer from tool internal memory to PC
See page 71, Chapter 16 of the Scan Tool User Guide

http://genisysotc.com/pdfs/555302_re...010_manual.pdf

Interesting that there is no mention how to view datalog files on PC

The first photo is very interesting - it clearly shows that engine is running lean as ECM keeps adding fuel until BLM hits upper limit of 150 O2 is bouncing while BLM climbs from 126 to 150. Then burps O2 voltage drops to 0.18Volts - which causes drop in RPM. Blip in the TPS recovers.
Photo #2 is also interesting, but you can definitely see oddball BLM drops and corresponding O2 voltage movement.

However need to look at the whole datastream.

//RF

kbaer28

I tried the USB transfer, it wasn't reading any of the thumb drives I tried. Called tech support, they said it only supports the old 128 or 256 mb thumb drives. Trying to locate one of them now. Other than that, they said there is no way to export the information. I am going to try some other things to try to get the logs on here for the experts here to dissect.

kbaer28

I can create a custom datastream, what values do you want to see? If I just record the whole thing, it doesn't always grab all of the data for whatever reason.

RFmaster

128 kB USB are relics by now - I may have one laying around.

BAT Voltage
IAC
CTS
IAT
VSS
MAP
RPM
TPS
INT
O2
BLM
Knock counter

//RF

kbaer28

I found an older scan tool that I could export the logs. But they are rather short because it is older. I will try to get some longer ones but maybe you can get some info off of these.

kbaer28

some more

kbaer28

And last

RFmaster

OK I have downloaded all your data files and had a chance to look through the data. The older tool that you have is very sneaky - it fills blocks of 10 identical readings into a data file. There are no time stamps (or engine run time) to go by. I had to import this this data into excel and do some massaging to get something meaningful.

1) IAC - I do not believe that this tool is reporting it correctly. The counts are too high. Pull you EVO and see what it reports at idle - stock engine should be around 10 to 30 counts. Otherwise perform IAC reset. Procedure is posted in the technical articles.

2) MAP voltage is jumping a lot under steady state driving (Log #4). TPS is flat 1.63 to 1.65 while MAP voltage is bouncing between 3.78 and 4.43 Vdc, which is a lot! That should not be happening under steady state throttle on a flat road (IAC is at 128, but I suspect this is incorrectly reported by this tool). All other sensors are rock steady:

CTS 151 to 156F
IAT 68 to 79
O2 is bouncing in a CL (0.24 to 0.92) which is OK
RPM show jumps, but I suspect it is due to MAP providing ECM with inconsistent voltage.

Conclusion(s) - reset IAC, MAP sensor either acting up or actually reporting true values. Bad MAP???? or a vacuum leak in the hose - TB vacuum port.

//RF

kbaer28

Wow, you are definitely the guy to talk to! You know your stuff.

When I put in the new distributor, I knocked off the vac hose to the MAP sensor, it left a code, I didn't see I had knocked it off, and went to get a new MAP sensor. As I was installing it, I saw I knocked it off. But I decided to try the new MAP anyway. It didn't change how it ran at all. So I left the old one installed. Unless I got a bad sensor from NAPA.

When I pulled the logs, I tried to stay, for the most part, in the worst running section. At those RPM's the engine was in a constant stumble. Waa Waa Waa Waa Waa waa waa. Very odd. That is why I suspect the high map readings?

I will double check the IAC readings and go through the reset.

Thanks so much for the help so far.

RFmaster

@kbaer28

I did a bit more number crunching tonight (computers are good for that). Given the two MAP voltage extremes 3.78 to 4.43 Vdc this equates to intake manifold vacuum bouncing between 6.0 and 2.5 in-Hg. While TPS is only 1.63 to 1.65 indicates either a very large load or vacuum leak since the throttles are just opening up. But first things first. Do you really have a bouncy intake vacuum that is driving MAP nuts???

Well, I used this setup in the past




Requires a conventional vacuum gauge duct-taped to the windshield, a vacuum T and some vacuum hose.

MAP port gets signal right from between the two TB barrels. Closest vacuum port that may screw things up is PVC port - did you check your PVC????



Brake booster gets its vacuum through the side wall of intake pedestal on the driver side. Vacuum boosters almost never fail, well almost. But this time I would check its hoses for integrity to rules things out.

//RF

kbaer28

I will be out of town this weekend so I will not be able to work on the car until monday but just to clarify, this is what I should check:

1. IAC Reset
2. Check vacuum going to MAP.
3. Check vacuum to booster
4. Check PCV. The PCV goes into the driver side valve cover correct? I removed it and when I moved it, it clicked, is there any other way to check it, or what should I be looking for.
5. In your pic, you say that section of the TB is prone to leaks, do you suggest putting in a new base gasket? Or just spraying carb cleaner in that area?

I think that's everything?

RFmaster

OK on items 1 to 5. New TB base gasket is a good idea - I use FelPro 60903 base gasket. Look for cracks in TB base or anything that might be loose or look out of place. I had a TB with loose vacuum port, which caused weird idle issues.

Double check IAC - reset it, and then use your EVO scan tool to monitor it. At idle it should be 10 to 30 counts while, MAP voltage should be 1.25 to 1.50 Vdc, TPS should be around 0.65Vdc +/- 0.2 on a stock LO3. Of idle, during steady cruise, IAC should not be running 128 counts all the time. GL.

//RF

kbaer28

Finally got some time to work on it. On the evo the iac is around 180 cold and holds steady at 38the when warm. I am going to put in a new base gasket and see if I can figure out the iac reset. I will report back what I find.

kbaer28

Put in the new TBI tuneup kit. Disassembled and cleaned IAC. It still sits around 40 counts at idle. Never gets between 10-30 like you said it should be. What would this mean?

When you said reset, what did you mean? The idle adjustment? In drive, it idles at around 550-600, do you still want me to go through the reset of the idle? Or what did you mean exactly?

It snowed here, so I am unable to drive it to see if anything I did fixed anything or to test the actual MAP vacuum. I will report back when I am able to drive it.

Schurkey

TBI base gasket varies with TB design. These are the two I know about. If there are others, I haven't seen 'em. Gasket part numbers are in the photos.

Genopsyde

iTrader: (2)

Kinda skimmed through this, but have you ruled out a faulty O2 sensor and/or plugged catalytic converter?

kbaer28

I used the top gasket.

The cat has been removed and O2 appears to be working.