89 trans am gta 350 tpi. stock car 155k miles. only modifications i can think of is radiator fans have been rewired to turn on by manual switch. this was done long before problem occurred. also recently got ebay intake piping and k&n filter since the stock air duct after the maf sensor got a hole in it.

How accurate is this test for MAF sensor, unplug it and if runs better then its your MAF? my original maf sensor broke, so i ordered a like new but used Delphi AF10320 from amazon. i regret ordering a used one now.
when its plugged in on start up it will idle rich and misfire sometimes running terribly smoke everywhere. when its unplugged it seems to run almost perfect. other then a misfire i think once in a while.
there is no codes unless the maf gets unplugged of course.

other things i checked and for sure are working.
1 No air leaks. checked and vacuum tested.
2 electrical faults to maf sensor, maf relay, burnoff relay, IAC, coolant temp sensor, TPS, fuel injectors, o2 sensor
3 Fuel injectors themselves, fuel filter
4 fuel pump and pressure regulator, also fuel pressure is good.
5 spark plugs, wires, cap, rotor, ignition coil
6 coolant temp sensor
7 TPS
8 IAC
9.emission equipment. A.I.R system, EGR, vapor canister. PCV valves. Catalytic conerter temp and back pressure. oxygen sensor new
10 Alternator, battery charge.
11 engine compression all 8 were even and just above 150psi i believe it was.

the only things i feel like would matter at all that i haven't checked is sticking valves which i heard can be a possibility of a car that has barely been driven for 5 years. even though the compression check was good....and engine timing, because my balancer doesnt have a timing mark! so it could be distributor or the ignition control module i havent checked.
can the base engine timing change by itself with like wear of components or something? or only if distributor is messed with?
but regardless of all this crap i checked. the car still seems to run great with maf sensor unplugged anyways. so it could just be the maf sensor.
Thank you!

 

is there any experts at reading ECM data? i finally logged my data and all i can see wrong is blm and int is low and o2 sensor bouncing up to .900s, just confirming the rich condition.
i have a tunerpro Log for
1. car cold open loop to warm closed loop, with the maf unplugged. cant do this with maf plugged in car will run BAD.
2. closed looped idle with maf unplugged.
3. when car is turned off and plug maf back in closed loop warmed up.
4. car is turned off reset codes/ecm, turned back on closed loop warmed up with MAF plugged in. no codes.

 

is there any experts at reading ECM data? i finally logged my data and all i can see wrong is blm and int is low and o2 sensor bouncing up to .900s, just confirming the rich condition.
i have a tunerpro Log for
1. car cold open loop to warm closed loop, with the maf unplugged. cant do this with maf plugged in car will run BAD.
2. closed looped idle with maf unplugged.
3. when car is turned off and plug maf back in closed loop warmed up.
4. car is turned off reset codes/ecm, turned back on closed loop warmed up with MAF plugged in. no codes.

 

I don't have TunerPro loaded up with the $6E defs at the moment but........

What is the idle MAF grams per second? Should be around 5-6 on a stock 5.7. General rule is 1 gram per second, per liter of displacement.

BLM and INT being low means it pulling out fuel. This could be due to the MAF data, the O2 data, or in the case of adding fuel, a misfire will drive the O2 lean.....

Changing the intake ducting is a no-no with a MAF system. It can dramatically alter the MAF scaling. Especially at idle.

GD

 

The MAF elimination test is a good one. It tells you that everything but the MAF is working well enough for the engine to run well without it. With no MAF sensor, the engine runs in default fuel based on a speed and TPS map stored in the MEMCAL. If it runs significantly worse with the MAF connected, you have to focus on the MAF sensor and circuits. What does ECM data tell you about MAF value? Is it unusually high? I can't remember right now if the 89 MAF is a signal generator that produces a frequency, displayed in Hz or if it produces voltage ranging from .5 to 4.5v. If it is a signal generator, it is very susceptible to RFI from leaky plug wires or other sources. Be sure your harness is routed correctly. I would strongly suggest you try a different MAF sensor.

 

The 89 is a hot-wire MAF that's 0-5v, not a frequency MAF. Those came later.

GD

 

thank you for replies already.

i had no idea it was so sensitive the ducting i bought is very similar to stock however at least after the maf sensor. and i didnt really have a choice no one seemed to sell a used stock duct when i tried buying one. also this issue occured before i got the new duct, after the problem started and removing the stock duct a few times it got a hole in it where the it clamps to throttle.
https://www.ebay.com/itm/RED-COLD-AI...53.m2749.l2649

i will attempt to post screenshots of the tunerpro data.

 

No maf, code 33 set. after 5 minutes of idle, closed loop. also with no maf the o2 reading bounces around alot more. from low .90s all the way to 900s sometimes. maf reading sits in 6.xx but unpluges anyways.



maf in, no codes ecm reset. car already warm, screenshot is after 6 minutes of idle. closed loop. o2 sensor will most of the time be on 800s. as for the maf reading it will sit in the 9.xx's pretty much whole time.

 

9.75 grams/second is much too high for 675 RPM. That's the source of the rich condition that the BLM/INT is trying to correct. That's about double what it should be. Probably not the ducting alone that's to blame, it's probably the MAF itself.

This is why we are all going to Speed Density ECU's - so we can eliminate that old junk that you can't buy decent replacements for. I'm a LINK dealer and we just developed a 7165 ECM replacement harness that is 100% plug and play with the stock wiring. Only change is to reroute the MAF to a MAP sensor like the 90-92 cars. It's super nice to be able to tune in realtime and datalog with a modern laptop, etc. Hit me up if you are interested.

GD

 

O2 should toggle between .200 and .900 in closed loop. Sitting above .800 means the system is stuck rich. The low Integrator(short term fuel trim) number is the system trying to pull fuel. The Block Learn(BLM or long term fuel trim) is on its way down also. Normal MAF should be around 6gm/s at idle. The data values in MAF disconnect default are much better with some lean correction, which is normal for an older engine. Like I said, this proves the MAF or circuits are the problem. Again, try a different MAF. In too many diag and repair sessions to count on these systems, I have never seen a case where the MAF elimination test was wrong. The fault is either the sensor which in your case, being the MAF is used and of questionable origin, is almost a given. Or it's the sensor circuit causing the value to be skewed. As GD said, get away from the crappy early GM MAF system as soon as you can. SD is much better.

 

I have swapped and tuned 3 different MAF TPI cars to SD keeping the 165' ECM. Basically delete the MAF and burn-off relays. Then add a MAF. Then use a Holden 308 based code that runs in a 165'. A little tuning and they run great.


​​​

 

True, it can be done, but the 165 ECM is seriously old junk. It disconnects constantly while logging, and good luck for the new guys attempting to learn tuning by chip burning. Not that it can't be done - it absolutely can. But you're honestly talking about a learning curve that's going to take months, if not years. Sure there are products like the Moates Autoprom.... but I own one and it's not all they make it out to be. The ribbon cable interface is fragile and prone to interference... and yes it can get the job done on a dyno, but I personally find it dangerous on the street.

Sure you can buy a cheap chip burner and use some poorly documented Holden ROM on 30 year old out of production electronics with flaky, slow logging and no real time tuning support.... yeah you totally can. I've done it also and it SUCKS.

I'll take a real-time USB interface with software designed to run on Windows 10 any day over that mess.

And sure you can get rid of the MAF that way. Still limited to the factory tables and the rest of the factory hardware.

You get, at most, exactly what you pay for with respect to this subject. You can pay very little and have a massive learning curve and huge frustration, or you can get some modern, decent hardware and get on with tuning and driving. If you already have the experience as a tuner, and with all this old chip based GM hardware - great! But I say don't make the mistake I did by learning all that $hit. Just move to the 21st century and leave all that for the history buffs.

GD

 

thanks ASE Doc and GD for the help!

new stuff i tried
1. so i tried a Autozone duralast MAF for the heck of it and returned it shortly after. the engine would surge over and over again, showng maf readings ranging from 9 to 45. so that was junk.
2. i reverted back to original intake air filter and duct for only before the maf, the ebay piping is still used for after the MAF. tried the autozone maf same result. i tried the Delphi maf it ran a bit better then usual and the maf reading would drop to the 8's sometimes instead of always 9's.
3. surprisingly i tried the original Bosch MAF and it appeared to work as it should again with using the stock air filter and duct. i guess i assumed the Bosch maf was bad when i initially started troubleshooting but it was only not working correctly because of the hole in the duct. and then more recently was not working because of the new intake. unless its intermittently not working correctly. its sits at 5gm/s at idle hardly moves from 5.

anyways the car idles much better now but still not right.
1. it will make a thumping noise occasionally which you car hear from the tail pipe. i thought this was misfire but GD you say its only misfire when it when it runs lean, not rich. so i dont know what the noise is. when it occurs i don't really notice a change on the datalog. other then the o2 sensor hits high 800's i guess and the spark advance in rel. to ref and TDC goes up a bit for a second.
2. after idling for the last 5 minutes when engine was hot the BLMs slowly dropped back down to 108 while INT stayed at 128. BLM's would rise to 128 whenever giving throttle.
3. the car will also continuously release smoke from tail pipe even after 20 minutes. it looked more white then black but hard to see in dark. smoke was stinging my eyes like crazy.
4. lastly car will backfire bad out of the exhaust not intake i think anyways. when releasing throttle from anywhere above 3k rpms. and you can see smoke puff out the top of engine. when in park.
5. datalog shows the car is in 3rd gear true when idling in park. when car is turned off it changes to false. probably irrelevant.

any ideas? could this be a timing issue? the distributor is 30 years old and ignition control module...other then distributor or valve problem i really dont know what else it could be. as for the valves though engine compression test was even 150's across all 8 though.
i have been putting off checking or replacing the distributor for the last thing on the checklist since my engine doesn't have a timing mark, and i haven't looked up how to set base timing without one...

most recent datalog screenshot, closed loop fully warmed up. current running time almost 5 minutes. turned engine on and off a few times over 25 minutes to see what would change.


 

Ok well sounds like it's still rich.

A misfire will drive the O2 sensor lean and the ECM will react by adding fuel (percieved lean condition), and the BLM/INT will be higher than 128. So it's not entirely a misfire problem - thought it does sound like you have that also. Possibly from the overly rich condition.

The MAF is now reading correctly - excellent. So now you need to know why it's still rich at idle. We see that the O2 can cycle so it's likely still working. Possible suspects are leaking injectors, or a ruptured fuel pressure regulator diaphragm. Could also be the source of the misfire.... smell the vacuum line for the regulator.

If you haven't checked timing you absolutely should. That can absolutely affect your AFR and cause the ECM to react.

GD

 

i went through the fuel system already following service manual procedures mostly. fuel pressure would hold at 39psi i believe at idle. fuel pump 2 years old now, but less the 500miles on it. fuel pressure regulator passed that test by seeing if its leaking into intake. also did the leak test to see if the fuel system was leaking anywhere. no leaks at pump check valve, regulator or injectors. but even though it all passed i bought new regulator and injectors anyways because i couldnt figure out whats wrong. no new fuel pump since its already new.

i still dont understand the checking timing reason though. if i never messed with the distributor and it used to run perfect. the base timing can just change by itself or something? but yea im about ready to buy a distributor next i guess. and learn how to make a new timing mark.

 

39 psi at idle is too high. Idle vacuum should be ~10 psi. Standard GM fuel pressure is 43.5 psi. So you should be at about 33.5 to 34.5 psi at idle. An extra 5 psi of fuel pressure can easily cause the rich condition it's correcting for.

GD

 

well. it might have been 35-36 its been a couple months since i did all this troubleshooting i cant remember. and this was when the car was mostly cold still since other then disconnecting the maf it was hard to get to warmed up closed loop idle. and i already returned the fuel pressure tester. what would be the fix for this anyways if it was slightly high pressure a new fuel pump? the regulator is new acdelco and the fuel injectors i have tried the original injectors that were in car when i bought it probably original ones, and the ones i bought which i regret buying as well because i didnt know about southbay injectors, until after the return period ended for the standard motor products FJ47 i bought and tried. the fuel pump has less the 500 miles on it and 2 years old ACDELCO EP241 fuel pump.

as for vaccum i used a vaccum tester plugged into the intake through the evap canister line, and on the gauge it was in the green zone/good moto zone on 17inmg i believe it was, and passed the testing for blipping the throttle and holding rev and all that. 17inmg i guess converts to 8.3psi. once again the engine was not fully warmed up when tested probably still in open loop.

 

Well all I can tell you is there's too much fuel. Be it injectors or regulator or ? The O2 works and the MAF is within proper range at 5 grams @ idle. So unless someone has switched out the PROM or custom tuned it.... it's pulling out about 8% fuel with the INT at 108 and the BLM at 128. Which really won't be all that noticeable really once it's corrected for it. So I suspect you have other issues that are more concern at this point. The misfiring and all that other behavior.

White smoke could be coolant - rich would be black and it would have to be down in the 9's to see it.... which 8% wouldn't do. And just because you don't see a lean spike on the narrow band O2 from the misfiring doesn't mean it's not misfiring. With a narrow band it's just lost in the noise from 7 other cylinders.

GD

 

Like GD said, you likely won't see a response from the O2 sensors or fuel trims from a single sporadic misfire. That misfire could be caused by a number of things. Top of the list being injectors with old worn pintles that are not producing a good spray pattern. Second being a poorly performing ignition. If still running the original distributor, it's suspect. Third being worn valve guides.

The 108 BLM is certainly rich, but it is in range. I am confused though with your saying the INT stays put while the BLM changes. INT is short term and should toggle high and low to keep the O2 sensor's toggling centered around 450mV. You may not see INT toggle watching it on datastream. You might have to select custom data with only INT, BLM and O2 voltage to get fast enough refresh rate, but it does toggle. BLM is long term and should be more stable as it is a base correction to keep INT centered around 128. If it starts out at 128 and works its way to 108, this means it is adjusting to correct for a basic rich condition. I would like to see BLM stay between 120 and 136 as this is a small correction range. However, if all else is functioning okay and the engine runs well, 108 is within the system's correction range and if it's the best you can do, it isn't a no go on an old high mileage system. The white smoke could be steam from the catalyst burning off alot of carbon built up from the engine running so rich before. If the white smoke continues after some miles of driving, look at coolant loss into the motor. You may also want to look at the smoke to be sure whether it's coolant vapor(will smell like coolant) or oil smoke, which is very light gray.

I think what GD means to say is that fuel pressure should drop about 10psi with idle vacuum applied to the regulator. Idle vacuum should be around 18in on the gauge. 17 is a hair low but for a high mileage motor it's not terrible. BLM will change as the engine moves through load and speed ranges. One data PID you might look at is Block Learn Cell. IIRC, the early GM should have 19 cells. Each cell is a particular block of the engine's VE map or base fuel map. The VE map is a three dimensional map that starts at zero rpm and zero load and ends at max speed and max load. Throughout the map from one end to the other is every possible combination of speed and load that the engine might see during operation. Each block learn cell is a portion of the VE map and BLM will be different for each of these cells depending on basic fuel correction for that cell.

 

alright i put back in the new fuel injectors again fuel pressure is at 38. as for 33.5 to 34.5 psi at idle you say, i read other posts that say for our 5.7 tpis it is supposed to be around 40 not that low. either way i have no way of lowering the pressure anyways. the pressure regulator is new acdelco. the SMP injectors are new, the acdelco fuel pump is basically new. the return line is not clogged at all.

after putting back in the new injectors. and using the stock intake box and pipe before the maf. and using my original bosch maf. the car started normal again, and would occasionally pop or misfire or whatever it is out of the exhaust. the car was still letting out a bit of excess exhuast, but i dint see any smoke other then regular start up white smoke. the datastream was all good during idle up until car warmed up.

the next day same results except this time i took for test drive. seemed great. later i watched the datastream and now the BLM was in the 150s the whole time so running lean now while driving. so at idle its a bit rich and driving its lean now. i turned off the car, fuel pump wouldnt turn off so i pulled the connection to the oil pressure sender. it stopped it. probably need a new sender but after reconnecting it was good again.

probably unrelated but now after starting it again 10 minutes later. it runs like complete FN garbage again!! what the hell! cant even datastream it because the engine would probably blow up! after a few seconds got a maf code 33 , but other then 10 seconds of running it twice i had to turn it off. the bosch maf must be working only sometimes or something, becuase unplugging it makes normal again. waited about 30 minutes plugged in maf and turned it on seemed like it was gonna run like **** again, but i tapped the maf a few times and it seemed to make it work normal again for now.

 

I have gone through multiple MAF sensors from different suppliers looking for a single good one. The remans are all junk and no one cares because these systems are so old and practically obsolete. It's possible too that you have a faulty MAF sensor circuit, either in the ECM or in the wiring. A circuit fault, in the sensor ground, signal or reference, will have the same effect as a faulty sensor. To find it you will have to monitor voltages on each circuit when the fault is present and when it's not present. Most likely, it's just a crappy sensor.

 

The math doesn't lie. I don't know what to tell you, but a normal idle vacuum is 18-22 in/Hg. That's 8.8 to 10.8 psi. Static fuel pressure is supposed to be 43.5 psi. If you then subtract 8.8 to 10.8 you get.... drum roll...... 32.7 to 34.7

That's the math. I don't care what anyone "says" or "thinks". That IS THE MATH. And it's correct. When I tested my idle fuel pressure while doing lap scope's of injector voltage offsets, my fuel pressure through a Snap-On pressure transducer is about 33.4 psi.

If your fuel pressure is higher than this, either your engine is not developing proper idle vacuum (or it's not present at the regulator), or your base fuel pressure is too high. It's that simple. Till you correct the abnormally high fuel pressure, you will have low BLM's at idle.

GD

 

Sorry GeneralDisorder... the math is not the math... you are not applying the math correctly. You just don't subtract the vacuum pressure from the fuel pressure... there is a lot more going on in the fuel pressure regulator. You just got lucky in how the numbers work out. See my new thread on how a fuel pressure regulator works.

Engineering is the mathematical application of science.... but you have to apply it correctly, just because they use the same units, does not mean they relate directly.

Also, all my TPI engines run 36-38 psig at idle... and they run just fine.

 

A couple of psi wouldn't amount to all that much difference in flow. For reference I've tested injectors at 40 psi, then jacked the pressure up to about 80 psi and they don't double in flow - not even close in fact. They gain about 15-20%. The load against the pintle creates a lot of additional latency getting the pintle to open. Of course this is relative pressure across the injector. 80 psi with 40 psi of boost reacts just the same as 40 psi fuel pressure.

GD

 

Nope, a few psig does not make that much of a difference... that my point with the idle fuel pressure... going from 33 to 36 psig will not change the performance.

Double the pressure does not double the flow. Flow versus pressure is a squared relationship...(pump affinity laws). to double the flow, you need four times the pressure (assuming perfect conditions).

 

Probably not all that much, but he does have a rich idle so it's worth checking manifold pressure and seeing if the regulator is out of spec. If, based on a manifold pressure of 10 in/Hg, you expect to see 33 psig and you actually see 36 psig, then that's about 30% off. It would impact the fuel trims.

It is almost certainly not his only issue though.

GD

 

36 psig against a setpoint of 33 psig is an error of 9.09%..... not 30%.

The math is the math!!!!

 

That wasn't the math I was doing. I said based on a manifold pressure of 10 psig..... if you start with 43 psig fuel pressure, 10 psig manifold vacuum but your regulator is only dropping the pressure to 36 psig then you have 7 psig of offset being provided by the regulator and the injector is now seeing an effective pressure of 46 psig across it when it would only see 43 psig at WOT (manifold vacuum near 0 psig). What I mean is that the REGULATOR is 30% out of spec at idle. It should be dropping the fuel pressure by an additional 3 psig for a constant 43 psig across the injector at idle. 3/10 is 30%.

GD

 

Here's a very recent related video (there's a whole series I've been following leading up to this - this was posted this morning). This shows how a 19% increase in fuel pressure (this is a TBI that's about 3 psi over the max spec) can throw the idle fueling so far off that the ECU can't even bring the mixture under control. If you watch the whole series you will see that the BLM and the INT are maxed out and the O2 is still pegged rich at idle and won't come down under control.


OP's car is only 9% over on fuel pressure, but his ECU is in control of the fuel with the BLM at 108 and INT at 128. It's under control, but it's not ideal. And that's VERY LIKELY because of the high regulated fuel pressure. Just because the ECU is able to compensate, doesn't mean it's in perfect operating condition. You could tune around it if you wanted to burn a new chip, but the regulator isn't doing it's job.

GD

 

Final update to close this thread. possibly save someone headaches if the same issue. after a year on and off of the problems, and the extensive troubleshooting i listed above. apparently not enough testing even though it frustrated the hell out of me. I do not know if all the new parts i added contributed to my car running good now. but after i pretty much tried almost everything anyone can check and test.

it turned out to just be the maf relay or maf burn off relay. i tested both relays as one of the first things i checked with a 12v,multimeter, alligator clips, probe light. since starting with the service manual working under the troubleshooting MAF code sections for code 33 and 36 i think it was. at first the car didn't set codes but eventually set code 33 or 36. i first figured it was the maf because doing the unplug test the car always ran good. so first i got a delpi maf.

after it didnt work i continued testing the long list of other possible problems. after testing everything else and following advice for fuel pressure etc, and replacing parts etc. i figured the maf is garbage because its said aftermarket mafs are all garbage for our cars. so i decided to try one more maf anyways a original rebuilt bosch. the car worked seemingly well for about 6 months or something not driving much though with the code 36 constant. problem came back with car running like crap. so since i tried everything else already i decided to just replace the 30 year old relays before i moved onto speed density map sensor or something similar. the car runs good again. even though the code 36 popped back up again...still runs normal but i guess i will see....

so yea screw testing relays that are that old. just replace them.

 

This might be a bit of a thread resurrection, but I may be interested in further details. I'm getting sick and tired trying to tune the 165 MAF. When I'm not trying to figure out what in the world the ECM is doing, it's always disconnecting!

 

Sure - send me a PM. And yes the 165 is primitive and can be frustrating. Especially the disconnecting behavior.

GD