Have a bit of a head scratcher

[Derrick852]

Hello all, creeped for a few months here but finally hit a wall with the truck and wondering what the community thinks about it.

A little background - I won the truck (69,000 miles) in a draw basically (I know, crazy), and have been going over the truck for a few months now fixing just about everything under the sun. The truck had gone through many owners, non of whom were particularly mechanically competent. 31 years of boosting and owner swapping took a bit of a toll on the poor thing, but I have it pretty well squared away at this point.

However, I seem to have a problem with the ECU/MEMCEL (I think) and I am looking for a bit of clarification. The truck will start up on a high idle and take a while to come down, when it gets down to idle it chugs and will sometimes stall. Now, for the good stuff.....

How/Where is the BARO value stored? To my understanding the ECM looks at the MAP on KO on stores that value as the BARO, and then corrects the MAP signal with the engine running with the stored BARO value to end up with a calculated value. So, I have seen 2 different values for BARO, .470 and .580, neither of which are correct. Initially, I just thought it was a bad MAP sensor, so I chucked a new AC Delco unit in there and it did the same thing so then I really started looking. I have the full 5V and good ground at the plug, and with the engine running the ECM is seeing .53v on the MAP which according to the charts I've found online works out to around 30kpa which matches the analog boost gauge in the truck that shows around -.4bar vac (not the greatest gauge in the world, but its analog and it works for this purpose). So if I take the full pressure reading (from the weather) that's been anywhere between .995 and 1.005 the last few days, subtract off the stored BARO value, and then add that to the MAP line, I get exactly what I should be seeing on the MAP.

So I think I've narrowed the problem down to wherever that BARO value gets stored. BARO shows an address 0xD5 and 4.18v scaled to .580 on Freescan, while the MAP shows .53v scaled to -0.010 and an address of 0x1B. The different addresses and the fact the MAP only has 3 lines got me thinking the BARO value must be inferred and I need to find out how/where that value is stored. Not really sure how it gets .580 from 4.18v, it almost seems like a short or a failing volatile memory chip. I have had some problems with the logging software stopping communications when the truck really starts to chug, it sees a header but no data comes from the bus. This could be glitchy old software/cables, but they do occur at the same time and figured it was worth mentioning.

Obviously I've omitted countless hours of Diag here, and just tried to focus on what I see as the problem. I can't find a damn thing wrong anywhere else with anything.

So before I start chucking ECMs at this thing, what do you all think?

Is the BARO value wrote to volatile memory on the ECM and not the MEMCEL? From my reading online is seems the MEMCELS are read only after burning, so I am fairly certain its contained within the ECM. Parts are hard to find for this thing, I thought it wise to pause for a bit and see what the interwebs thinks.

THANKS.

 

[Derrick852]

I guess I could add a little more here, I have put approximately 600 miles on it since I got it. When I got it the previous owner said it would occasionally puke a code for the O2 sensor, I did not get if it was high or low or what. I did, on 2 occasions have the CEL come on while cruising on the highway, but it went off shortly after both times - Never looked into it further.

The ignition system decided to die one day in my driveway (Suspect ICM but no way to confirm), and that lead me on quite the adventure, but resulted in a perfectly repaired/timed system.

The truck always has had a hanging idle sometimes, and on a few occasions sputtered during idle recently which has led me to this point. It is what you would describe intermittent, and seems to need time to load/heat soak whatever it is. My gut says electrical.

The IAC functions as is should, and will even close up all the way trying to bring the idle down. Readings will show 0, and the idle is still high. With the vacuum present (30-40 bar "ish" @ 4000' ASL)on the analog gauge there is no way I have a massive air leak. Fuel pressure is bang on, never faulted, and the plug read when I changed plugs was a slightly rich even burn on all 6 plugs.

With my MAP readings causing a lean condition (confirmed with high V on O2, hanging idle, and a very unscientific smell test of the exhaust), the only thing I can conclude is the engine is way lean due to faulty MAP readings due to a corrupted BARO value.

Truck will show O2 High, IAC something, and MAP low codes now. All due to the MAP reading being faulty. Started with O2, then IAC, and now MAP low. I have no reason to believe there is anything wrong with the IAC as it functions and responds when commanded, and the position makes sense. The O2 shows what I believe to be the correct voltage (Sensor is new, I didn't do it, I suspect someone was chasing this problem). Quite the adventure.

It took me some time to arrive at this, but I am starting to think I am looking at a slowly degrading electrical component that's been showing signs for years. There has been someone who went over the engine looking for vacuum leaks already. I think this problem has been coming a long time, perhaps a few people have tried to fix it, and has now almost "Perma Broken", and I finally have the damn thing on a laptop to see what the hell is going on.

I'm going to confirm the voltage the ECM sees at Key On with my Fluke meter at the ECM pins for the MAP and just make sure that matches up to what it should be (which I am sure it will), and that just leaves something in behind that not getting/storing the value correctly.

Thanks for the response. I'm going to do a bit more digging here and keep this post updated.

 

[cosbysweater]

so i believe on alot of the older ecus (maybe not this one unfortunately) the baro reading is taken off the map sensor at the first key on cycle.

edit just confirmed that. pulled this straight from all data.

With the ignition "ON" and the engine stopped, the manifold pressure is equal to atmospheric pressure and the signal voltage will be high. This information is used by the ECM as an indication of vehicle altitude and is referred to as BARO. Comparison of this BARO reading with a known good vehicle with the same sensor is a good way to check accuracy of a "suspect" sensor. Reading should be the same ±0.4 volt.

A Code 33 will result if CKT 452 is open, or if CKT 432 is shorted to voltage or to CKT 474.

i ended up going through a few distributors chasing issues on my typhoon, the only one that worked right was a brand new cardone unit. the ac delco "reman" ones are complete garbage. mine was idle hunting and doing all kinds of weird stuff and the new distributor fixed it. that is after i fixed the intake leaks, make sure your lower plenum gaskets are not leaking, easy to miss

 

[Derrick852]

Thanks man! I wish I had an All Data account, don't fix enough stuff to justify the cost. Thanks so much for confirming my suspicions for how the ECM calculates BARO. I'll confirm the MAP pin sees the voltage its supposed to at key on and go from there.

In regards to the distributors, I have a funny story about that. The one I replaced in the truck was the original Delco Remy style, with what I'm sure was the original ICM that failed. I replaced it with a AC Delco reman (not the Remy style, along with everything else in the ignition system) lol, shortly after I had this issue pop up, and I went into the reman looking for answers. I found that they used some cheap made in China ICM (I actually sent an Email to AC Delco explaining my displeasure of this, talks are still ongoing), replaced that with a real GM ICM and since the problem persisted, that caused me to get some scanning gear and start looking into the ECM, and that's where I found the BARO issue.

Creeping the forums you always get vac leaks, it was one of the first places I started thinking about, but based off of my Vac readings at idle on the gauge, the readings and behavior of the IAC, if I have a leak, its small, and I don't think it would be causing the weird idle. The data readings on that front look good, and the gauge sources the vacuum from a T in the line to the map, so I see the same thing on the gauge the MAP sees.

 

[cosbysweater]

yea those ac delco remans are absolutely terrible, horrible quality pickup and icm in them. i installed mine and it ran better until i drove it and it dies almost immediately. i got no tach signal while cranking so i just grabbed a new cardone unit and threw it in and put my original icm and a new delco one in the glove box. the coil pickup is what failed on mine i believe. The factory gauge is pretty hard to use for diagnostics, my car had a ton of small vacuum leaks and a blown out lower intake gasket and the factory gauge was barely bouncing at all but i could see my rpm fluctuating quite a bit. also make sure the map is the correct part number.

how does it sound out the tail pipe?

 

[gmemony2]

The data readings on that front look good, and the gauge sources the vacuum from a T in the line to the map, so I see the same thing on the gauge the MAP sees.

It's always been my understanding to not T off the MAP vacuum line or fpr. How much it actually effects it I cannot say. Most times people are coming off the vacuum line for the brake booster; adding a fitting for the gauge. Of course, this is assuming an aftermarket gauge...

 

[Derrick852]

I tapped into the wire leading to the ECU that the map sensor output comes into the ECM on. I used my Fluke meter, set to MIN/MAX to catch any voltage spikes and rolled the key on. Showed 2.1v which equals around 90 Kpa, current weather is 97 Kpa so that's plenty close enough for this quality of sensor. Inside of the ECM however BARO was again showing as 58 Kpa with a voltage of 4.18v (This is around 170 kpa on a 2 Bar GM MAP Sensor based off voltage).

There were no observed voltage spikes at key on, and the voltage the ECM reports for the MAP while the engine is running is correct, and scales to atmospheric conditions. It just doesn't save/update the BARO correctly.

I have concluded its a failed ECM or MEMCAL. When I pulled the ECM, it was already a Delco Reman so that was a super awesome sign. The MEMCAL is a AYBM 4053, so one of the earlier ones so no loss there (Maybe the original one, perhaps the ECM was swapped trying to fix this issue, and the re used the MEMCAL as they are basically unobtainable at this point and the remans don't come with them)

When I learn where everything gets stored I'll update this post again to save others the same headache should this occur to them.

 

[Derrick852]

I have no idea how to work the quotes thing here so....

How does it sound out of the tail pipe? - Dry, lean, rough, chuggy, kinda popping almost but nothing like a lean bike on decel or anything like that.

As for T'n off of the MAP line for a gauge, that does make sense, the extra volume of the lines and gauge would slow the response time. I didn't put it in there, but I will eventually move it. I never put much thought into this before you mentioned it. Thanks.

 

[Derrick852]

Quick update here.

So I managed to get some answers on where the BARO value is stored. It is stored on a RAM/FLASH chip on the PCB of the ECM.

I'm going to replace the ECM with a Blue Streak Reman one from Rock Auto and see what happens. They say "Plug and play", see how that goes.....

 

[Derrick852]

Well this is a bit embarrassing. After talking to a very smart person, the BARO value actually contains a "bug" from factory, and the value is not applied to any live sensor feeds. What was happening is Freescan was displaying the values wrong, while displaying some other ones correctly and sent me down a completely unnecessary rabbit hole. In addition, the idle control solenoid was reading one position, while being in a completely different position.

I found a copy of EFILive, fired that up and went to work. It looked like a classic vac leak. Checked everywhere with a propane torch on a running engine with the intercooler removed. Didn't find anything. But the noise coming from the ICS seemed weird.

Pulled the throttle body, replaced the gasket and ICS and reassembled. I needed to adjust the throttle stop, it was set way to far in (the butterflies were open to much), probably about a full turn.

She idles at 725 RPM like a dream, ICS reads 32 with the engine at temperature.

So yeah. That was dumb....

 

[DaveP's Ghost]

I stayed quiet. You seemed to have convinced yourself what was wrong.

But yeah, when the IAC counts are Zero, the first thing to try is to close the throttle plate, and see if idle control improves. If the throttle plate is completely closed and the IAC counts are still zero, there is usually a vacuum leak. Many early GM OBD I calibrations have a Code 35 'Unable to control idle to desired speed". (The 92 Sonoma GT with the CPI engine has this code.) The cause of Code 35 is always a vacuum leak. Unfortunately, the SyTy calibrations do not contain the Code 35 DTC.

 

[Derrick852]

The plug was still in the throttle body stop, so I never even bothered to try it, I just assumed the factory value would be correct. Lesson learned. Time to replace a leaky heater core and whatever else this stupid thing decides to throw at me, I can't imagine it has much left in it.