Dismiss Notice

Welcome To CK5!

Registering is free and easy! Hope to see you on the forums soon.

Score a FREE t-shirt and membership sticker when you sign up for a Premium Membership and choose the recurring plan.

Rough running until warm up

Discussion in '1973-1991 K5 Blazer | Truck | Suburban' started by 4by4bygod, Aug 25, 2004.

  1. 4by4bygod

    4by4bygod 1/2 ton status

    Joined:
    Nov 13, 2003
    Posts:
    3,859
    Likes Received:
    0
    Location:
    With My Tinfoil Hat
    Hi all!

    My truck just started doing this.. it idles rough and runs with a shudder, until the engine warms up and the thermostat opens. after that, it's fine. Any suggestions before I start replacing stuff? The O2 sensor,cap ,rotor and wires aren't a year old yet, and neither is the egr valve. maybe a coolant temp sensor?

    It's a bone stock 87 305 w/TBI, and 206,000 miles on it.

    Thanks, Tom /forums/images/graemlins/1zhelp.gif /forums/images/graemlins/usaflag.gif /forums/images/graemlins/k5.gif
     
  2. 1-ton

    1-ton 1/2 ton status Premium Member

    Joined:
    Apr 17, 2004
    Posts:
    2,884
    Likes Received:
    165
    Location:
    Las Vegas, Nevada
    Sounds like the computer on your truck may not be going into "open loop", and is stuck in "closed loop".

    Open Loop is where the computers electrical circuit is not a complete circuit (i.e. open). This "open loop" condition occurs when the trucks engine is cold, and the computer does not control the engine. At this time the fuel injection system defaults to full rich, thus functioning like the choke on a carburetor, until the engine is warm.

    Closed loop is where the circuit is complete (i.e. closed). The "closed loop" condition occurs when the trucks engine is warm. The trucks computer is then allowed to control the engine, thus controlling the mixture of the fuel to air.

    A few sensors are responsible for telling the computer when to go into "open loop" and "closed loop". The O2 sensor and temp sensor is a couple of them.
     
  3. Blue85

    Blue85 Troll Premium Member

    Joined:
    Jul 26, 2000
    Posts:
    9,084
    Likes Received:
    486
    Location:
    Grand Rapids area
    [ QUOTE ]
    Sounds like the computer on your truck may not be going into "open loop", and is stuck in "closed loop".

    Open Loop is where the computers electrical circuit is not a complete circuit (i.e. open). This "open loop" condition occurs when the trucks engine is cold, and the computer does not control the engine. At this time the fuel injection system defaults to full rich, thus functioning like the choke on a carburetor, until the engine is warm.

    Closed loop is where the circuit is complete (i.e. closed). The "closed loop" condition occurs when the trucks engine is warm. The trucks computer is then allowed to control the engine, thus controlling the mixture of the fuel to air.

    A few sensors are responsible for telling the computer when to go into "open loop" and "closed loop". The O2 sensor and temp sensor is a couple of them.

    [/ QUOTE ]

    Hmmm...That is mostly all wrong, so I had better come in and correct some things.

    1) "Open loop" means that feedback is not used in the control. "Closed loop" means that feedback is used; in this case from the O2 sensor.

    2) The computer controls the engine in either mode. In open loop, it reads the throttle position, temps, and MAP to calculate the amount of fuel it "thinks" the engine should need. In closed loop, it continuously monitors the oxygen content of the exhaust and adjusts the fuel mixture up and down based on this reading.

    3) While closed loop is more precise and efficient, open loop IS NOT "full rich". You don't need a choke becuase the computer richens the mixture without one.

    4) The computer basically makes the switch based on coolant temperature. It also needs to see good O2 readings.

    5) If the truck was stuck in open loop, it would not smooth out after it warmed up. It would run too rich and get bad fuel economy. The problem with this vehicle is occuring during open loop operation. I don't see how a vehicle could get stuck in closed loop. When you first start the engine, the O2 readings are so bad it shouldn't run at all.

    The only things I can think of that would cause this are:
    -improper timing
    -vacuum leak around intake/throttle body.
    -bad MAP.
     
  4. 1-ton

    1-ton 1/2 ton status Premium Member

    Joined:
    Apr 17, 2004
    Posts:
    2,884
    Likes Received:
    165
    Location:
    Las Vegas, Nevada
    [ QUOTE ]
    Hmmm...That is mostly all wrong, so I had better come in and correct some things.


    [/ QUOTE ]

    Your an ass!

    This guy had some 30 something reads, but no replies, until I tried to shed some light on his problem. You are one of those "ass bites" that will not reply to anybody's problem, but will wait to pounce on the first person that dares to. All you did was reiterate most of what I said. I was probably working on some of the first computerized automobiles when you where still shitting your diapers...don't tell me what Open and Closed loop is. I am well aware of what it is.

    The term "open loop" originally meant, "open circuit". Meaning that when the O2 sensor was cold, it would not pass voltage, which meant that the circuit was not complete. I attended Arizona's Automotive Institute back in 1978 when Datsun and Volvo's where some of the first computerized vehicles in America. I am sure some things have changed since then, but most of the theory is still the same.

    I also spent 15 years as a Heavy Duty Diesel mechanic, and I bet I could out wrench your ass any day.
     
  5. TopOff

    TopOff 1/2 ton status

    Joined:
    Jan 16, 2002
    Posts:
    1,760
    Likes Received:
    0
    Location:
    Folsom, ca
    My guess is a coolant temp sensor.

    Buy a sensor check the resistance when it is cold. Compare it to your sensor when the engine is cold. If the resistances are similar, it's not the sendor.

    Good luck.
     
  6. 4by4bygod

    4by4bygod 1/2 ton status

    Joined:
    Nov 13, 2003
    Posts:
    3,859
    Likes Received:
    0
    Location:
    With My Tinfoil Hat
    Hi all!

    Thanks for the replies. I put in a new coolant sensor (the one in the manifold)and nothing changed. I got a new MAP sensor, and after I hooked it up, the truck wouldn't start. It would just crank away, and not fire. I've had coils go bad on me with no warning, so I swapped in a coil I knew was good, just in case. still wouldn't start. I know jack about MAP sensors, so why it would have anything to do with the truck not starting is a mystery to me.

    I put in the old MAP sensor that I hadn't thrown away yet, and now the truck runs as good as it did when I started this whole mess. /forums/images/graemlins/frown.gif

    I re checked the mileage I have on the cap / wire / rotor, and I replaced them 20,000 miles ago, and I have 9,000 miles on the plugs themselves. maybe that's part of my problem? Then again, if they were bad, wouldn't the truck still run rough after it warmed up? I dunno, I'm just thinking out loud.

    At the parts store, I had a choice of 2 MAP sensors, a black one, or a gray one. I got the black one, because it was identical to the one I took off. Any idea why I'd have a choice of two? Should I have gotten the other one?

    I hate to sound like one of those parts replacers we laugh about, but this one's got me stumped.

    Tom /forums/images/graemlins/k5.gif /forums/images/graemlins/usaflag.gif /forums/images/graemlins/1zhelp.gif
     
  7. Mastiff

    Mastiff 1/2 ton status Premium Member

    Joined:
    Jul 21, 2002
    Posts:
    2,755
    Likes Received:
    48
    Location:
    Tucson, AZ
    Your original symptoms made it sound like something changed on the motor such that the fuel delivery tables stored in memory were no longer correct, but that the computer would adjust for it once it went into closed loop mode.

    I'm not sure what is going on when you replaced the MAP sensor. Did you pull power to the ECM after you replaced it? This would be a good idea so you erase any learning the ECM did with the old MAP sensor.

    I've been hacking on this stuff a lot recently, but I can't imagine just stabbing in the dark without any data. Do you have a laptop computer? If so, I'd highly recommend getting WinALDL for free and seeing what the computer is doing. The worst part is that you'll need to make or buy an interface cable. They have instructions on the WinALDL site though.

    Good luck.

    Oh, I forgot to ask. Have you driven around much since you've had this problem? I'm wondering if the ECM has had time to compensate for any new weirdness. Not that this is necessarily a good thing, but it'd help to know whether the issue is one that the ECM is capable of compensating for.
     
  8. k5freak44

    k5freak44 1/2 ton status

    Joined:
    Jun 1, 2004
    Posts:
    2,164
    Likes Received:
    0
    Location:
    Plano, TX
    [ QUOTE ]
    [ QUOTE ]
    Hmmm...That is mostly all wrong, so I had better come in and correct some things.


    [/ QUOTE ]

    Your an ass!

    This guy had some 30 something reads, but no replies, until I tried to shed some light on his problem. You are one of those "ass bites" that will not reply to anybody's problem, but will wait to pounce on the first person that dares to. All you did was reiterate most of what I said. I was probably working on some of the first computerized automobiles when you where still shitting your diapers...don't tell me what Open and Closed loop is. I am well aware of what it is.

    The term "open loop" originally meant, "open circuit". Meaning that when the O2 sensor was cold, it would not pass voltage, which meant that the circuit was not complete. I attended Arizona's Automotive Institute back in 1978 when Datsun and Volvo's where some of the first computerized vehicles in America. I am sure some things have changed since then, but most of the theory is still the same.

    I also spent 15 years as a Heavy Duty Diesel mechanic, and I bet I could out wrench your ass any day.


    [/ QUOTE ]

    Comeon guys, keep fighting, dont stop now, RELEASE YOUR ANGER /forums/images/graemlins/weld.gif /forums/images/graemlins/shame.gif
     
  9. 4by4bygod

    4by4bygod 1/2 ton status

    Joined:
    Nov 13, 2003
    Posts:
    3,859
    Likes Received:
    0
    Location:
    With My Tinfoil Hat
    <<Your original symptoms made it sound like something changed on the motor such that the fuel delivery tables stored in memory were no longer correct, but that the computer would adjust for it once it went into closed loop mode.>>

    <<I'm not sure what is going on when you replaced the MAP sensor. Did you pull power to the ECM after you replaced it? This would be a good idea so you erase any learning the ECM did with the old MAP sensor.>>

    <font color="red"> </font> All I did was disconnect the battery while I was changing the sensors. Dumb question, but where is the ECM on this thing? my chiltons book doesn't tell me crap. Could the ECM be bad?


    &lt;&lt;I've been hacking on this stuff a lot recently, but I can't imagine just stabbing in the dark without any data. Do you have a laptop computer? If so, I'd highly recommend getting WinALDL for free and seeing what the computer is doing. The worst part is that you'll need to make or buy an interface cable. They have instructions on the WinALDL site though.

    Good luck.&gt;&gt;

    No laptop, unfortunately. Any solutions for us desktop pc guys? <font color="red"> </font>

    &lt;&lt;Oh, I forgot to ask. Have you driven around much since you've had this problem? I'm wondering if the ECM has had time to compensate for any new weirdness. Not that this is necessarily a good thing, but it'd help to know whether the issue is one that the ECM is capable of compensating for.
    &gt;&gt;

    <font color="red"> </font> Yes, I've done a lot of driving, a lot of short hops, along with a few highway trips of at least an hours duration. The trucks behavior hasn't changed at all.

    Thanks for the help. I'm going to break down and chnge the cap/rotor/wires and see if that changes anything. I'll also disconnect the power to the ecm as suggested, and see if that other MAP sensor still keeps it from starting.

    I'm close to just taking it to the shop and having them hook it up to their diagnostics and see what's up.

    Tom /forums/images/graemlins/usaflag.gif /forums/images/graemlins/k5.gif
     
  10. Blue85

    Blue85 Troll Premium Member

    Joined:
    Jul 26, 2000
    Posts:
    9,084
    Likes Received:
    486
    Location:
    Grand Rapids area
    [ QUOTE ]
    You are one of those "ass bites" that will not reply to anybody's problem, but will wait to pounce on the first person that dares to.

    [/ QUOTE ] I guess you haven't read too many of my posts. Oh well. /forums/images/graemlins/dunno.gif [ QUOTE ]
    All you did was reiterate most of what I said.

    [/ QUOTE ]. Not true. If you compare what I posted to what you posted, there are several key differences. You probably know enough about engine controls to recognize this and you simply mis-typed things in your original post. If you don't see any errors in your original post, then you are confused about this stuff. [ QUOTE ]
    I was probably working on some of the first computerized automobiles when you where still shitting your diapers...don't tell me what Open and Closed loop is. I am well aware of what it is.

    [/ QUOTE ] OK, so you know what it is. Your description was wrong was my only point.

    [ QUOTE ]

    The term "open loop" originally meant, "open circuit". Meaning that when the O2 sensor was cold, it would not pass voltage, which meant that the circuit was not complete. I attended Arizona's Automotive Institute back in 1978 when Datsun and Volvo's where some of the first computerized vehicles in America. I am sure some things have changed since then, but most of the theory is still the same.

    [/ QUOTE ] "Open Loop" and "Closed Loop" are part of fundamental control theory and are not specific to engine controls.

    You probably can out wrench me. Since we are on the topic of professions, I am an automotive engineer, meaning I have my Masters of Science in Engineering in this stuff, so I do have a grasp of the theory side.

    Now let's set this down and get back to this guy's problem.
     
  11. 1-ton

    1-ton 1/2 ton status Premium Member

    Joined:
    Apr 17, 2004
    Posts:
    2,884
    Likes Received:
    165
    Location:
    Las Vegas, Nevada
    [ QUOTE ]
    "Open loop" means that feedback is not used in the control. "Closed loop" means that feedback is used; in this case from the O2 sensor.

    [/ QUOTE ]

    If you cannot see the similarities between this statement, and what I said, then you are trying to hard to be disagreeable.

    [ QUOTE ]
    The computer controls the engine in either mode. In open loop, it reads the throttle position, temps, and MAP to calculate the amount of fuel it "thinks" the engine should need. In closed loop, it continuously monitors the oxygen content of the exhaust and adjusts the fuel mixture up and down based on this reading.

    [/ QUOTE ]

    This is where the only difference between what I said, and what you said occurs. Yes, your statement is true in all TPI and later 80's TBI vehicles, but it is not true for early 80's TBI and the very early-metered carbureted computer controlled model vehicles. In the early days "open loop" meant the computer had no control at all on the vehicle. These vehicles had a mechanical ability to run full rich on the idle circuit without any input from the ECM.

    [ QUOTE ]
    While closed loop is more precise and efficient, open loop IS NOT "full rich". You don't need a choke becuase the computer richens the mixture without one.

    [/ QUOTE ]

    What ever! An engine runs rich until it is warm, and what term you want to argue is correct for that I don't care.

    [ QUOTE ]
    The computer basically makes the switch based on coolant temperature. It also needs to see good O2 readings

    [/ QUOTE ]

    This is correct, but I did not mention it in my post because it was already long winded enough as it was.

    [ QUOTE ]
    If the truck was stuck in open loop, it would not smooth out after it warmed up. It would run too rich and get bad fuel economy. The problem with this vehicle is occuring during open loop operation. I don't see how a vehicle could get stuck in closed loop. When you first start the engine, the O2 readings are so bad it shouldn't run at all.

    [/ QUOTE ]

    The first part about "open loop" is correct, but I disagree with the second part about "closed loop". A vehicle can start and run in either mode. All an engine needs to run is some fuel, air, and fire regardless if the mix is exactly correct or not. It may not run well, but it will run.

    Also, this is where our two worlds as you the Engineer, and me (used to be) Mechanic split. I have seen the difference between book theory and real world operation of motor vehicles. I cannot tell you how many times I have seen an ECM not pick up a trouble code from a bad input sensor; even though the books I studied said this was not possible.
     
  12. 1-ton

    1-ton 1/2 ton status Premium Member

    Joined:
    Apr 17, 2004
    Posts:
    2,884
    Likes Received:
    165
    Location:
    Las Vegas, Nevada
    The whole idea behind "open loop" and "closed loop" may be an out dated term, which does not even really apply anymore when it comes to modern ECM systems. It is kind of like when the term "floppy" applied to the 5-1/4" drive media, but carried over to the 3-1/2" drive media; even though there was nothing "floppy" about 3-1/2" drive diskettes.
     
  13. Mastiff

    Mastiff 1/2 ton status Premium Member

    Joined:
    Jul 21, 2002
    Posts:
    2,755
    Likes Received:
    48
    Location:
    Tucson, AZ
    [ QUOTE ]
    The whole idea behind "open loop" and "closed loop" may be an out dated term, which does not even really apply anymore when it comes to modern ECM systems. It is kind of like when the term "floppy" applied to the 5-1/4" drive media, but carried over to the 3-1/2" drive media; even though there was nothing "floppy" about 3-1/2" drive diskettes.

    [/ QUOTE ]

    The term "closed loop" is well defined in engineering control systems. It means that parameters are adjusted based on feedback control. "Open loop" means that there is no feedback and parameters are computed directly. In the case of an ECM "closed loop" means that the fuel delivery is adjusted to achieve a desired measurement from the O2 sensor. In open loop, the fuel delivery is based only on input parameters (RPM, MAP, temp, etc.).

    By definition, the ECM is also in a closed loop mode of sorts at idle when it adjusts the IAC to achieve a desired RPM. Tuning spark to eliminate knock from the knock sensor is also a form of closed loop control.

    In the context of engine control though, when people say "closed loop", they mean with respect to the oxygen sensor.
     
  14. 1-ton

    1-ton 1/2 ton status Premium Member

    Joined:
    Apr 17, 2004
    Posts:
    2,884
    Likes Received:
    165
    Location:
    Las Vegas, Nevada
    I was taught that the term "Loop" meant "Circuit". Is this correct or wrong? Also, in "closed loop" wouldn't all inputs (i.e. O2, temp, MAP, Barometric, Throttle, and RPM) have input and contribute to the ECM decision making process, and not just only the O2 sensor?

    Another question I have about today's ECM systems is how they read the inputs. In the early days ECM's had a constant connection to the inputs, but when I started get out of the mechanic trade, I learned that because of the increased amounts of ECM inputs they started sampling input readings at a predetermined staggered time intervals (for example every 0.5 seconds). Is this correct?
     
  15. Mastiff

    Mastiff 1/2 ton status Premium Member

    Joined:
    Jul 21, 2002
    Posts:
    2,755
    Likes Received:
    48
    Location:
    Tucson, AZ
    [ QUOTE ]
    I was taught that the term "Loop" meant "Circuit". Is this correct or wrong? Also, in "closed loop" wouldn't all inputs (i.e. O2, temp, MAP, Barometric, Throttle, and RPM) have input and contribute to the ECM decision making process, and not just only the O2 sensor?

    [/ QUOTE ]

    I'm an electrical engineer during the day, and I never hear anyone say loop in place of circuit, but maybe people in some circles say loop. I'm not sure.

    I've been on the steep part of the ECM learning curve for the last few months, so I don't know every detail - but in closed loop, I think the ECM will mostly do whatever it takes to get the right O2 reading. That is, within certain bounds. At a certain point it'll give up and assume the O2 sensor is broken. Even in closed loop, I know it does at least monitor the TPS so it can do the accelerator pump emulation. But within certain bounds, I'm pretty sure it'll ignore MAP, RPM and temp to make the O2 happy.

    Having said all that, the O2 feedback loop is relatively slow. I thing the open loop fuel delivery is the starting point even in closed loop. Say engine conditions in terms of MAP, RPM and temp are changing rapidly. The ECM will make an immediate fuel computation based on these parameters, plus any learning it has done previously. If the O2 reading ends up wrong, the ECM will start from that open loop baseline and start adjusting to get the O2 reading correct. This adjustment from the baseline is the source of information for the learning.

    Like I said though, I'm no expert so I'm open to corrections.

    [ QUOTE ]
    Another question I have about today's ECM systems is how they read the inputs. In the early days ECM's had a constant connection to the inputs, but when I started get out of the mechanic trade, I learned that because of the increased amounts of ECM inputs they started sampling input readings at a predetermined staggered time intervals (for example every 0.5 seconds). Is this correct?

    [/ QUOTE ]

    Well, the ECM is microprocessor based, so in reality it has to only work at sampling intervals. It can't even deal with data rates faster than the clock rate of the processor. But yes, the ECM does grab samples of the sensors through analog-to-digital converters at certain intervals. This isn't really a bad thing as long as the sample rate isn't too slow. What would an ECM do with O2 sensor readings at 100 MHz or something anyway - nothing in an engine changes that fast anyway.
     
  16. yeababy

    yeababy Registered Member

    Joined:
    Aug 30, 2004
    Posts:
    18
    Likes Received:
    0
    Location:
    sweden
    hi 4by4 not sure if an 87 ecu is same place as an 88 but mines behind the glovebox
     
  17. 1-ton

    1-ton 1/2 ton status Premium Member

    Joined:
    Apr 17, 2004
    Posts:
    2,884
    Likes Received:
    165
    Location:
    Las Vegas, Nevada
    [ QUOTE ]
    Well, the ECM is microprocessor based, so in reality it has to only work at sampling intervals. It can't even deal with data rates faster than the clock rate of the processor. But yes, the ECM does grab samples of the sensors through analog-to-digital converters at certain intervals. This isn't really a bad thing as long as the sample rate isn't too slow. What would an ECM do with O2 sensor readings at 100 MHz or something anyway - nothing in an engine changes that fast anyway.

    [/ QUOTE ]

    In the early days of ECM technology I used a standard Ohmmeter to take input readings, but as ECM technology changed I had to use a special Ohmmeter that had the capability of taking sample readings at 0.5-second intervals, in order to get a correct measurement. If you tried to use a standard Ohmmeter on today's ECM systems, you would get an improper "mean" average reading instead of full value.

    Another question I have about ECM's is: Do they use non-volatile flash memory, or do they use volatile RAM memory to store trouble codes?
     
  18. Mastiff

    Mastiff 1/2 ton status Premium Member

    Joined:
    Jul 21, 2002
    Posts:
    2,755
    Likes Received:
    48
    Location:
    Tucson, AZ
    [ QUOTE ]
    Another question I have about ECM's is: Do they use non-volatile flash memory, or do they use volatile RAM memory to store trouble codes?

    [/ QUOTE ]

    The trouble codes and "learning" are stored in volatile memory, at least on the older ECMs. I'm not sure about the newer ones. The calibration PROM is non-volatile, either UV eraseable or EEPROM.
     

Share This Page