rough idle/hesitation at cold start - fuel pressure regulator issue?

Discussion in 'Injection Tech' started by Theodore, Jul 10, 2020.


  1. Theodore

    Theodore Farmall Cub

    Joined:
    Dec 2, 2019
    Messages:
    7
    Likes Received:
    0
    Trophy Points:
    1
    First-time poster/long-time lurker. Am working on a 1990 Grand Wagoneer (AMC 360) project that already had EFI (GM TBI) installed on it when we got it. Have been working thru running/driving issues ever since, and have it the best it's ever run, with the exception of a couple issues:
    1) Rough idle/hesitation at cold start for a minute after startup
    2) Backfire at high RPM at highway speeds, when running at 70 mph, let off throttle, decel a bit & get back in it; can hear & feel backfire thru the exhaust - sounds like popcorn popping.

    Am running Affordable's GM TBI fuel injection system, along with 1320electronics.com bluetooth OBD1 scanner & have tuned initial timing (10 deg), base idle(750), along with IAC & TPS. Have found/fixed numerous - having replaced all vacuum lines in the engine bay, and added a Schrader valve to the fuel supply line, about 12 inches back from the throttle body. Am thinking i may have a fuel delivery issue.

    When cold(sitting overnight), fuel pressure at prime shows 12 lbs.; once primed, the fuel pump cuts off, and pressure instantly drops to zero. The pressure drop seems key to the issue, as it seems like it should hold pressure. To figure out which direction to look (forward or backward), i've run a few tests:
    1) inspected the fuel injectors for leaking - so far, ive seen a nice pulsing spray pattern, with no leaks/dribbling.
    2) Changed fuel filter - saw no real change.
    2) Crimped the supply line between the fuel pressure gauge & the throttle body, and primed the system: to see if the fuel pressure drops indicating an issue with fuel pump, or its check valve. Pressure rose to 60 lbs & held - indicating the issue is on throttle body side of the equation.
    3) crimped the return line near the throttle body & primed the system; & saw the pressure rise to 60 lbs & hold, which which makes me think fuel pressure regulator. Have installed a FPR rebuild kit a few weeks back, and saw no change.

    Am stumped. It seems like the fuel line should hold pressure at prime & not instantly drop off. Everything i read says, that seeing a pressure spike when crimping the return indicates an issue with the fuel pressure regulator, and crimping the supply ahead of the throttle body proves the
    fuel pump check valve is good.

    Where to go from here?
     
  2. Bill USN-1

    Bill USN-1 Moderator Staff Member Moderator

    Joined:
    Dec 25, 2002
    Messages:
    17,130
    Likes Received:
    292
    Trophy Points:
    83
    Location:
    Whidbey Island, WA
    I don't see any data logs or BLM tables to see where your tune is.
    The system does NOT retain pressure when the pump is off.
    It has bleed holes in the throttle body to bleed of pressure and air should a vapor lock occur.
    So there is no way to maintain pressure and has nothing to do with your issue.
    There's a LOT of bad information on the web.
    I have all the basics posted on this site.
    basics in timing table and fuel tuning in separate FAQ threads and in the step by step FAQ.
     
  3. Theodore

    Theodore Farmall Cub

    Joined:
    Dec 2, 2019
    Messages:
    7
    Likes Received:
    0
    Trophy Points:
    1
    Thanks, Bill - glad to have that myth busted; have read many conflicting reports there.

    Thanks for being willing to look at a data log; have uploaded one for review. Log includes a cold start after sitting overnight, a slow drive thru town to the Interstate at speeds up to 70mph. Engine exhibited rough idle at start for the first minute or so, and saw the backfire at speed during the last minute or so of the log. The Grand Wagoneer is stock AMC 360, TF727 transmission.
     

    Attached Files:

  4. Bill USN-1

    Bill USN-1 Moderator Staff Member Moderator

    Joined:
    Dec 25, 2002
    Messages:
    17,130
    Likes Received:
    292
    Trophy Points:
    83
    Location:
    Whidbey Island, WA
    The system needs tuned for your engine.
    Your charging system is low.
    Could be the charging system or where you're powering the EFI from.
    You should be 13.8-14.2v
    check out www.madelectrical.com
    Attached is the BLM table as discussed in all my tuning FAQs.
    You're lean at low rpms and then it get better as rpm goes up.
    I also attached the log file in excell.
    Capture.JPG
     

    Attached Files:

  5. Theodore

    Theodore Farmall Cub

    Joined:
    Dec 2, 2019
    Messages:
    7
    Likes Received:
    0
    Trophy Points:
    1
    Thanks for taking a look. Based on the advice, i dove into tracing wires & while there were a few splices visible, there were more buried within the wire loom.

    When i started:
    - battery voltage at stop - 12.95
    - battery voltage at idle - 14.5
    - battery voltage per ECM at idle - 13.0

    - cleaned the battery to body ground: got .2 volts running
    - found & cleaned ECM to engine ground: no change
    - traced the ECM's power(pink) wire that goes to the coil. Tracing found multiple splices, with wire colors changing 3 times by the time it reached the coil. Cut all that out, respliced in a single wire, and soldered, and noticed it seemed to run better & i saw 13.7 per ECM at idle on Sunday evening - woohoo!
    - there's also a 12 gauge wire running to the (+) battery terminal, which also has a splice - am seeing no voltage drop on it, so have left it for now.

    Took it out for a test run last night, and noticed the ECM was reporting high-12's/low 13's again - dangit. Even at 13.2V, i noticed the engine ran better, and backfiring was significantly reduced, as well as, power improved climbing a steep hill on the Interstate here. Definitely think we're headed in the right direction.

    Am wondering where the ECM monitors 'battery voltage' from? Reason i ask, i've not yet found a wire that measures the same voltage as what the ECM is reporting. Am either seeing values in the 12.6 (at the coil), or 14+, but nothing in the 13 range. Thinking out loud, i think i need to test the positive battery terminal vs. ECM ground wire. Will take a look at it later on today & report.

    Have also uploaded a data log from last night's test run - similar to before. Cold start, with a run thru town to freeway.
     

    Attached Files:

  6. Bill USN-1

    Bill USN-1 Moderator Staff Member Moderator

    Joined:
    Dec 25, 2002
    Messages:
    17,130
    Likes Received:
    292
    Trophy Points:
    83
    Location:
    Whidbey Island, WA
    Keep in mind that the system adjusts and learns as you drive in closed loop.
    Any time you disconnect the batt or ECM it clears that memory.
    So it will improve the more you drive.
    That's why i custom tune all my systems, so it doesn't have to relearn every time you disconnect power.
     
  7. Theodore

    Theodore Farmall Cub

    Joined:
    Dec 2, 2019
    Messages:
    7
    Likes Received:
    0
    Trophy Points:
    1
    Hey, Bill, took a look and moved the ECM ground to a place where I could measure voltage with my multimeter - to one of the thermostat bolts, cleanest contact cleaner. I feel like I'm missing something. Back-probing the ECM positive connection at the battery, and the ground terminal at the thermostat, I see 14.6V. The ECM reports 13.2V bAttery voltage. Does this make sense to you?
     
  8. Bill USN-1

    Bill USN-1 Moderator Staff Member Moderator

    Joined:
    Dec 25, 2002
    Messages:
    17,130
    Likes Received:
    292
    Trophy Points:
    83
    Location:
    Whidbey Island, WA
    Not being there I can't answer specifics.
    BUT
    open voltage is different from a load.
    You can connect wires to the + and - post of the batt and 13v.
    But connect a fan or radio or anything else to those wires and the voltage will change.
    With the engine running you "should" measure 13.8-14.2v at the battery.
    Depending on the load, the alternator may need to put out more than that.
    So make sure charging voltage is correct and the rest should follow.
    www.madelectrical.com for tips on the 10si sense wire
     
  9. Theodore

    Theodore Farmall Cub

    Joined:
    Dec 2, 2019
    Messages:
    7
    Likes Received:
    0
    Trophy Points:
    1
    Am wondering if we're not saying the same thing. I pulled another data log tonite from Jeep idling in the driveway for a while, after having cleaned the ECM power connections (positive at battery positive; ground at thermostat housing). Based on the comments above, it seems the charging system is working well with voltages consistent at a few different locations, and no voltage drop on ECM power wires.

    Engine not running:
    - Battery voltage per multimeter: 12.86
    - Voltage at ECM connections per multimeter: 12.86

    Engine running at idle:
    - Battery voltage per multimeter: 14.44
    - Voltage at ECM connections per multimeter: 14.44
    - Battery voltage per ECM: ~13.0

    ? - Does it make sense i seem to be missing 1.44V? 14.44V vs. 13.0V at idle? or, does this seem right to you?

    Thanks.
     

    Attached Files:

  10. Theodore

    Theodore Farmall Cub

    Joined:
    Dec 2, 2019
    Messages:
    7
    Likes Received:
    0
    Trophy Points:
    1
    Am writing to follow up: Do you know which connections the late 1980's GM TBI ECM reads 'battery voltage' from? I've got the ECM 'pink' power-on wire going to the positive terminal on the coil, which in turn is fed from the BAL (ballast resistor) terminal of the starter relay. In testing voltage w/a multimeter, the BAL terminal shows 12.5-13.2V with the engine running. At first, i thought i'd found the issue, but it turns out that's not where voltage is read from - keep reading...
    As a test, i disconnected the ECM 'pink' power wire from the positive coil terminal and reconnected it to the positive battery post to see what the ECM saw - no change: battery voltage reported in data logs still showed 12.5-13.2V, whereas battery voltage measured via multimeter shows 14.6V.

    It seems like I'm missing something here, but i'm not sure where to look.
    Appreciate any insight anyone can provide.
    Thanks.
     
  11. Theodore

    Theodore Farmall Cub

    Joined:
    Dec 2, 2019
    Messages:
    7
    Likes Received:
    0
    Trophy Points:
    1
    Following up - making progress on the low voltage issue over the last couple weekends. I've been able to clean up the wiring (more on that below), and have seen voltages rise as measured by the ECM as changes were made; moving from ~12.5V (logs showed ranges 12.2V-12.8V) to a high of 14.2V now. With all that, i still feel like i'm missing something here - voltages measured by multimeter show consistently higher by .2-.4V than what the ECM reports for Battery Voltage, and while the engine is running better than it ever has, we still have the base issue of hesitation at cold start, intermittent dying when going into gear, and high speed backfire when letting off the throttle at 70mph, coasting to 60, and getting back in it. Right now, am focused on getting the wiring fixed, and raising the battery voltage measured by the ECM as high as possible.

    When measuring battery voltage w/engine stopped & running, i see nearly no voltage drop between battery voltage tested w/a multimeter and whatever terminal i'm testing at the ECM, but the ECM reports .2V-.4V less than that. Am wondering if i've got a bad connection in the ECM harness connector?

    For example,
    ------------------- Engine ----Engine -----Measured per ECM
    ------------------- Stopped --Running ----& shown in logs
    ------------------- ---------- ---------- -----------------------
    Battery Voltage 12.5 ------ 14.5-14.6 - 14.1-14.2
    ECM Pin A6 ----- 12.5 ------ 14.45
    ECM Pin B1 ----- 12.5 ------ 14.45
    ECM Pin C16 --- 12.5 ------ 14.45

    Over the last couple weeks, i've broken apart the wiring harnesses (factory & EFI) & run voltage drop tests, finding a myriad of bad splices/ connections along the way. As part of that, i've cleaned battery terminals, cleaned all the grounds i could find, found and connected a previously
    unconnected 'green' 12-ga ground wire that was connected to the negative battery post - stuffed in the headlight hole. All of this got me from ~12.5V measured at the ECM to ~13.2.

    After reading & re-reading your posts, and talking to AFI & babywag on fsjnetwork.com, i worked my way thru the wiring harnesses to the ECM itself, the main relay & fuel pump relay & began cleaning connections there with DeoxIT D100 contact cleaner. Corrosion was pretty amazing, as evident from the bubbling. That's when i found a major cause of voltage drop - the ECM fuse. The mix of corrosion & cracked connection was responsible for ~1V.

    Cracked fuse
    [​IMG]

    Battery voltage as reported by ECM jumped from 13.2 to 14.2, after cleaning ECM & relay connections and replacing the cracked fuse.
    [​IMG]

    DeoxIT D100 electrical contact cleaner
    [​IMG]

    i've also uploaded the latest set of data logs, would appreciate your taking a look & see if anything jumps out. One thing i notice in the logs is voltage drops the longer the engine runs... So far, i'm only driving it in a 6-10 mile loop that includes in-town & highway driving, so am wondering how far it would continue to drop if i drove longer/further? Am planning on taking it out later on this weekend, for a longer test drive & see what happens.

    Right now, the ECM is in mounted in a box under the hood, on the driver's fender where the rear window washer bottle should be. Given the hacked up nature of the EFI harness, am planning on getting a new one & moving the ECM to inside the Jeep - just wanted to get it running properly first.

    Wondering: could engine heat be affecting measured voltage?
     

    Attached Files:

Share This Page