By on February 1, 2019

1988 Honda Accord, Image: OP/Grassroots MotorsportsBrian writes:

Sajeev,

This is a weird one, but I figured you would enjoy it. I have owned an ’88 Accord LX-i five speed hatch for a while. One day, driving along, I noticed that it seemed to be coasting easier than normal. When I approached a red light, I found out why: 3,000 rpm was my new idle speed. I stopped, and before I could even think of why this was happening, the idle returned to normal. Once underway, 3,000 rpm was again the new idle speed. Subsequently, I tried many things. This is not related to the brakes, not related to the throttle input, not related to absolutely anything other then wheel speed. In the most stark example, idling on a slight incline, I can just release the parking brake and, once rolling, the idle jumps to 3,000 rpm. Using only the parking brake to stop once again, the idle returns to normal. No CELs or anything else strange happens during this.

I found that it would idle normally if I disconnected the IACV. This worked fine, but when using the A/C it can no longer compensate, so that was not ideal. I also could make it work if I disconnected the speedometer cable, so I did that for a while before really missing my speedometer and cruise control. I tried another way, which was to disconnect the electrical connections between the speedometer and the rest of the gauge cluster. This works, but I get no cruise control, and a CEL only if I coast with no throttle input for too long, which is strange.

I have tried bleeding the IACV, replacing the IACV, replacing and adjusting the throttle position sensor, replacing the entire gauge cluster (which had the same issue, but seemed to change the high idle RPM weirdly enough, but still wonky-high).  Also, I did check all grounds and the solder joints in the ECU.

Here is my long standing build/upkeep thread, and here is a terrible video.

Sajeev answers:

Damn son, your comprehensive diagnosis/repairs make it tough to wave my magic wand and proclaim a resolution! But I bet you either:

  1. Tragically bought a defective throttle position sensor (TPS).
  2. Didn’t adjust said TPS correctly.

Your symptoms remind me of a looney month with a 1987 Mustang GT convertible where everyone thought I was rearin’ for a race: whenever I engaged the clutch/popped it out of gear (after warming up) the 5.0 would rev around 4,000 rpm. No check engine light, no joy when swapping the IAC (IACV in Honda-speak), but it fixed itself after swapping a working TPS.

While it appears that installation on your Accord is harder (video NSFW-ish), while there’s plenty of digital ink’s spilled over Honda TPS tuning and while I’d never doubt your skills…do re-check your work on the throttle position sensor.

What are we missing? Tell us, Best and Brightest!

[Image: OP/Grassroots Motorsports]

Get the latest TTAC e-Newsletter!

Recommended

32 Comments on “Piston Slap: According to the Throttle Position Sensor…...”


  • avatar
    gtem

    Maybe I missed something but this 100% sounds like a vacuum leak from the intake manifold? The fact that applying the brakes (power brakes with booster pulling vacuum from said manifold) affects things makes me believe this theory even more. Let the car idle, open the hood and spray around with some starting fluid, or an unlit butane torch (just the gas feed). If the RPMs shoot up even higher, you’re getting close.

    I LOVE the 3rd gen Accord hatchbacks, LXi, 5spd, be still my beating heart! My favorite Accord hands down, possibly favorite Honda period.

  • avatar
    EAF

    OP, if your old Honda is set-up anything like my old Mitsubishi, I would *guess* the issue is with the FIAV (fast idle air valve). It is either stuck open or coolant lines leading up to it are clogged.

    Engine cold starts – FIAV is open – high idle for warm-up – coolant builds heat – melts wax in FIAV – spring gradually shuts FIAV – engine idles down.

    Either the valve is stuck open (may be able to clean it) or coolant is not freely traveling to it. Goodluck!

    • 0 avatar
      EAF

      Brian, I carefully re-read what you wrote. A stuck FIAV would not cause a raised idle by simply coasting in neutral, idle would be raised whether coasting or standing still.

      Guess #2; some ECUs have logic in them based on the speed sensor in the gearbox. While moving, based on speed, the ECU may keep the throttle slightly open to prevent a coast stall. It LIKELY controls this using the step motor (IACV) that you’ve already replaced. As Sajeev stated; if it can be set, perhaps the idle speed or the TPS settings are slightly off. Maybe the new IACV is defective out of the box – I’ve had this happen as well!

      Tough one Brian!!!!

      • 0 avatar
        gtem

        Yep, unless it’s OEM Honda, I would absolutely be suspicious of any new part like an IACV.

        • 0 avatar
          dukeisduke

          Especially if it’s an AutoZone (ValuCraft) part. Chinese-made junk. I helped a friend replace an ignition coil in an early ’90s Civic. He bought a ValuCraft coil, and the thing was so oversized we practically had to hammer it into the distributor cap. Absolute garbage.

          • 0 avatar
            JohnTaurus

            I haven’t bought from Auto Zone for years because either I get the wrong part, the part is defective off the shelf, or the part fails in short order. Went through 3 coil packs on a 1997 Taurus until I demanded my money back and bought one elsewhere. The car is still running with that one 3 years later.

            F’n hate auto Zone.

        • 0 avatar
          Tuna55

          it was known-good-used

    • 0 avatar
      Tuna55

      The engine temperature doesn’t change the idle in any way. I checked coolant flow and cleaned the FIAV already. This still would not explain why the speed affects the speed.

  • avatar
    StudeDude

    I would check the power brake booster and check valve for the booster. It’s unusual but possible that the problem is coming from that direction, especially since it’s a 31 year old car.

  • avatar
    dukeisduke

    Shouldn’t you just be able to test the TPS with a VOM, by hooking up the leads to the pins coming off of the TPS, and moving the throttle while watching the resistance value?

    I’ve only ever replaced one TPS, on my old ’95 F-150. It was on the bottom of the throttle body, and I managed to replace it without disconnecting the coolant lines (TB heater) at the TB. It had some adjustment, but I just eyeballed it and placed the new one like the old.

    The old one wasn’t bad (it had 66,000 miles), but several people told me that they usually failed before 75k, with the normal failure mode being stalling at idle. I replaced it around the same time as the IAC, which was bad (it would periodically stick closed on startup, which would cause a slow idle for a couple of seconds, then the thing would pop open, and normal idle would return. It wasn’t an easy thing to diagnose, as the EEC-IV would throw codes 333 and 334 – EGR and EGR valve position sensor codes, because the slow idle would confuse the PCM about how much EGR it should be getting. There were no IAC-related codes in the PCM, so it took unplugging the IAC connector while running at idle to reproduce.

    The 5.0 had a one-piece IAC (solenoid and valve in one unit), so it couldn’t be cleaned. I’d have been in luck if I’d had the 5.8, as those use a two-piece IAC, so you can remove the valve portion and soak it in parts cleaner.

    • 0 avatar
      Tuna55

      Can you find a procedure for that? I was unable to see anything with my DMM due to perhaps the connector style.

      • 0 avatar
        SPPPP

        The best way I have found to check sensors while installed is to buy some long pins (say, 2″) with the plastic balls on the ends. Carefully insert them into the back end of the connector housing, alongside the wires, but be careful not to tear up the weather seals (if equipped). Slide them in gently until you feel them contact the metal connectors inside the housing. Then you can probe the pins and see the voltage. If you don’t have enough hands to do that and work the throttle body, then use a couple of short leads with alligator clips to keep the electrical test circuit in place.

  • avatar
    cbrworm

    I don’t remember what the EFI was like on those. There isn’t a separate throttle closed sensor is there? What about a throttle closed/idle adjustment that isn’t built into the TPS?

    The throttle body is clean – the butterfly doesn’t have to be held far enough open at idle for the cpu to think it is not closed?

    I have driven cars before that would do something similar, but at ~1,500 rpm and they would drop when your ground speed was just above zero, this was by design – I think to reduce both drive lash and emissions.

    I know the carb’d version of those hondas had crazy vacuum line routing – is it possible there is a vacuum valve stuck open or closed? Possibly something that opens during coast that has a vacuum leak on the other side instead of an orifice.

  • avatar
    SPPPP

    I believe that EFI cars with manual transmissions usually command high idle in situations when the car is rolling in gear and the clutch is out. This is to help you not stall it if you let the clutch back out and you don’t have enough throttle applied to keep the engine running. When you are at a stop and the clutch is out, this isn’t really needed as much, because you expect to have to use the throttle to pull away. It probably also helps to reduce emissions.

    If that strategy applies to this Honda, then consider that this is a flowchart-type control strategy. Speed is one part. The clutch being out is the second part. (The car being in gear is the third part. But I don’t think Honda installed a separate sensor for the car being in gear, so presumably that is inferred by the computer based on speed.)

    Disconnecting the speed input to the ECM (engine computer) would prevent the high idle from kicking in. This would match with the symptoms you have reported.

    So why would this strategy be operating all the time? Could it be that your clutch position switch is faulty or misadjusted?

    I would suggest checking that first.

    Also, you mentioned that the car idles normally with the IAC unplugged. I think that may indicate a slight vacuum leak, as the car should be at a low idle with no IAC. This gives the ECM more control over the actual idle speed.

    To back up the vacuum leak theory, you can try checking the timing advance with a timing light. (Though I am not sure how easy that is on your car.) But what you would be looking for is, does the timing advance jump around at idle? Or is the timing advance stuck at a relatively low number? If it is the first case, then it indicates that the ECM is playing with the timing to control the idle speed, as most EFI cars do. If it is the second case, then it indicates that the ECM is pulling timing to slow down the idle speed, but has reached its limit. This indicates a vacuum leak.

    Oh, and one more thing. Certain types of IAC solenoid have an internal plastic bellows. That eventually cracks and fails, letting air in, and causing high idle. If that has happened, unplugging the IAC won’t make any difference. You said you replaced the IAC. If that was a used part, I would look at it carefully to be sure it wasn’t also bad.

    • 0 avatar
      SPPPP

      Ok, one more last thing. It appears that a lot of Hondas have TWO clutch switches. One is for the starter interlock. Presumably, that switch is working on your car, or you would have noticed. The other switch is for cruise control (and, I would guess, idle strategy). That would be the suspect here. I don’t know for sure if your Honda has two switches, but it would seem likely.

    • 0 avatar
      Tuna55

      This is an interesting reply. As far as I knew, the car didn’t have a clutch switch. It does not have a neutral safety switch. It does have a brake switch for both the lights and the cruise control. I looked on Rockauto and indeed I see a clutch switch! I don’t know where it might live, because I do not recall seeing one on the pedal like the brake switch, but I will now look. I cannt see, however, how that gets me to 3,000 RPM. Before this failure mode I never noticed an increase in idle speed with clutch application. I have not been able to check the problem on/off because it literally returns to normal idle speed before I can get out of the car. Perhaps I can put the front end on jackstands and have a helper in the car while I check the timing. If it’s in reverse, it’s not -that- dangerous.

  • avatar
    pwrwrench

    This is another one of those times where having the OEM workshop manual should help.
    For a 1988 Honda someone may have posted the manual online somewhere. Or you might find a good used one on craigslist or fleebay.
    It may take some careful reading, but there is likely full info on the idle system.
    Others have made good suggestions: Replacement parts that are non-functional (IACV, TPS), Vacuum leaks, TPS out of adjustment range.
    Making a full diagnosis with the Honda service info will probably give the best and quickest fix.

    • 0 avatar
      Tuna55

      AS far as I knew, the IACV and TPS were known good used. I swung the TPS from max.min adjustment and wasn’t able to see a change. I know that’s not a real adjustment, but I figured it meant that the ptoblem was elsewhere.

      • 0 avatar
        pwrwrench

        Someone mentioned a possible malfunctioning oxygen sensor. That would be a good place to look and at the wiring from the sensor to the PCM. What I found in 20 years of working on cars with a Check Engine light was that things could go bad, often wiring or connections, but not have a fault code in the OBD. Or the code would have little to do with the problem.
        Also look for mechanical problems. There was a mini-van that had a intermittent stalling problem. Many parts in the FI and ignition systems were replaced. The actual problem was a loose valve seat.The cylinder did not have low compression. Until it came completely out and stuck the valve open.
        My thought is one has to have an open mind. The shop manual should show all the things that affect the idle speed. You might have to study the wiring diagram to find some. Some cars have a power steering pressure sensor that is in the loop. I think this was also mentioned by someone.
        I have seen people replace many parts only to find a bad connection at a sensor. Since the idle problem seems to be related to when the car is moving checking the Vehicle Speed sensor and its wiring is another place to look.

  • avatar
    ryanwm80

    Sajeev is usually spot on, but you need to know for sure it’s the TPS and not something like a sticking throttle body or a vacuum leak. I would scan the ECU even though the CEL isn’t on. I don’t know how Honda’s OBD-I system worked, but with the Ford OBD-I it was possible to have a trouble code even though the check engine light was off, and I had this happen with my 1988 Mercury Sable LS – there were situations in which the engine ran poorly (no throttle response or a rough idle) but most of the time it was OK, and the check engine light was off. I tried scanning the computer and got a trouble code for no response from the O2 sensor, so I installed a new one and it’s ran like new since.

    • 0 avatar
      Tuna55

      “Checking codes” is a blinky light. No light when the idle issue happens.

      The throttle body is never used during these events, and is not sticky in any way regardless.

  • avatar
    Michael S6

    Had the exact problem with my 1988 Acura Integra where the rpm shot up at random times. It took 3 trips to the dealer to finally fix the problem. Unfortunately, 30 years later I don’t recall the diagnosis, but I think that it had to do with cold weather start setup.

  • avatar
    Art Vandelay

    I don’t know what is wrong with your car…but looking at the clean design vs. current models I can see everything that is wrong with new Hondas.

  • avatar
    maryland1966

    I also have no clue same as the two guys above. Not sure why I posted but wanted to note that I like the color of this vehicle.

  • avatar
    ToolGuy

    Brian/Tuna,

    Are you still working this problem?

    • 0 avatar
      ToolGuy

      Brian/Tuna,

      Step 1: Re-read “Zen and the Art of Motorcycle Maintenance” (yes I’m serious)
      Step 2: Take a nap (yes serious)

      EAF and SPPPP talked about the ‘command high idle’/anti-stall functionality and purpose.

      From your video, the computer on your car is commanding high idle. It is doing it successfully. It is just doing it in circumstances that you don’t want/need. Put yourself in the place of the computer. It can command the throttle just fine – stop looking there. Most of the signals it is getting, seem to be fine – put those on hold.

      *Something* else is going on. My suggestion: Get a good old-fashioned dial vacuum gauge. “T” it in so the car still runs ‘normally’ (closed system, no hoses unplugged). Use a long enough hose to the gauge that you can view it from the dash while you drive (could tape it dial-side-in to the base of the windshield, or could run the hose through the window) – and of course the hose should be sturdy enough that it doesn’t collapse under vacuum.

      Now watch and listen. What is the computer ‘seeing’ in terms of vacuum, and under what conditions? Is there any correlation/connection between the vacuum gauge readings and the ‘command high idle’ condition?

      Some of the best diagnostic advice I ever got was from a friend’s dad: “Look for horses, not zebras.”

    • 0 avatar
      ToolGuy

      Also curious after scanning again through some of the suggestions – have you ever done a dry/wet compression test? (Extremely helpful for peace of mind and ruling out a bunch of stuff.)

Read all comments

Recent Comments

  • 07NodnarB: The Hummer…what do I think of it you ask? I don’t. I don’t think about it. That’s...
  • RHD: Everything is questionable. That’s one of the basic tenets of philosophy as well as science. However,...
  • RHD: There are hundreds of millions of tons of mine tailings that are waiting to be reprocessed. New holes...
  • RHD: That idea is what powered mechanical toys a century ago. It’s not a new idea, but certainly has potential...
  • Old_WRX: cartunez “I absolutely love the political comments on this site.” Yeah, where else can you find...

New Car Research

Get a Free Dealer Quote

Who We Are

  • Matthew Guy
  • Timothy Cain
  • Adam Tonge
  • Bozi Tatarevic
  • Chris Tonn
  • Corey Lewis
  • Mark Baruth
  • Ronnie Schreiber