NA Idle Problems

Original Author: Phil Barnett (prb) - 2013

Back to Maintenance


One of the common issues Miata owners face when their cars begin to age is idle problems.

The idle problem the engineers needed to solve is one of unknown variable load. To solve this problem, the engineers introduced the ISCV or IACV. ISC and ISCV are short for Idle Speed Control Valve and IACV is an acronym for Idle Air Control Valve. When you see either of these two terms, they are referring to the same part of the system.

Sensors on your engine give the computer advance warning of variable load so it can make better and faster decisions. All of the sensors for IACV are related. A/C load. Alternator load. Power steering load. Transmission/clutch load. Fast throttle up (this is a violent change in air velocity in the intake manifold). The loads they present are all different and each compensation factor is built into the firmware of the computer.

Connecting GND to TEN at the diagnostic connector parks the IACV valve to a known spot in it's range. Once it is parked we can adjust the air bleed screw while the computer is ignoring all of the sensors. This sets the system so the compensation factors built in the computer firmware have a real world baseline so it can make correct decisions.

As we have witnessed, the computer is able to adjust the idle rpm and maintain it there. The only control mechanism is the IACV.

So we come to a decision fork: Input problems (bad sensors or polluted input), decision problems (bad computer) or execution problems (bad IACV).

Is it the computer, the sensors, the IACV, a bad adjustment of the bleed screw or a vacuum leak? Maybe inaccurate engine temperature sense. A vacuum line leaking air in somewhere. A bad PCV valve.

One thing to note: The computer in these cars is not only the most expensive part of the idle control system but it has been incredibly reliable over the decades since these cars were built. So if you decide to blame it, change it last.

If the computer makes bad decisions from bad input, the IACV follows the bad instructions. This sometimes leads us to blame the computer when the computer is fine and is acting on invalid input from it's sensors.

Did the IACV not return to park (stuck or failed) so we properly adjusted the air bleed screw? Is there a vacuum leak which changes the range of things and the computer makes not so good decisions. Any variety in setting the base idle could set up the IACV to hit the end of it's range before it completes the mission.

Many of the sensors are just on/off switches. Intermittent sensors lie to the computer. We can test switches. Testing them while the car is rolling is only marginally more difficult than static testing.

From what I see, the only variable load is the alternator. I believe the power steering sensor is a switch, but I've never measured it so I'm not sure.

None of this is a solution. It's just me trying to make as full of a description as possible so the problem can be conquered in a structured way.

Understanding a problem often leads to it's isolation and thus it's solution. And if we have to shotgun the problem, at least we can determine that list in ascending order of cost

For the record, many people have cured this problem by replacing the Idle Speed Control mounted under the throttle body.


Generally speaking, you find vacuum leaks by doing the following:
  • Connect TEN to GND to turn off the computer controlling the idle speed.
  • Disconnect one vacuum line at a time from the intake manifold and close the leak with your finger.
If the engine changes speed between hose attached and finger closing the leak (ignore that moment it's leaking wide open) you have a leak in that vacuum circuit. Generally the engine will run at a lower RPM after the vacuum leak is fixed and you'll have to let more air in the idle air bleed screw (turn it counterclockwise)

I would attack the problem in the following order, least expensive to most expensive.
  • Not the least expensive, but if you have a 1990 or early 1991, you may have a Short Nose Crank. These are known to fail and cause engine running issues. Search engine use with the words MIATA SHORT NOSED CRANK should get you more reading than you wanted.
  • Check the clean air crossover tube for leaks and cracks.
  • Adjust the throttle cable to remove any excess slack. Be sure that the throttle still closes completely. There should be a very small amount of slack with the throttle plates closed.
  • Check for vacuum leaks as stated above. If you find one, fix it and then set the Base Idle. Keep going until you have eliminated every vacuum circuit.
  • Disconnect and block the purge vacuum line to the charcoal canister. If it runs better with it disconnected the canister may be full of bad fumes. Also, the purge valve should not allow vacuum to the canister at idle rpm, so if that line has a vacuum on it, you'll need to replace the purge solenoid.
  • Set the Base Idle to 850 rpm with TEN and GND connected at the diagnostic block.
  • Set the ignition timing to factory spec. This will interact with the idle speed so you'll have to go back and forth until you get both of them correct.
  • For the NA6, the factory spec for setting the dash pot is to adjust its screw until it just touches the throttle pulley flange at 2500 rpm. Lubricate it while you are there.
  • Remove, Clean and Tighten your ground points.
  • Make sure you have the correct injectors installed. Blue for 1.6, tan for 1.8.
  • Check the clutch pedal blue plastic rivet is in place.
  • New PCV Valve.
  • If no automatic transmission, check the plug on the back of the manifold for leaks. If automatic, check the modulator for diaphragm leaks.
  • Check for vacuum leaks. You can pinch off small tubing with a pair of pliers. If you can pull a vacuum hose from it's nipple with little friction, that hose needs to be replaced. You can use an unlit propane torch to locate the source of a vacuum leak. When the propane arrives at the leak, the engine will change speed.
  • Check the Injectors for air leaks or replace the injector seals.
  • Test the Brake Vacuum Booster for leaks from the point it attaches to the intake manifold. (use something like a Mity-Vac to see if it can hold a vacuum)
  • Check that the throttle plate is returning to the stop.
  • Check for a binding throttle cable.
  • If you have a 94-97, clean the MAF sensor with an approved cleaner.
  • Test and adjust the Throttle Position Sensor.
  • Check for loose connections to all the sensors and the computer.
  • Test all the sensors to see if they are actuating.
  • Test that coolant is flowing through the ISCV.
  • New spark plugs and wires.
  • Check the fuel pressure and see if the fuel pressure regulator vacuum line is wet. If it is wet with fuel, the FPR is bad.
  • Replace the engine thermosensor that has multiple wires in it's connection. The one that goes to the ECU.
  • Clean the ISCV. Instructions are in the Miata.net Garage. New throttle body and ISCV gaskets.
  • It could be an internal leak. Remove the EGR valve and test it for leaks. If you can blow through it, it's bad. This is a vacuum leak you can't hear when the engine is running.
  • Replace the EGR valve.
  • Replace the alternator or have it checked by an alternator shop. Ask them to insure the diodes are all good. If one diode goes bad, it can put noise on the 12V system which confuses the computer. Alternators have fixed quite a few idle issues.
  • Replace the ISCV. If you haven't replaced them yet, new throttle body and ISCV gaskets.
  • And I hesitate to add, replace the engine control computer. Obviously, it would be best to do a substitute test before laying out this kind of coin.