2009-10-03

Idle Speed Control Valves and safe failure modes

Finally. It has taken me a great deal of wrenching, cursing, and no small amount of money, but my 1992 Miata is finally running the way the good Lord intended, as she rolled off the factory floor in Hiroshima so long ago, the same year I started high-school in rural Connecticut. Back then I had no idea that this car would eventually be waiting for me, more than three thousand miles away and two continents later.

Well... of course, that first statement should come with a caveat or two. My example could do with a new soft top, a new timing belt and accessory belts, really, new rubber all around.

That said, it finally runs right. Yes, I'd say my example had some deferred maintenance. So far I've replaced:

  • Shock absorbers (Koni Sports)
  • Thermostat (was frozen shut)
  • ISCV (failed, probably due to overheating when thermostat was stuck shut)
  • Plug wires
  • Air filter element
  • Clutch slave cylinder
  • Brake pads and fluid
  • Battery (Duralast 8AMU1R)


Is that really it? It feels like I've done so much more. Maybe the fact that I've flushed the coolant three times while I was fighting the overheating issue, the fact that I've changed the oil and transmission fluid, spent so much time dealing with the sneaky idle control problem... I feel like I have more into the car than just a simple dollar amount.

The shocks are a good example. Far from being drop-in replacements, the Konis have a 12mm shaft that includes an internal adjustment rod - which of course won't fit through the factory shock mounts. So, I had to drill them out... under my carport, without the benefit of a drill press or even a bench vise. I have to admit that it's high time I set up a proper machine shop.

My latest adventure: replacing the idle speed control valve.



The old ISCV is on the right, with air intake tubing attached, and the new ISCV is already mounted on the throttle body. In the Miata (and I assume other Mazda cars from that era) the idle speed control valve can open to allow a calibrated amount of air past the throttle plate at idle to compensate for load. The best example would be the air conditioning compressor. Of course, as good engineers would, Mazda's realized that it would be safer for this valve to fail open. Far better to have the car able to run the A/C at a light if the solenoid actuating the valve fails. The alternative? The engine would either stall or do horrible things to itself at 200 rpm. So, the ISCV fails open, and your car idles at an annoying 1800 rpm. Yeah, that's how much extra throttle you need from a 1.6 engine to run a tiny air conditioning compressor.

Well... mine failed. Since my original ISCV would stop working as the engine reached operation temperature (later ceasing to work at all, cold or hot) my best guess is that the overheating problem I inherited with the car resulted in an broken conductor within the solenoid housing. If I can muster up the motivation, I'll grind open the housing and see if my suspicions are correct.

Either way, the 'new' one works perfectly. My baby now idles quietly and without fuss right around 800 rpm. It's beautiful. Even the valve assembly itself looks newer, which is strange as it was pulled from a '91. Either it was a replacement piece itself - possible - or that particular NA was parted out early on in its life. The $50 in cash from my local specialist was a lot less painful than the $500 new replacements fetch online.

So... my baby is now back to some kind of baseline. Handling and engine performance are 'stock-like'. This winter, it'll be ready for some proper modifications. I haven't decided exactly what I'm doing, but a manual steering rack is definitely going on, and the power steering hardware coming off. The steering is precise, but much too light.

One thing: if you take off the ISCV for inspection or cleaning, you probably won't be able to re-use the gasket that seats into the throttle body. At least, I couldn't. The gasket seated (I think) and promptly disintegrated.


I'll note that the gasket had swelled badly from being cooked for several hundred miles, and the rubber must have been pretty badly deteriorated. Since I was unable to source a replacement gasket, this time I went with a rubber/cork one that I cut by hand. I was considering a silicone sealant, but I'm weird and I'll probably want to pull the throttle body for inspection, at which point it'll come apart without a mess.

In retrospect, I should never have pulled the old ISCV for cleaning. It didn't do any good, and the gasket failure wasn't the best thing in the world to happen to my engine. Those missing pieces of gasket were nowhere to be found, which meant they probably went through the engine. That's not as big a deal as the rust on my old ISCV, which makes me think that small amounts of coolant were leaking into the intake.

Not good. So: don't do what I did and attempt to re-use that gasket. It wasn't broken for very long, though, and I think my car is ready to forgive me.

No comments: