Checking a Honda 3 pin IAC valve [Archive] - Honda Odyssey Owners Club Forum

: Checking a Honda 3 pin IAC valve

07-06-2015, 12:39 AM
I recently had one of those "perfect storms" happen to my wife's 1999 Odyssey, where 2 things break at the same time by coincidence, making the troubleshooting process very difficult. I thought I would share some info on this forum because I could not find the info I needed after searching the internet for hours.
That info I couldn't find pertains to 2 unrelated engine parts:

1) How the sparkplug tubes are secured to the heads
2) The electrical resistance between the terminals of the Idle Air Control (IAC) valve.

My experience started with a rough idle which progressed to the car running very rough over about a week's time of daily driving. Also, the MIL came on. I read the codes which indicated a misfire in multiple cylinders. So I got some new spark plugs. But when I was pulling the coils off, the one on cylinder 6 broke. The top of the coil separated from the part that goes down into the tube. So I got a new one of those. But when I tried to get the rest of the old coil out, I couldn't. It wasn't just stuck in there, it was literally fused to the sparkplug tube! I tried everything to get that sucker out, and ended up grinding it out with a die-grinder. But while doing so, I ground through the tube... oops. Knowing that ground metal filings and crap were now in the cam and rocker area of the engine, I took off the intake plenum and rocker cover to clean it all out and to make it easier to remove the tube that I now needed to replace. BTW, you cannot buy the sparkplug tubes as they are considered part of the head. Good thing I have a metal lathe. So I made a new tube. I didn't want to destroy the head, so I searched for info on how the tubes were held in, but could not find that info, so I assumed they were press-fit... and they are. When I got the old tube out (I had to weld a slide-hammer to it) I discovered what caused the coil to fuse to it; the sparkplug was loose. I mean reeeeally loose... as in, about to fall out loose (I am not the one who put those plugs in, the previous owner did). But it was still firing that cylinder for who knows how long, causing fire and heat to leak past the plug into the tube which melted the coil to the tube. So with the problem now solved, I put the engine all back together with the new coil, tube and all new sparkplugs, and did an oil and filter change after cleaning all the grinding crap out with kerosene in a sprayer. Started the engine and it now ran smooth, BUT it wanted to idle at 2000+ rpm. The ECM won't let it idle that high, and the result is the engine revving up to about 2000rpm and then dropping down to about 1000rpm over and over. I did some research and learned this is because the ECM sees that the throttle is closed (TPS sensor) but the engine rpm is over 1600, so it figures that the vehicle is decelerating. During deceleration the ECM turns off the fuel injectors to save gas. So the rpm goes up over 1600, ECM cuts the fuel, rpm drops below 1600, fuel is turned back on, and this cycle continues indefinitely. And when you drive around like this, at stop lights everybody thinks you are a minivan driver that wants to race!:cool:
So at this point, I figured I had a vacuum leak because I had missed a hose while hooking everything back up, or possibly caused a brittle hose to crack while working on the engine. Spent another hour or two searching for the leak, but could not find one. So I adjusted the idle speed screw all the way down, but the problem persisted. Next thing to check was the IAC (Idle Air Control) valve located on the bottom of the throttle body. It allows the ECM to change the idle speed by controlling a small rotary valve that connects to an air channel that bypasses the throttle plate. I found that Honda uses a different type of IAC than other manufacturers. Most IAC valves have 2 electrical connections which will allow the ECM to control the rpm by varying the voltage applied. Honda uses these also on some of their other engines, but not on my 1999 VTEC 6cyl. These engines have a 3 wire IAC that is PWM (pulse width modulation) controlled. This method of control is sometimes referred to as duty cycle. I searched and searched the internet for info on testing this IAC valve, but found only erroneous info from people who thought they knew what they were talking about. I had a feeling these people were full of $#!+, and it turned out I was right.
Without any information on testing this IAC, I decided to take a chance and buy a new one anyway for about $90. Before installing it, I tested the resistance between all the wires so I could make that information available to others on the internet. After installing the new IAC valve, adjusting the idle speed screw back up, and clearing the codes again, the engine runs as it should. :D What an ordeal! What started as (what I thought would be) a simple plug change, turned into a multi-day frustration laced with anger and cussing. I say this as if I'm surprised, but in reality I know that small jobs often end up being big ones when it comes to vehicles.

I do not know what circuitry is inside the IAC, so I don't know if resistance is a valid method of checking it. But I do know that in my case, I could definitely identify my old IAC as bad since 2 of the wires showed full continuity (a short circuit) where the same wires of the new one did not.

And here is the information you seek regarding testing the Honda 3-wire IAC valve:
Measurements were taken with my BK precision 388A multimeter which I checked for accuracy on a 10Kohm 1% resistor which showed 9.98K.
I'm calling the pin farthest from the throttle body pin 1, and closest pin 3.
If diodes are in there, polarity matters (checking resistance requires that the meter apply a small voltage which means there can be a difference in the reading if the test leads are reversed when semiconductors are present).

New replacement IAC valve manufactured by BWD in the US (p/n 31072):
(-)1 to (+)2 = 9.9K
(-)1 to (+)3 = 2.21M
(-)2 to (+)3 = 2.18M

For comparison, here are the readings from the old broken Denso IAC.
(-)1 to (+)2 = 0 ohm (shorted)
(-)1 to (+)3 = 59.1K
(-)2 to (+)3 = 59.1K

Reply if you wish, but I will not be following this thread.

08-02-2016, 02:17 AM
Very useful information!
192.168.o.1 (

08-11-2017, 06:20 PM
Thanks a lot for these info!
movers baltimore md (