I have investigated this phenomeon very closely. I had a high idle condition on my 87 420sel. Applying a variable voltage to the idle control valve allowed me to control the idle. But when it was connected to the idle speed controller in the car it would not bring the idle low enough. It seemed like the valve idi not have enough authority to do the job. The idle speed controller was putting out the maximum current that it was designed to do. It was just that the valve would not close down enough.
I installed a new one and the problem was solved. However before I installed the new one compared it to the old one on the bench and did several measurements to find the differences between them. I measured current verses stroke and all sorts of other electronic measurements. I could not find any differences!. They performed the same on the bench. I cant explain why one worked and the other did not. I assume it was because of some mechanical wear related to the sealing surfaces or something like that.
In one case I tried to clean the valve using brake cleaner (a good solvent) but it didnt do any good. I ended up buying a new valve and the problem was fixed. However, these valves are way expensive. So on the other car I spent a lot of time cleaning the valve. This time I used carburator / choke cleaner. It did a better job of disolving the carbon junk. I kept working on it until I could not get any more junk out. I was surprised how much stuff came out. I spent almost 2 hours soaking, spraying, and repeating the process. To my surprise, it did fix the problem. I also replaced both of the hoses since they were hard. On hind sight, I should have spent more time on the first one and I would not have had to spend the $ to buy a new one. Live and learn.
However even if the electronics is OK the valve seems to have an authority problem. Let me explain what I mean. It started with my 87. The idle was always hi. About 1200 RPM. I could not get it to come down. When disconnecting the valve connector - the idle went up. When applying 12 volts to the valve - the engine died. I connected a variable voltage source to the valve and was able to regulate the idle speed quite nicely. I was then convinced the problem was in the control module since the valve seemed to work. NOT. The valve had the ability to control the idle, I demonstrated that. So why didnt it work?
The control module was fine. It seemed that it was not able to supply enough current to the valve to bring the idle speed down far enough. Thats what I mean when I say it didnt have enough "authority" to control the idle. It was trying but it did not have enough strength. I bought a new valve and installed it and low and behold everything was fine. So now I was determined to understand what the problem was.
I took the "bad valve" apart - I mean completely apart. I could find nothing. I checked the stroke vs current & voltage signature from the new valve and checked it against the bad one. No observable difference. The bad valve had no observable difference to the new one except it was not shiny. So I accepted the fact that there was still so something I did not understand and I let it go at that until my 91 had the same problem.
This time I reasoned that there was something physical in the valve portion causing the problem. Thats when I got real agressive on cleaning the thing. I worked on that thing for hours until I could get no more carbon out of it. And it worked.
There is another phenomom used to drive these valves. The control module modulates the current applied to the valve with something called "dither". Dither is needed to overcome something called "stickshion" . Stickshion is best described as something like sticky friction. Its where the valve does not want to move shoothly and porportionally with the applied current. Because things get sticky it wants to jump rather than slide smoothly. So the electronics applies a low level rapidly varying current to it to constantly wiggle it and keep it free so it wont stick in one place. You can feel the dither when the engine is running. It feels like the valve is buzzing. This buzzing effect makes the valve move smoothly even though it is full of gunk and other sticky stuff.
My conclusion is that over time these valves get crudded up and wear mechanically and the result is that it takes more current to drive the valve shut. After a while the control electronics cannot supply enough current to bring the idle down and hence the idle creeps up. Cleaning it can cure it if it is not worn out. Dither makes the valve move thousands of times per minute and the resulting wear changes the stroke / current characteristic of the valve to something beyond what the electronics can compensate for. So try cleaning it really well and if that dont work, grab your wallet.
I got too many cars!! Insurance eats me alive. Dave
78 Corvette Stingray - 3k
82 242 Turbo Volvo - Manual - 270k
86 300e 5 speed manual - 210k
87 420sel - 240k
89 560sl - 78k
91 420sel - 205k
91 560sel - 85k
94 GMC Suburban - 90k
97 Harley Davidson Heritage Softail - 25k
00 GMC Silverado 1 ton 30k