Mystery Problem! Can't get idle low enough, Is it a mortal sin to mess with the throttle stop screw?

Engine is b18c1 (USDM GSR) with mild cams very similar but slightly more aggressive than stock Type-R cams.

I've decided to tackle a problem that's been bothering me a while, I can't get my idle as low as where it's supposed to be. It should be 850 rpm. The lowest I can possibly get it to is 600 rpm and that is with the IAC unplugged (which should not be unplugged during normal operation) and the idle screw in the TB completely closed. With the IAC plugged in the lowest I can get it is 1k RPM unless I completely close the idle screw and move the IAC duty cycle all the way to the lowest possible position (which can't be right) then I can actually achieve 850 rpm but clearly something is amiss if the duty cycle must be set at basically zero. Under those circumstances 850 is the lowest I can possibly achieve, can't get it lower without unplugging the IAC again and then it would be 600.

I'm using an Integra Type-R throttle body I got off ebay (I know ebay, but where else you gonna get something like this and very hard to find at that).

Hondata has some pretty clear instructions on setting idle:

---------- Hondata Instructions ----------

1. Warm up the engine.

2. Set the idle speed to 850 rpm and move the duty cycle adjustment to the middle.

3. Unplug the idle valve.

4. Adjust the bypass screw until the engine runs at 450 rpm. If you can not reduce the idle to 450 rpm, then you have an air leak. Fix this before proceeding. If the engine will not run with the idle valve disconnected then your bypass port is clogged. Fix this before proceeding.

5. Plug the idle valve back in.

6. Clear the idle valve error. You should now get a smooth idle at 850 rpm.

7. Increase the engine speed to 2500 - 3000 rpm and abruptly let off the throttle. If the engine speed dips below 850 rpm, move the duty cycle slider to the right. If the engine speed hangs above 850rpm for some time, move the duty cycle slider to left.

Some engines prefer 900 or 950 rpm idle speed.

------------------------------------------------------------

Hondata's inustructions state under no uncertain terms that I must have a vacuum leak. I'm not so sure that's the case, well at least not in the traditional sense. I have emptied cans of carb cleaner at every leak source I can think of (and a 600rpm vacuum leak would have to be a pretty big leak no?). I've sprayed all the hoses, sensors to the manifold, around all the gaskets (using thermal gaskets), etc.... I did find a small leak at the check valve to the brake booster and I've fixed that. It had an effect of about 50 rpm.

Here is what I've tested aside the above:

1) With the IAC unplugged I've put my finger over the hole where the IAC draws from in the TB with the idle screw open a little bit so that idle is 700rpm and I could feel it sucking on my finger. This achieved the same minimum 600 rpm with my finger plugging the hole.

2) With the IAC unplugged and the idle screw completely closed so that the idle is at 600 rpm I plugged up the hole with my finger and could feel no sucking. This had no affect whatsoever on the idle rpm and to me this confirms that I don't have a leak at the idle screw.

3) Now with the IAC unplugged and the idle screw completely closed so idle was at 600 rpm I put a piece of cardboard to block the throttle opening and the engine choked out and shut off rather quickly. I couldn't hear it trying to suck any air from anywhere else and it vacuumed the cardboard to the TB opening rather nicely. So this makes me think that I can't have a vacuum leak as big as 600 rpm beyond the butterfly.

4) Lastly with the IAC unplugged and the idle screw completely closed so idle was at 600 rpm I sprayed a quick shot of carb cleaner at the closed butterfly, this shut the engine off rather quickly as well.

I have a theory that whoever owned this TB previously may have messed with the throttle stop screw that sets how much the butterfly closes and I believe exists to prevent the butterfly from sticking in the bore. There's usually some yellow paint on the stop screw and nut but I only see evidence of the yellow paint in the threads of the screw, none on the nut. However it may be that there was yellow paint on the nut when I received and given the thing was filthy and cleaned and shined it up nice I may have removed that paint myself.

Word in HondaTech (and the service manual) is that this screw is set at the factory and is to never be touched. There are of course all sorts of posts on HondaTech with people fighting about adjusting and then others saying that any time you adjust it an angel loses it's wings and so on.

I'm thinking perhaps the butterfly is not closed enough when the throttle is closed and that a previous owner had changed it and this is where the 600 rpm worth of air is slipping by. I fear it could also be from when I cleaned the TB when I received it. I did hit the bore with the finest grit dremmel polishing wheel, the one that is sort of made of a scurbby pad. I think it's 400 grit. It took the oxidation layer off the bore but let's say there was zero dust produced from this and although I think on a microscopic level it took some material off and made it less round on a non-microscopic level I don't think it changed. Currently there is a uniform slight film on either side of the borders of the butterfly where it meets the bore and I figure if I had made a mistake here there would be a spot around the bore with a gap in this film.

It should be noted that with the IAC unplugged and the idle adjustment screw completely closed I can't hear any obvious air being pulled in around the butterfly. Putting my finger in there I can sort of feel some air rushing by but it's so minor it might be in my mind. Given a quick pulse of the carb cleaner in there shuts the engine off or a piece of cardboard it's gotta be pulling air in somewhere in that side of the housing.

So questions:

a) Have I missed something here I should be checking?

b) Is it a mortal sin to adjust the stop screw? And if not, what's a good procedure for this? I've read several. An interesting one I read is to calibrate the TPS with the throttle all the way open and then when you close the throttle the TPS provides a reference for how far off the set screw is. I don't think this will work as I've already calibrated my TPS spot on with the stop screw the way it is so I'd expect the same result. I'm thinking my best move is to just back the stop screw off half a turn and then see what I get, but looking for guidance.

c) Attached is a photo of the the TB. There are 3 holes in the inlet side of it. The big hole (circled in green) on the side is for the IAC and it seems also where the idle adjustment screw gets it's air from too. But I don't know what the 2 tiny holes (circled in red) are on the top of the bore just before the butterfly. I think 1 of them may go to the charcoal/evap canister, but I can't account for the other tiny hole. I tried putting my finger over these too and didn't feel any suction, though they are so tiny and one is rather close to the butterfly it's hard to be sure I completely pinched them off.

There is no FITV (or fast idle valve) on this Throttle body nor is there a provision to put one on (so no there is no block off plate for where one could be).

I'm out of ideas other than playing with the stop screw. Anyone have a brighter idea?

If you can't close the throttle more, try reducing the ignition timing to achieve the 450 RPM listed in the Hondata procedure.

That was an interesting idea. So I just tried that and the lowest I could get it down to was 500 rpm with timing set at 0 degrees. I had an option to go into negative timing in Hondata but didn't feel good about that. Fact is even if 0 degrees got it to 450 a) Smells like a bad solution to me, and b) If Hondata wants me to adjust the screw until I get to 450 rpm I have a feeling that this should not be the minimum either. I'm assuming that lower than 450 should be achievable if everything was correct.

I do believe I can still close the throttle a slight amount more by adjusting the throttle stop screw that is set at the factory and most people say should never be touched... I have no assurance that the previous owner of the throttle body did not adjust that stop screw despite not being supposed to. Some people adjust idle by moving that screw (and then others flame them on HondaTech).

I'm happy to play with that screw but I want to be sure I've exhausted all other possibilities first.

If it were a vacuum leak wouldn't one of the 4 tests I did above have had different results? Especially when I covered over the TB with cardboard and the engine quickly shut off. If there were a vacuum leak capable of keeping idle that much higher than expected when everything, including the idle adjustment screw, was completely closed off I'd like to think the engine would not have shut off and would still have enough air from the leak to keep going if perhaps only at a very low speed like 300 rpm or something.

One slight other update, I've figured out what the 2 tiny holes go to from some experimentation, they are both to the same input port for the evap system. Probably they have 2 holes instead of one for redundancy given how tiny they are.

Adjust the throttle stop screw or replace the throttle body with one that isn't worn-out and leaks.

I think David's last post is on the money.

Different engines use different methods of setting the idle speed, in the old days this was a simple mechanical stop that held the butterfly(ies) slightly open. in time, and largely because of 'running on' prevelant in early emission engines, methods were introduced for idle speed regulation that used solinoids to hold the butterfly open, or air bypass where the throttles would be fully closed, sometimes both - this was so the air supply to the engine was completely removed and so prevent auto-ignition being possible.

What I suspect is happening is that the throttle butterly is allowing air past it, and the normal air bypass is relatively ineffective in reducing the airflow into the engine to the level required for the required slower idle. Usually this is because of the closed stop being slightly mis-adjusted and the butterfly is being held very slightly open. However, you mentioned removing some material from the body bore, and this may be enough to allow the excess air through - what I would suggest is removing it and holding it up to a light in the closed position - if there is an air leak you will easily see light between the throttle blade and the body. If so, you may be able to reduce it by adjusting the closed stop. If you're lucky, it will have a screw adjuster, if not you may need to file the stop a little - id you do this I would strongly recommend you try and leave the merest clearance between the blade and the body, as if they interfere it will cause an annoying 'sticking' just as you open it, as they can slightly wedge in place if they make contact.

That said, I'm not sure I would be too concerned if the idle was slightly elevated - what exactly is it, anyway, I've gone through your post several times and you have a LOT of different numbers for it? Remember, your engine is using a mixture of aftermarket and different Honda model parts and, if it's 950-1000 rpm, it may be quite acceptable.

I assume you're using the raw data from the ECU, and not the tachometer which may be a little out at idle?

Oh, also carefully check for damage around the bypass screw, you may have damaged it allowing a through passage.

@Gord realistically you’re probably right, the idle I have isn’t exactly unacceptable which is 1k rpm. And David too, I was looking for some sort of affirmation before touching that throttle stop screw that I hadn’t missed anything and there was no other option other than accept the idle being a little off.

Interestingly I just thought to fact check Hondata with the factory service manual for a Type-R, which is more similar than different to my setup, and it turns out Hondata was a bit off on expected base idle. The book says base idle should be set to 550+-50 which is higher than all the other B-Series and means techniclaly I can acheive within base idle spec closing the base idle adjustment screw all the way which gives me 600. Not the ideal way to accomplish it but an option. The fast idle when warm and with the IACV operational is supposed to be 850+-50 and I cannot achieve that but 1k is close. When it comes to “with the IACV operational” I may see different results if I try a different IACV.

So I think it boils down to either you’re right maybe best to leave well enough alone, or I can manipulate the throttle plate stop screw. Will have to sleep on that one.

Interesting update, mystery solved. It turns out there is a light sealant applied to the throttle plate at the factory and when I cleaned it I removed that sealant. I had wondered why that some of the black scum around the perimeter of the plate was so stubborn to be removed. Now I know, doh! Learned something.

I happend to take some photos back when I was cleaning and you can see some of the sealant.

Two photos attached before and after. The sealant is the black scum on the throttle plate.

I learned this by emailing my TB boring guy George from MaxBore.com (who is fantastic). Sending it off to him to check it out and re-seal properly. The product is called Molykote 321 Dry Film Lubricant and is a teflon spray and is a rather expensive spray can.

Final update, problem solved I can now acheive any idle I’d like (within reason) without ever changing the throttle stop screw. I figure this may help someone else and since I can’t answer any of the hard tuning questions yet I’ll leave this contributuon.

I ended up fixing it myself, MaxBore suggested that would be more sensible and directed me to buy CRC 3084 Dry Moly Lube grey color because the other product is no longer available (In California at least). Same stuff different brand and less expensive. Grey color is important, you want it to be opaque and you want to be able to see where it is. I don’t think there are other colors available anyway.

The procedure is paint it around the perimiter of the throttle plate with the plate closed. First clean all the surfafes involved with acetone. Next spray the coating into a plastic cup, use a 1/8” paint brush, dip in the cup and paint it on (do not spray it directly on, use the brush). Paint just the perimiter only and only on the manidold side of the throttle plate (though you could do either, in my case this made more sense since the tiny evap holes would have been painted over on the other side), takes 15 minutes to cure. Do a second layer to build it up some (it’s very thin). After cured shine a light through the throttle body and see if you missed any spots. If you did re-apply there. Next crack open the throttle plate, this will break the seal you just made somewhat. Shine the light again, see how much light comes through now and evaluate if you want to try and make it any better. You can only get it so good and I reckon it will change some over time anyway. When you paint it you will end up with some of the stuff on the aluminum bore and on the throttle plate since you are painting that seam. I also painted the seam where the plate slips through the cross bar in there, don’t known if needed, seemd like a reasonable idea. Theoretically I think you could scrape off the excess on the bore and the plate (with something non-marring) but if you did a good job painting it’s good enough. Technically the important goop that should not be scraped off is in the tiny gap between the plate and the bore and it ran in there when it was wet and filled it in. When wet it’s very runny though it did not run through to the other side. From the factory they did not scrape any of the excess off so lesving it there gets you that factory look at least.

Attached are a photo showing how much light shined through before and after. Quite a bit shined through all around the plate before. After there were a few tiny spots but you can’t even see them in the picture. There is also a brighter picture to see what it looks like painted up.

Alternatively if you don’t want to use the coating you could send it to someone like MaxBore and they can machine a more precise throttle plate for you. It turns out the factory plates are die cut and not super precise and that’s why they have the coating. This also reduces the change of the plate sticking. I figured if the coating was good enough for Honda then it was good enough for me.