Am trying to diagnose a dancing idle between 1200 and 1500rpm on my '91 JDM FC and having a hell of a time. Bought a vacuum/boost leak tester from Boostpro in an attempt to help me. Got an adapter and fitted it up to the turbo. Pressurized it to 0.5 bar and it lost pressure within a second or 2. So I went further downstream and plugged it in at the throttle body. Did the same thing. But I can't hear any huge volume of air being pushed out, which is the strange thing. Sprayed soapy water all over and could see no bubbles.

I even tried a suicide attempt today in the form of putting a smoke bomb in my throttle body and seeing if I could see where the smoke was coming out from, but I couldn't spot any.

The vacuum test should just be plug in, pressurize and check for leaks, right?

 

I have no experience with a vacuum tester, but you may want to spray some stating fluid around and when you heard the engine bog you found your leak.

 

I've been spraying carb cleaner like a **** all over when the idle's going up and down and it just kept on doing it.

 

its been my question for a while.
i know its probably covered a million times in other threads, but some of us dont have 2 hours or whatever to search and read a thousand threads that mention the word but dont talk about it in any detail.
are people afraid the boogeyman will get them if the write a sentence or two response?

 

Maybe its not a vacuum leak.

Check:
1) trouble codes
2) timing and advance
3) idle mixture
4) TPS high & low resistance values, dropouts, and adjust for 1k ohm at idle throttle position
5) throttle roller is off the fast idle cam
6) thermowax/fast idle system operation
7) BAC valve resistance, operational check

Does it do the same thing if you have a jumper in the initial set connector (that fixes the BAC to a constant duty cycle)?

 

1. Might try to make one (even the later JDM FCs don't have a check engine light)
2. How can I check that when the idle is going up and down above 1,000rpm (which will mean the timing is changing)
3. Don't think later JDM engine can change that
4. Tried setting to both 1k ohms and 1vdc. Both have same effect on idle
5. No cold start system
6. No cold start system
7. No BAC

Only way I can make it stop is by disconnecting the TPS sensor, which then gives me a misfiring idle at around 1,300rpm.

 

Like the idea of the smoke bomb but wouldn't it be more prudent to perform the smoke test first followed up immediately by pressurizing the system and then check for leaks as the smoke should should be forced out the engine following the path of least resistance.

 

That's a good call, and something I might try later on today.

 

Tried pressurizing the system after smoke bombing and all it did was coat my intake manifold and throttle body with the dye from the smoke bomb, so no luck there.

Also made an LED tester for the ECU. Turned on the ignition and no lights came on. Don't think I can mess up something as simple as connecting an LED to 2 connectors and grounding out another plug, but stranger things have happened!

Only thing that makes me think it might not be a vacuum leak is what happens when I disconnect the TPS. When I do this the idle stays a little high (1,300rpm) and starts misfiring a little. Strange thing is that when I start the engine up initially it runs with a high idle (1,500rpm) for about 20-30 seconds and then it's like a solenoid flicks and it starts going up and down. Nothing to do with the cold start/fast idle cam or BAC as they've all been removed.

 

Why not just start fresh with all new vac lines? I never had luck pressurizing mine either. You can also try to use a stethoscope with the round end taken off. That will allow you to hear any sucking a vacuum leak would make. I vote you just replace all your vac lines that way you can eliminate that as a problem. I just read that you removed your BAC, why? It doesn't hurt performance and makes adjusting idle much easier.

 

They're all new. Have rerouted a few as I've taken out things like PRC solenoid, carbon cannister etc.

 

Plugs the exhaust. Then put your intake plug in and pressurize the system to 15 psi. Let the pressure build. You'll probably have trouble hearing a leak with the compressor running.

So get some kids bubble juice and put it in a spray bottle. Starting at the intake plug where you're pumping air into the system, start spraying the intake tract. Keep moving slowly and methodically down the intake track being sure to spray in all the hidden areas, the backsides of this hoses and what-not.

When you see the bubbles growing, you've found your leak. Finding leaks on the 13B is a pain in the *** because of the way the LIM and UIM are arranged. YOu may find it necessary to remove the TMIC and replace it with some PVC pipe so you can have access to the injectors and rats nest.

 

Made an ECU code checker this evening and the good news is that no ECU codes are being thrown out. However that does mean that I'm no closer to working out what is causing this erratic idle.

 

do you have the ACV still?

it turns out you can't pressure test a stock t2, because there are a couple of things that work from the PRESSURE from the air pump.

and with the engine off, there are direct connections from the intake manifold to the ACV/atmosphere

so to pressure test a t2, you have to bypass half of the stuff plugged into the intake, which kind of defeats the whole point..

bouncing idle, have you checked/set the TPS yet?

 

Does ur rpm still bouncing after the car fully warmed up (or say "after 10mins of driving").
Mine does this too when the motor is cold, but after a few mins of driving (when water temp around 180*F) then it stop.

 

How long ago was it since the idle was below say 800 rpm and what have you done to the engine since then? Misrouted a *vacuum* line or two?

What happens if you use your hand to try to close the throttle linkage more? RPM's go down?

 

It's not run at all for over a year (I've had it about 3 months and have got it from not firing up at all to here with a fluctuating idle). I've never seen a stable idle around 800rpm.

Could be a misrouted vacuum line - will try to make a vacuum line map for you to show where everything's going.

Throttle plates are closed as much as they can I think. Again, will try that when I get home, but I don't think they will close any more.

 

I have never tried to pressure test a T2 with an ACV installed. It may indeed present some problems. But I don't really have your car in front of me so I can't say. What I will say is this though. When you are pressure testing you are not actually looking to see if it holds pressure, as strange as that sounds. Don't focus too much on what the gauge says, because air will leave out of the exhaust ports. All you're doing is trying to find if air is escaping from a place it should not be escaping.

So let's say I have a bad UIM gasket causing a vacuum (and boost) leak. I pressurize my engine. I feel air escaping around the mating surface of the UIM and LIM. I spray a little bit of slightly soapy water on the area and I see bubbles. That confirms it. I replace the UIM gasket and now all of the real "leaks" are fixed. I pressurize my engine again and yes, the gauge doesn't fully hold pressure because air is escaping out of the exhaust. But I still found the leak so I'm good to go.

 

Thanks for that. In that case it might not be a vacuum leak at all. But I'll be damned if I can work out what it is.

Might try and get the engine fully warmed up tonight (on private roads due to me not having any plates at the moment ) and see what happens to the idle then.