Cold Start Problem-no fast idle

Your '83 is a K-Jet without Lambda (with the Cold Start Air Valve in addition to the AAV) -- on the K/KE-Jet with Lambda systems the cold RPM isn't so high.

 

The high idle device you speak of is only on 80-83 models and is not on his 85. The cold running equipment his car has that is not working is still intact on your car. Without that system working the car prob will not idle on its own until warm and very probably is also affecting systems that are supposed to function a particular way hot or cold.

 

any ideas why the dropped the redundent bypass? doesn't seem that anyone kept it working.

 

Originally Posted by spiderseeker
...Frequency valve always buzzes...



Is the frequency valve "buzz" that noticiable? I checked the connection on mine cleaned the contacts but really didn't notice the "buzz" .
How obvious is it suppose to be?...

 

They had trouble with the cold start phase of the emission test. A higher idle will be cleaner and heat up the cat faster.

 

Depends on your hearing I guess. Put your hand on it, it should feel like it is vibrating. Also disconnect it while running, the motor should stumble and falter and maybe die.

 

Although the p-p amplitude may be about right, that isn't the type of O2 sensor output waveform that I'm familiar with for KE-Jet closed-loop operation. IME on the KE-Jet with Lambda it's a near sine wave shape (not a quick pulse). I wouldn't expect a K-Jet with Lambda to have a significantly faster mixture response time than a KE-Jet with Lambda so I wouldn't yet say "healthy" for that part of the system (but confirming/denying the control pressure behavior of the warm-up regulator vs temperature would be your next step anyway).

Unfortunately, we're in a cold snap for a few days here in CO so I wasn't planning on running my TR for a while, but when I next do I'll take some similar shots of the O2 sensor outputs on the oscilloscope. And if anyone's got a similiar scope shot of their K-Jet with Lambda running closed-loop at idle, please post (doesn't have to be an F -- any make using K-Jet with Lambda would be OK).

 

Here are some shots of the type of voltage signal coming from one of the O2 sensor on my KE-Jet US TR when at warm idle running closed-loop (the amplitude is 500 mV/div and the timebase is 200 ms/div). These transistion from lean-to-rich and rich-to-lean (which take about a second total) occur anywhere from right after one another to several seconds apart (i.e., the system doesn't do a "correction" until the O2 signal strikes one of the boundaries). I sketched in the faint areas of the traces with the dashed white line by hand just to show the slowish "S" shape better:
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Steve I too look forward to your answer

 

Steve -- Well, at least it's good to know that your WUR is quasi-OKish (if the WUR temperature for your "cold" test was in the mid-50~mid-60 deg F range). Have you been able to confirm that the system does run closed-loop at warm idle -- i.e., do you see similar lean-to-rich and rich-to-lean oscilloscope transitions on the O2 sensor output at warm idle? For example, although you've reported (privately) that the duty cycle of the electrical signal going into the FV coil is present, and does seem to be different for different modes, it doesn't gaurentee that the FV is actually affecting the fuel pressure as it should (not saying that this is the problem, just pointing out that you've quasi-verified an "input" and not the actual "output" from the injectors -- which takes a piece of gear so rare and complicated that I'm starting to doubt if it even exists ). If it does run closed-loop at warm idle, I'd take that as a good sign that the FV is OKish. I also like Brian's prior suggestion for a quick and easy way to see if the FV is working -- at warm idle, is there a bad reaction or no reaction if you unplug the FV?

One other thing you might try is doing some more in-situ testing/confirming of the AAV. For example, if you remove the air hose that goes to the AAV at the air box end -- does it really suck in extra air at cold start-up and not suck in extra air when warm? Or similarly, if you "pinch" that air line closed does it affect the cold-idle and not affect the warm-idle?

 

Try this out.. With the engine running using a pair of needle nose pliers, squeeze the main air hose which comes out of the aiflow meter too the AAV, note any change in the engine RPM. If it does not produce any change in engine RPM..you AAV is stuck CLOSED.

Check for 12V at the AAV connection. There should be constant power here with the engine running or key on engine off. If you have power and ground(very important to check ground also)..remove the AAV and try to clean it out..this sometimes works, but not all the time. On some of the valves there is a set screw on the side. Sometimes you can tweek this guy for a bit more air to enter the engine. Measure resistance of the AAV with an Ohm meter. If it comes up open circuit..you have found your problem.

You can also do an on car test. Stick the valve in the freezer for about 10 minutes. It should FULLY open. Now take it back to the car and plug it in, don't hook up the hoses. Turn the key on. Watch the valve inside. With in about 5-10 minutes it should close up gradually. If not replace the AAV.

 

stepping back to the theory for a second, I understand that the AAV creates an airway around the throttle plate. So, more air into the engine. Since the air is coming from the metering box, you get more fuel too. It has exactly the same effect as holding down the accelerator pedal a little.

The WUR, frequency valve/O2 sensor, and cold start injector all mess with the fuel mixture by either fiddling with the fuel pressure (so the injectors squirt more or less fuel), or in the case of the cold start injector, you have another injector dumping fuel into the intake plenum.

In the case of the WUR, it's enriching the mixure when it's cold, and on some cars under acceleration if the WUR has a vacuum sensor.

So how would making the mixture richer/leaner effect the idle speed?

The other puzzle piece is the throttle position microswitch (on a 328, anyway). I assume the computer is looking at it and messing with the frequency valve. While the switch unit is capable of detecting closed throttle and wide-open throttle (WOT), the WOT switch is not used...only the closed switch. If the switch is bad, what happens?

p.s. the right-angle hose exiting the AAV has a tendency to kink at the bend if the main air hose is pushed too far rearward, so check that when everything is assembled.

 

if I understand this stuff right, when the car is warm, it runs closed loop and the frequency valve is mucked with by the computer/O2 sensor output to adjust the fuel mixture to hit the emissions target.

when the car is cold, it runs open loop, the frequency valve runs at a default
50% (I think) duty cycle, and the air/fuel mixture is determined by the AAV and WUR responding to engine termperature and time, along with the base fuel mixture setting in the metering head.

when you yanked the hose to the AAV, what did you do to the open hose end and the open barb it connected to? Plug one and/or both?

The scope output looks bad to me...I'd expect a square wave something close to a 50% duty cycle (+/- 5%). What you have implies the base fuel mixture is off.

When the car is warm enough to run closed loop, the frequency valve is compensating, but in open loop you're pretty far off. 'course, that assumes that an incorrect mixure would cause the rpm's to drop...hence the question I was asking about the relationship between rich/lean and idle speed. I would guess that too lean would cause rpm's to drop, but too rich is ok assuming the plugs aren't fouling...hopefully someone knows.

if you unplug the frequency valve, the duty cycle is zero, which I think would max lean out the engine.

You can use the frequency valve waveform to adjust the mixture. Assuming the thing is running closed loop at the correct idle speed and the O2 sensor is ok, tweak the mixture to get a 50% duty cycle. At that point, you could theoretically unplug the O2 sensor and you'd see no difference in the emissions measured with an exhaust gas analyzer.

With the base mixture correct, the WUR and AAV should produce the right mix in open loop.

Did the car have to go through emissions? Maybe someone fiddled with the base mixture to get it to pass, and therefore hide a different issue.

Time to grab a scotch and wait for people to tell me how I screwed all this up

 

I apparently don't understand it either...if you pulled off the AAV hose and let extra air suck into the plenum, that would lean out the mixture since that air is not being measured by the metering plate (which would add more fuel to compensate). But you said the rpm's went up, and then when reattaching the hose the rpm's went down. I suppose more air gets in with the open hose, and if there is enough fuel being dumped in already the rpm's would increase.

Anyway, my understanding is that it's only the AAV that is responsible for fast idle (QV's and 328's), and it does it by adding more air, and indirectly more fuel, to the engine.

When an engine is cold, the fuel will tend to condense on inconvenient places like the cylinder walls, so extra fuel is dumped in to make sure there is enough hydrocarbons floating around to blow up effectively. That's what the cold start injector does, and to a lesser extent the WUR. The cold start injector is pretty much out of the picture in a few seconds, so the WUR does the remaining enrichment until the engine is warm. Then the WUR does nothing until you stomp on the gas, at which time its vacuum sensor adds more fuel.

If the cold start injector is bad, you would have a hard time starting because you wouldn't get the extra fuel squirted in for a few seconds during cranking and initial start.

The thermo-time switch could fail in two ways. If it never closes the ground side of the circuit for the cold start injector, you'd have the same hard start problem. If it is always grounding the cold start injector, you'd have lots of extra fuel dumping in all the time. You can test the second case by pulling off the plug on the cold start injector and making sure you don't get 12V measuring across the two pins after the engine is started (or when the coolant is hot and the ignition is on).

If you do have a bad thermo-time switch and someone leaned out the mixture to offset the extra fuel getting dumped in, I don't know how that would effect idle.

Another thing I don't understand is what effect the fuel ratio has on idle speed. If the air into the plenum was constant, but you mess with the amount of fuel, what happens to the rpm's?

Yet another thing I don't get is: you see an initial idle speed of around 1K, and then it drops as the AAV closes. That's what's supposed to happen...but then the idle for some reason pops back up to 1K again...does that happen when going into closed loop? If you yank out the O2 sensor output plug to force open loop (the single wire, not the two wire O2 heater plug), does the idle drop again?

Check the throttle microswitch on the side of the throttle body. When mine got cruddy, the idle tended to pop up a couple hundred rpm...why, I never got around to figuring out....it's something the computer was doing. If your switch is bad, and someone misadjusted the main idle to offset, you'd get the symptoms you are seeing. The true idle is low at 800 rpm, and it hits 1000 during fast idle and at closed loop with a bad microswitch, but is low when open loop with a closing/closed AAV.

When the system is running right, it has no problems getting past california emissions.

Between the gas analyzer, the scope, and the fuel pressure gauges, you can
test everything except the injectors themselves. If the O2 sensor is good, the gas analyzer is redundant. One of the FI books talks about using a dwell meter to set the base mixture on cars with lambda, but the scope is better.

I guess at this point I'd check the thermo-time switch, then make sure the base mixture is right, then see what happens. If you do mess around with the base mixture, I'd love to hear what the effect is on the idle as you lean/enrich the mixture. Remember that you need to remove the allen wrench and cover the hole with your finger before paying attention to the reading, and don't rev the engine with the allen wrench in the adjust screw...you can bend stuff inside the metering head.

I think that's everything I understand and misunderstand about CIS...now I'm hoping someone will fix the mistakes and make it coherent

 

Basicly the fast idle system on the 308 is a controled vacuum leak. It is controled by restriction and the AAV provides this restriction. If you have no restriction and the AAV is working correctly, the idle has no choice but to go up. Maybe it time to look at the mechanical side of things..like pinched or blocked hoses etc.

Going by what you have tested..everything in the system seems to be operating as it should. All fuel pressures appear to be ok..you have power in all the right places, and you have tested the AAV for propper function..which by all accounts has passed. All thats left is a mechanical issue really.

 

on my 328, fast idle is more like 1500-1800rpm. 2500 would be the cars that used the on/off vacuum operated valve bypass path around the throttle plate (like the '80 gtsi), rather than the progressively decreasing idle/AAV approach.