TPS vs Throttle Stop?

Just bought a 1996 Geo Metro 1.0 with 180K, had a couple burned valves, so I did a valve job. It will start fine when cold, but dies as soon as throttle is opened even slightly, and will not start after that. If allowed to idle for 3 or 4 minutes, the problem vanishes like a fart in a high wind.

I suspect the previous owner may have adjusted the forbidden Throttle Stop Screw, as it does not have either a cap or paint on it. Book says TPS should have a range of 57 mV at throttle close to 497mV at wide open. To set it the book says to place a 3.5mm guage between the throttle stop and the throttle linkage, and adjust TPS to 97mV in that position. Done that, with above results. the only odd part is that I only get about 400mV at wide open, but the operation is as above. Replaced plugs, wires, cap, rotor; checked timing, EGR, MAP, MAT,etc and all check out okay.

ECM does not set a code.

I am back to suspecting that the Throttle Stop (which is the baseline for the TPS reference) is off.

Anyone got a hint how to get it back where it should be? Like a factory setting for throttle plate to body opening?

Proper magic spell?

Well, here’s what I’m thinking about this… if you’re getting the initial voltage correct but only getting 400 millivolt at WOT, the problem ought to be with the top end, right?

Since you obviously have a digital volt meter, probe the TPS and slowly operate the throttle arm. Does the voltage change smoothly? If there are no points on the movement of the TPS that have any holes (that is, drop to 0v) or shorts (full reference voltage), then I don’t think the problem lies there.

Riddle me this: have you checked for vacuum leaks? Fuel pressure?

TPS voltage is smooth throughout range of plate motion.
I only mentioned the upper end voltage because it calls into question whether or not the lower end voltage should be located closer to the closed position- ie: is the resistance too high to start with, thus giving the correct voltage in the incorrect position. (Obvious thing to do here would be to replace the TPS and see if the voltages are different, but I’m loathe to start throwing parts at a problem until I’ve exhausted the diagnostic possibilities)
Fuel pressure is good, with no variance, and filter is new.
I’ve been unable to find a vacuum leak, but it’s probably worthwhile to put a guage on it. I wish I had a real shop manual, with things like vacuum ranges at certain ponts such a PCV, MAT, Vapor Canister, etc. Also, I can’t understand why the ECM doesn’t think something is wrong. I believe the downstream O2 sensor may be original, but it seems to produce a signal when hot, and there is no restriction in the exhaust or Cat Before the valve job, it was setting a code for catalytic emissions, but doesn’t now.
It seems to be related to engine temperature somehow, as if there is something expanding when it warms up to meet specs, but failing when cold.
I just can’t FIND it. Need a magic bullet instead of a shotgun.
(Scratches head and rolls eyes)

Hmm, wierd one. The particularly wierd part is that cracking the throttle plate only slightly would affect re-starting the engine. My first thought is a flooding issue. Letting it run 4 minutes allows the raw gas to dissipate enough that you can open the throttle…

It will start fine when cold, but dies as soon as throttle is opened even slightly, and will not start after that.

Can you expand on that? How long does it remain unable to start?
Are the plugs soaked?
Got spark?

If allowed to idle for 3 or 4 minutes, the problem vanishes

So if you allow it to idle for 4 minutes, the problem goes away. Once warmed up, if you turn it off and immediately try restart it, does the problem repeat or is it gone as long as the engine is warmed up?

Does the idle remain in spec under cold and hot conditions?

When it won’t start, does putting the pedal to the floor while cranking have any effect?

I’m pretty sure that the symptom is, in fact, flooding. (The cause is the tricky part)
After a start failure, it will usually have to sit for ten minutes or so before trying again with any success, again pinting to flooding.

Spark is consistant as evidenced by looking at a timing light while cranking.
I doubt that the plugs fouling is the problem, as they should remain fouled even after ten minutes in the closed combustion chamber, and fail to fire on the restart.

If the pedal is almost fully depressed after dying, it will start to fire after about 10 seconds or so of cranking, (clearing out some of the fuel, possibly) but starting in that setup is more of a labored chugging at 500 RPM or so, and as soon as the pedal is released it will die. Idling is out of the question.

Once warmed to operating temperature, subsequent starts are normal in every way.

Idle remains a little on the high side at the present TPS/Throttle Stop combination, being about 1000 to 1100 RPM. At other combinations, I have been able to get a fairly smooth idle at 850.

I’m just about ready to start throwing parts in, like another cap/rotor, plugs, TPS, O2 Sensor and a new air freshener, just to see if anything changes…

Please forgive me if this is irrelevant. Is this a carburated car with a choke? It sounds so much like an improper choke control I have to ask,
4 minutes warmup time is enough to open the choke as appropriate.
Less than that the choke is totally closed and flooding the engine
When you push the gas pedal to the floor in the olden days it would force the choke plate open, and the lean mixture would work out the extra gas and start.
If it is a temperature controlled spring that operates the choke that may need replacement or adjustment.

throttle body fuel injection with computer controlled mixture and timing.

I think you’re barking up the wrong tree with the TPS. The .097 volt difference is not significant enough to cause problems, and even when the TPS is completely out it doesn’t usually cause serious non-running conditions.

Have you checked the coolant temp sensor?

You won’t get 5 volts out, unless you have 5 volts in. I think the repair manual has a procedure to follow for a disturbed throttle stop screw. The manual probably has the throttle plate to throttle bore spec, also.

What is the difference between a cold engine and a hot engine? Temperature, or, temperature indication. Right? Temperature indications come from the engine coolant temperature sensor, and the air charge sensor. Compare each of their output to actual temperatures.

Ahhh…for a REAL shop manual…
Haynes just says to test the water temp sensor by putting it in water and checking that the resistance increases with rising temperature as you heat the water- no specs, either min or max, and no temperature profile. (sigh)
All Haynes has to say about the throttle stop is don’t disturb it.
Chilton’s is no better.
Is All Data any good for details?
Local library doesn’t have much that’s helpful, either.
I’ll probably just swap Coolant sensors with my Swift (same engine) and see if it behaves differently before buying a $40 part…

Your public library may have an online subscription to which you can access when you are at the library. Ask a librarian. If not, you can get your own online subscription for $25 a year. That’s cheaper than one wrong part changed.
Check for a throttle switch. Some engines had them. Yours? Maybe.
A poor connection, in a wire connector, can cause high resistance in that circuit. If a temp sensor had high circuit resistance, it would read much colder than reality. This article tells more about temp sensors:

You could also, assuming they’re the same part, check the resistance on the swift’s CT sensor and then compare it to the Metro’s. Who needs a stinkin’ manual when you can do independent research!

Consider yourself lucky it’s TBI, you can look at the spray pattern for additional information. When it’s running, how does the spray pattern look, nice cone shape or is it dribbling? Have you checked the fuel pressure? On my old TBIs, I had to rig up my own gauge and plumb it into the hardlines. I cut an end off a fuel line in the junkyard and rigged in a 20lb gauge in-line with some hose/clamp that fit over the hardline in the car.

I’m not positive about the Metro, but many GMs have two temp sensors. One is a variable resistance sender used to drive the gauge, the other is a switch that is used by the ECM to tell if the engine is cold or hot. In that case, you need to check the switch, not the sender. In my experience, these usually don’t result in cold start flooding. If it fails looking like a hot engine when it’s really cold, the result is a lean mixture at cold start. If it fails looking like a cold engine and the engine is cold, it starts and runs OK until it actually warms up. YMMV.

Stupid Question Number 1: If you really think the problem is the Throttle Stop Screw, have you tried simply adjusting the thing and seeing if the problem goes away? You certainly sound like a careful enough workman to put it back where it started if it doesn’t help things.

Stupid Question Number 2: Have you tried making a list of things that are supposed to happen during or after warmup? One of them may be kicking in early or late. Examples: Your engine is probably running open loop with no feedback from the O2 sensors during the time you are having trouble. Maybe the ECU fixes whatever is wrong when it starts running closed loop. And how about the EVAP stuff? My impression is that the cannister should not be purging until the engine is warmed up. Could it be scrambling the the mixture for you by purging early?

I’m not remotely an engine guy, but some of the folks here should be able to provide you with a pretty good list of stuff that starts working differently after a few minutes of warm up.

Progress of sorts: the Coolant Temp Sensor tested open- no resistance.
The amusing part was that when I removed it to replace it (without draining the system) NO coolant came out of the fitting in the thermostat housing. So I started the engine- still no coolant leaking out. Radiator and reservoir are full.
This is somewhat counterintuitive- shouldn’t the sensor be IN the coolant?
Looks like it’s time to remove the housing and look for condoms, dead mice or duct tape in the thermostat…

I am just going to respond to the AllData question. When I first subscribed for a Ford product, it looked just like a Ford DVD that I later obtained. The advantage was up to date TSBs.

Since they were bought by an auto parts store, there seems to be some changes. A couple of weeks ago I wanted access to the connectors on a module behind the glove box. AllData instructed me to remove screws in the bracket by reaching up from below. I could not even see the screws. I dropped the glove box beyond its stops for a better look. At that point I just slipped the module out of its bracket without removing any screws. It was a lot easier than their method.

There is a lot of info about trouble-shooting sensors. I think that is accurate and all the wiring diagrams and connector diagrams are accurate. I second the library proposal. If you have a library card and access from home, you might even be able to enter their database from home. Don’t forget everyone, they still do have full-text TSBs. With a newer car, that alone is worth the price.

Coolant Temp sensor replaced and working per specs, MAT sensor working properly. No change in weirdness- still dies after cold start if throttle is opened.
And still no trouble code set with either of two identical ECM units; driveability is fine, and I’m getting 36 MPG @ 70 MPH.
Still completely unsuccessful in obtaining info on the throttle stop screw- AllData and a couple of the Pay-Per-Answer sites have nothing.
Each adjustment I do seems to result in very little change, and no relief from the major symptom.
Anyone know of a good place to get a TSB for a 15 year old car?
Sure wish I could figure it out…

Still in crummo land.
The latest interesting development is the Check Engine light flashing on and off when I have botched a start and am cranking with throttle wide open to clear the flood. I can find no reference to a flashing light, other than the flashes at ignition turn on that indicate it is working. The only code it sets is the “cat out of range” I get after about 200 miles of driving. And the thing is using a quart of oil every 150 miles, with no visible leaks or visible smoke from the exhaust.
A little theory might help…It appears the major problem is instant flooding when the throttle is opened while in open loop. Could it be because the throttle plate is closed too far, and fuel is pooling on top which then gets dumped suddenly when the plate is opened?
What else happens when the throttle is opened, besides goosing the injector and adjusting the mixture and timing?
Do the O2 sensors even have input in open loop?
Anyone know who makes a real technical manual for Metros?

Have you done a compression and leak down test since you did the valve job?
Something is causing your oil usage, and that’s the best way to determine where the issue lies.


Great Googlie-Mooglie!
Comp test with crappy loaner from parts store gave me: #1: 120 #2: 85 #3: 75.
Oil in the cylinders made little difference. If I am to believe that guage, I am severely hosed here!
Is my new head dorked? Could the timing belt be off a notch?
No wonder I can’t get it to run right…