Extremely rich and rough idle

[AfricanSnowOwl]

I have an 87' GT 5spd, all stock except shorties, electric fans, o/r x-pipe, and 70mm TB. I've ran through the surging idle checklist multiple times, base idle reset multiple times, set and reset timing multiple times... So the car sounds like it has an E cam at idle. If I drive it, it bucks and hesitates at low rpm. It does this whether it's hot or cold. The idle will also sometimes climb on its own up to 1500ish and hold there for awhile. It's so rich it burns my eyes. The plugs all look identical when pulled, heavy black soot. Codes are 92, 42, 21, 11. Not sure why the code 21, the sensor is new and the voltage test across it checks out for different temps. Compression is 145 + or - 5 on all cylinders. Vacuum is around 14-16". Fuel psi 30, 40 if I unplug vacuum line. I am absolutely certain there are no vacuum leaks.

Things I have done trying to fix the idle: New TPS, IAC, MAP, ECT, ACT, O2 sensors, the 70mm TB, EGR deleted, all emissions equipment deleted and capped, replaced the injectors with known to be good 19# explorer injectors, new distributor, 3G alt upgrade, new refurbished ECU, "the fix" to the salt & pepper connections, replaced every single vacuum line whether it looked good or not, new PCV valve grommet and screen, added extra ground cable from ECU ground to block, clean and reseat HEGO ground and that strap ground cable, new plugs, wires, cap, rotor, fuel filter, fuel pump. I used Motorcraft sensors/parts where applicable.

I'm kind of lost as where to look from here. Doesn't seem to be a mechanical problem. Could it be a bad engine harness? All grounds are good. I have not pinned out and tested the O2 sensors yet, but I did test continuity from the IAC and TPS to the computer connector.

 

[MG0h3]

What are the codes for? Anything on the maf?

Did you have to make a major move with the distributor to get the timing right the first time you checked? Trying to rule out jumping a tooth; not common.


Sent from my iPhone using svtperformance.com

 

[AfricanSnowOwl]

It’s still the speed density. The codes are for running rich left and right bank, and the code 21 is ECT sensor out of range. It also has the canister purge valve code, but that’s expected since it’s currently removed and plugged.
I tried restabbing the distributor in various ways just to be sure. I also tried retarding and advancing the timing to their extremes just to see. It seems to idle the best at around 14 degrees.
Also forgot to mention I replaced the fuel press reg.

 

[MG0h3]

Oh ok must have been a late 86 build. Thought 87s had the EFI.

So I know you replaced and then tested/confirmed the ECT as good to go.

I would check the 5V ref and ground as it should be a typical 3 wire sensor.

The 5V ref and ground are often shared so maybe you’ll find a problem there and it will lead you to the problem.


Sent from my iPhone using svtperformance.com

 

[MG0h3]

And by shared, I mean most if not all sensors will share the 5V ref as well as the ground.


Sent from my iPhone using svtperformance.com

 

[AfricanSnowOwl]

So when I was testing the continuity of the TPS, I was able to verify that the 5v ref wire is good and showing 5v at all the sensors it should. I did notice that in the harness by where the EGR and TPS wires split, the 5v ref wire was degraded and the sheathing had separated, but it still had the 5v so I just put electrical tape over it.
I also verified the ground next to the battery for the ECU is good and that little connector has a good connection.
My thought was that if the harness had degraded by the tps egr split, maybe there were other areas I can’t see or get to that are the same or worse...

 

[BlksvtCobra01]

Oh ok must have been a late 86 build. Thought 87s had the EFI.

So I know you replaced and then tested/confirmed the ECT as good to go.

I would check the 5V ref and ground as it should be a typical 3 wire sensor.

The 5V ref and ground are often shared so maybe you’ll find a problem there and it will lead you to the problem.


Sent from my iPhone using svtperformance.com

87 has efi but MAF didn’t come till 89. 1988 on California Mustangs.


Sent from my iPhone using Tapatalk

 

[98 svt]

87 has efi but MAF didn’t come till 89. 1988 on California Mustangs.


Sent from my iPhone using Tapatalk

Exactly, the EFI was started in the 86s

 

[big dad]

I wouod pull the PCM and check for leaking capacitors, they usually have leaking green goo on the solder joints. If you don't know where it is, find out where it isn't.

 

[AfricanSnowOwl]

It's not the PCM. It was just replaced, but I double checked it anyways. All good there. I did measure the O2 sensor voltage at the PCM and noticed that it doesn't really cycle. It was holding around 0.9v, and would occasionally dip down to 0.6 or 0.7v real quick, but it didn't seem to be cycling at all. It did that on both the L and R sensors. Didn't change whether the engine was cold or hot.
I am absolutely certain the HEGO ground is as good as it could possibly be. The fusable link also tested fine. I did not test continuity from the sensors to the computer yet.
What would cause the sensors to not cycle and hold .9v? Is that something that would cause a crappy rich idle? Both of the sensors are less than 3 months old...

 

[Bdubbs]

Have you ever checked out the salt and pepper shakers?

Subbing.

Sent from my Moto E (4) Plus using Tapatalk

 

[AfricanSnowOwl]

Yeah the salt & pepper shakers are good. I've pulled them apart to do "the fix", as well as clean them a few times just to be sure. I've also verified the resistance for a lot of the sensors on both sides of the connector.

 

[Mustang5L5]

Code 21 always displays unless you run the test with the vehicle at temp. If you are pulling the codes with a fully warmed up vehicle AND still have a code 21, I would take a look at the ECT sensor.

Use a meter to test the sensor: Set it to resistance and check the values based on the estimated coolant temp at the time (fully warmed up = 180-192 degrees)

50 degrees F = 58.75 K ohms
68 degrees F = 37.30 K ohms
86 degrees F = 27.27 K ohms
104 degrees F = 16.15 K ohms
122 degrees F = 10.97 K ohms
140 degrees F = 7.60 K ohms
158 degrees F = 5.37 K ohms
176 degrees F = 3.84 K ohms
194 degrees F = 2.80 K ohms
212 degrees F = 2.07 K ohms
230 degrees F = 1.55 K ohms
248 degrees F = 1.18 k ohms

If the sensor is good, you can probe the ECT connector from the rear for return voltage and verify.

50 degrees F = 3.52 v
68 degrees F = 3.02 v
86 degrees F = 2.62 v
104 degrees F = 2.16 v
122 degrees F = 1.72 v
140 degrees F = 1.35 v
158 degrees F = 1.04 v
176 degrees F = .80 v
194 degrees F = .61
212 degrees F = .47 v
230 degrees F = .36 v
248 degrees F = .28 v

A bad ECT will make the car run poorly.

 

[AfricanSnowOwl]

I did test the voltage across the ECT sensor and it was reading what it should be based on various operating temps. I also tested it at the input to the computer and got the same voltage, so no issues there.
Something I just noticed today, I have a cheap-o fuel pressure gauge connected to the Schrader valve. I notice it never leaves 0 when I prime the fuel pump. When I start the car it reads a normal 30psi at idle. Turn the car off and the pressure immediately drops to 0 again. Is this an internal fuel leak?, or is it just what a cheap fuel gauge does? There is no fuel leaking from the regulator(it's a new factory style from LMR) and the injectors are refurbished explorer injectors I got off eBay a few months ago from a shop that had a really good rating. The pump is a Delco assembly I got off Amazon a few months ago. The hoses definitely show their age, but I couldn't find any external fuel leaks.

 

[Mustang5L5]

Check valve in the pump has failed. Car should still run though. I had a failed check vskve for years without any other issue.

but, key on prime should build pressure. Should spike to 39psi instantly. Cycle the key a few times. See if you get 39psi

while running oressure should be around 35-36psi and spike to 39psi when you blip the gas. If you unplug and cap the vac line, you should have 39psi. Smell the vac port on the regulator. Does it smell like gas ?

If you are only getting 30psi you should investigate that. Possibly a bad regulator and I’ve seen a split hose on the fuel pump Hanger in the tank cause similar issues

 

[AfricanSnowOwl]

No matter how many times I cycled the key, pressure stayed at 0. On the first start it never fires right up. It cranks for about 2.5-3 seconds before starting. If I turn it off and then go to start it again, it will usually fire right up. There’s no gas in the fuel pres reg vacuum line. Pressure at idle is 30, and close to 40 with the vacuum line removed.
Would a bad check valve cause an overly rich bad idle condition? I read it could be an injector leaking, but if I pull plugs they all look the same, which is dry and covered in heavy soot. Can a regulator be bad even though I can’t find a leak on it?

 

[Mustang5L5]

The failed check valve is an issue, but not a major one. Like I said before, I drove my car for about 10 years with a bad check valve in the pump before I got around to replacing the pump. It ran VERY well, but it held zero pressure with the engine off. Cycling the key would build pressure and then drop off. SO yes, it's faulty, but don't hang up over this thinking it's the cause of your issues now. Just means at some point you should consider replacing the pump.

Your reported pressures seem ok. 30psi is a bit low for idle pressure, but 40 psi with vac off is where you need to be. If you can maintain this pressure under load, then the pump is strong enough to overcome the failed check valve.

What is going on with code 21 however? If you clear all your codes, and then rerun the test with the vehicle at operating temp, do you still get that code 21? If so, I think you really need to troubleshoot that, and it might involve testing at the ECU pin (Pin 7 & 46) to see if you are returning a proper signal.

 

[AfricanSnowOwl]

I tested the voltage of the ECT at both the connector and ECU and it's the same correct voltage. The ohm test is also the same and good. I also tested the ohms at the connector for the ECT circuit and got 18K... got 17K for the ACT circuit. I think that is tolerable right? I also went back and re-verified that the 5Vref was at all the sensors it goes to. Also retested the IAC. Got zero ohms across the IAC terminals, 12v on the red wire and 0.5v on the other.

 

[Mustang5L5]

17K ohms would be roughly around 85-90 degrees or so.

But you still get a code 21 when you run the EEC codes with the car fully warmed up?

 

[AfricanSnowOwl]

I never noticed earlier, but the negative connection on my battery is just a standard cheap-o Autozone one. The other small ground wire coming off the negative connection just goes to the headlight ground. My computer ground connector just goes to the connection on the fender. I did also install a 4g ground wire from that fender connection to the block as part of the 3g alt upgrade, but neither goto the battery. Is this acceptable?, or could it be leading to some of my problems?
Also today I pulled the intake back off to test all the injectors. I put a noid light on them and they all had the same intensity and pulse. They also all had 12v at the red wire, and if I pulled the connector off they all had about 15ohms of resistance.