Running Rich Solutions?

[MarkV]

Well I got the Seville back. Here is what has been replaced:

All gaskets

Water Pump

Harmonic Balancer

O rings on injectors

fuel hose

Lifters

Camshaft

Thermostat

Spark Plugs

 

Long story short I took it to get smogged yesterday and it is running 'rich' and failed smog. The car is running great and drives well, passed all the tests except the CO %. I live in California a state which insanely believes that a 70's car should still be smogged.  Any ideas on what I can do to make it run less rich? Should I drive it more and see if it solves itself? Is there something else I need to replace? Its a 1977 with the original throttle body fuel injection. 

Edited June 10, 2018 by MarkV (see edit history)

[LINC400]

A mechanic at a gas station I worked at would lean out the fuel mixture so much in his Charger that the car would barely run. It would pass emissions, and then he would immediately change it back. He often did that right in front of the inspector. That part might not be advisable, but I did the same thing a few times with cars of my own. Just because it passes doesn't mean it is going to run better.

[misterc9]

Your best bet might be to put some highway miles on it to get everything cleaned out. Make sure you have fresh gas in it and that it gets fully warmed up. Then try again to see if there is any improvement.

[pfloro]

3 hours ago, MarkV said:

Its a 1977 with the original throttle body fuel injection. 

Hello MarkV:

 

Your Seville is beautiful...!

 

Let me first make a clarification and then provide some information.

 

The fuel injection system in your 1977 Seville is not a throttle body system but a port injection system.  There is a fuel injector at each intake port.  Most cars since the '90s have a more sophisticated type of port injection technically called

"timed sequential port injection".  In these systems, the computer pulses each injector individually in the same sequence as the cylinder firing order.  Each injector is commanded to squirt fuel into the intake port during the

intake stroke of that cylinder. This technology lowered emissions during idle and low engine speed operation.  On a GM V8 engine, throttle body injection uses two injectors mounted in the upper part of what used to be called

the carburetor. DFI (Digital Fuel Injection / as GM called it) is sort of a stepping stone between a carburetor and port injection.  The two throttle body injectors are alternately pulsed by the computer (both fire simultaneously during cranking). 

Your Seville uses bank firing to pulse the eight injectors in groups (probably 2 groups of 4).  There is no coordination or synchronization between the timing of the fuel squirt and intake stroke of a particular cylinder.  This less complicated

system worked reasonably well and met the late '70s emission standards.

 

Why all this background?  Because...

 

Your Seville uses an early analog electronic fuel injection system.  It was a joint venture developed by Bendix, Bosch and (I believe) Delco. The system has no "digital computer" and no built in diagnostics.  Specialized test

equipment was developed to specifically service this fuel injection system.  I suspect this equipment is now as rare as hen's teeth. Finding somebody who understands this technology and is able to service it is going to be

a bit tough (IMHO).

 

Your analog system was first offered by Cadillac as an option in 1975 on the 500 ('75 & '76) and 425 ('77 - '79).  The downsized 1979 Eldorado (and 1980 in CA) used the Oldsmobile 350 with this analog FI system.  

I once saw a triple white 1979 Eldorado Biarritz in an Arizona bone yard with every available option and it had your analog FI system.  It was a truly beautiful automobile.

 

Beginning in 1980, GM began offering a digital throttle body fuel injection system.  Although primitive by today's standards, it had build in diagnostics (OBD I).  It was reliable and provided good driveability & lower emissions.

GM throttle body DFI was used for many years across all divisions.  There are companies who sell systems to upgrade carb equipped cars (Affordable Fuel Injection).

 

Regarding your rich running condition:  The first thing I would do is check the fuel pressure at the fuel rail.  I suspect it should be around 40psi.  If it's too high, more fuel than necessary will be injected and the engine will run rich (high CO).

                                                                      I don't know if there is any easy way to "micky mouse" the system to get it through the smog test.  Hmmm..., I wonder if the fuel pressure can be adjusted downward to reduce the amount of fuel being

                                                                      injected into the cylinders.  For the smog test, this just might get you through it.  Your system must have a fuel pressure regulator but I have no idea where it is of even if the pressure can be adjusted.

 

Below are links which might be of help in understanding your analog FI system:

 

GM Analog Electronic Fuel Injection

Cadillac Analog Fuel Injection

 

Please keep us posted,

 

Paul

 

 

Edited June 10, 2018 by pfloro (see edit history)

[60FlatTop]

Try a piece of cardboard in front of the radiator and heat it to the limit just before you go in for the test. Experiment a little so you know when it is right on the edge of pinging on light throttle. You should pass the test, throw the cardboard out, and be good for a year.

 

[Bloo]

How much CO is it putting out and under what conditions? Idle? 2500rpm in neutral? rolling on a dyno?

 

I am unfamiliar with how California tests these days. What were your numbers (all of them), and what are the standards you have to meet with a 1977 car?

 

If the CO is too high, something is mis-adjusted or broken. The idea that a stock engine has to be way to rich to run correctly is just plain wrong. The idea that you can lean something out until it runs bad and pass is also wrong.

 

If the CO was 0.7 or less, and if California requires lower, the catalytic converter may not have been lit. The test should never be done on a cold car. Go to or three exits down the freeway and roll on in if possible. Don't shut it off unless you have to.

 

Post your numbers. Maybe we can figure out where to look.

 

 

[LINC400]

10 minutes ago, Bloo said:

If the CO is too high, something is mis-adjusted or broken. The idea that a stock engine has to be way to rich to run correctly is just plain wrong. The idea that you can lean something out until it runs bad and pass is also wrong.

 

 

 

I spent over $1000 taking a car in to try to get it to pass. Each time it was closer, but ran worse. Leaning it out until it passed, and then changing it back cost zero and got it passed.

[John348]

Check the Cosworth Vega Owners Website. They used a similar system, made by Bosch/Bendix, as well as the K-tronic system used on Volvos and VW's is similar. There is a booklet written for the Cosworth's on how to diagnose the system with a volt ohm meter.

You also might want to check those sites for forums and put it out there.

Just for laughs unplug the coolant temperature sensor while it running and warm to and see if anything changes. We all  have gotten spoiled by trouble codes and check engine lights. It does have a T.S.and controls the fuel enrichment system, through the ECU and it might even be unplugged already

 

PS I gave up on the EFI in my Cosworth and located the aftermarket weber carb set-up, now run great and never looked back.Sad to say it was just a bad system, difficult to troubleshoot without the the proper tools

Edited June 10, 2018 by John348 (see edit history)

[MarkV]

[Bloo]

Looks nice and healthy, only slightly too rich. Actually maybe quite a bit to rich, because the converter is probably covering some of it. The real story will be told when you lean it up a little. I don't remember how to do that. it has been too many years since I worked on one. I suspect any adjustments that existed were sealed. There may not be any adjustment for cruise. If that is the case it might be something like a bad coolant sensor (or a dirty connection on it), maybe a thermostat that is too cold.

 

Trying to look it up, back in a few minutes.

[John348]

The temperature sensor along with MAP control the enrichment. The temp sensor controls the duration of the fuel injector opening, as it gets warm it becomes resistive. I would look there first. should be a 2 wire lead I believe it they will be 1 black and 1 blue wire. I have to do some digging to find out which is the the line side wire. Later tonight I should have more information for you 

[padgett]

Let me preface this but saying I used to do a lot with smog engines back in the 70s. Had a Buick 455 that when they inserted the sniffer read zero. That said I'd start with  variable temperature sensors for both coolant and IAT. Next I'd hook a variable vacuum actuator to the fuel pressure regulator. In extremis I might add a capacitor to the injection nozzles to decrease the pulse width a bit.

 

All of this would require experimentation but can be done.

 

ps might also consider a fast light off dual catalyst system similar to a Mercedes M113.

[Bloo]

2 hours ago, LINC400 said:

 

I spent over $1000 taking a car in to try to get it to pass. Each time it was closer, but ran worse. Leaning it out until it passed, and then changing it back cost zero and got it passed.

 

I am not doubting you. I am glad that worked out and sorry you had to spend $1000 trying without actually getting it fixed.

 

For what it is worth 1.2 CO and 220 HC were the limits up here in WA back in the 80s/90s, higher limits in four steps back to 1974, and no testing for cars older than that. 99 percent of everything that passed through my bay (including things much older than smog regulations) spit about 0.7 or less CO and not much over 210 HC once tuned perfectly. These numbers are without catalyst, a car with a catalytic converter will be lower as soon as the catalyst lights off.

 

I am talking exclusively about stock vehicles. Cars with lumpy cams are a whole different game, at least at idle.

 

Yes, I am saying a whole lot of pre-catalyst cars could pass 1.2 CO and 220 HC easily, and usually did when well tuned, including one of my daily drivers back then, A 1966 plymouth fury with 375k and a real appetite for oil. It was too old to even require testing here in WA.

 

Engines that could not quite do that I tended to remember, like the earliest ford 260/289, the early OHV rambler six (based on the flathead block), cars that were converted to PCV by drilling a hole in the manifold, etc.

 

 

[Bloo]

42 minutes ago, John348 said:

The temperature sensor along with MAP control the enrichment. The temp sensor controls the duration of the fuel injector opening, as it gets warm it becomes resistive. I would look there first. should be a 2 wire lead I believe it they will be 1 black and 1 blue wire. I have to do some digging to find out which is the the line side wire. Later tonight I should have more information for you 

 

This is the way. I have been looking online and have not found much documention of the system, yet. My notes from long ago are inaccessible.

 

John348: Do you mean that the resistance of the coolant sensor goes up if the car gets hotter?

 

If hotter equals high resistance, dirty connections should have the opposite effect, making things too lean.

 

MarkV: I notice the thermostat was changed. I am fairly sure it should be 195*. Can you find out if maybe a 160* or 180* got slipped in?

[Bloo]

Verify that the MAP hose fits tightly at both ends. If it is questionable at all replace it.

 

Also pull the little vacuum hose off of the fuel regulator. There should be no raw liquid gas in there. If there is, the regulator is bad. It will probably be fine, but should be verified. This hose also has to fit tight with no possibility of leaks. If there is any doubt, replace the hose.

 

Also watch for either of those hoses sucking shut (flattening out by themselves) at idle. You can't have that.

 

 

[John348]

49 minutes ago, Bloo said:

John348: Do you mean that the resistance of the coolant sensor goes up if the car gets hotter?

 

 Yes, places resistance into the circuit as the temperature increases, goes back to the ECU and then controls the duration time of the injectors.  it might be something as simple as to low of a thermostat as you mentioned

 

32 minutes ago, Bloo said:

Verify that the MAP hose fits tightly at both ends. If it is questionable at all replace it.

 

I totally agree but finding one might be problem in itself.

 

 

Quote

 

If hotter equals high resistance, dirty connections should have the opposite effect, making things too lean.

 

As I recall when playing with the Bendix Injection set up in the Cosworth the wiring connections were not really weathertite it struck me as being odd, lead me to believe that a large part of the problems were due to extra resistance at the connections  of the Benix harness misleading the ECU with reading resistance that was not due to sensor input 

Edited June 11, 2018 by John348 (see edit history)

[John348]

I am sure Bloo would agree with me, what I would suggest for you to do first is with the engine cold disconnect the coolant temperature sensor, and take an ohm reading between the two connector pins on the sensor. Reconnect the connector and start the car and get the engine up to operating temperature. Shut the car off and disconnect the  plug and take an ohm reading across the two pins of the sensor. The reading should indicate an increase in resistance, if not there is the problem. I have to look for what the exact amount of resistance should be

I have to find the test manual for the bendix systems, I think it is in the other garage. I will look tomorrow  

[MarkV]

Is this supposed to be like this? It looks like it is missing a hose Right at the left rear of the throttle body

[MarkV]

Also of note that  I just noticed was the fuel economy meter is now always orange and never turns green. Maybe thats linked? 

[Bloo]

2 hours ago, John348 said:

3 hours ago, Bloo said:

Verify that the MAP hose fits tightly at both ends. If it is questionable at all replace it.

 

I totally agree but finding one might be problem in itself.

 

Is that some kind of special hose with a tight bend or different size ends or something?

 

2 hours ago, John348 said:

As I recall when playing with the Bendix Injection set up in the Cosworth the wiring connections were not really weathertite it struck me as being odd, lead me to believe that a large part of the problems were due to extra resistance at the connections  of the Benix harness misleading the ECU with reading resistance that was not due to sensor input  

 

True enough, IIRC nothing had weatherpac yet.

 

2 hours ago, John348 said:

I have to find the test manual for the bendix systems, I think it is in the other garage. I will look tomorrow

 

 

That would be an terrific you can find it.

[Bloo]

1 hour ago, MarkV said:

Is this supposed to be like this? It looks like it is missing a hose Right at the left rear of the throttle body 

 

I don't know. You really need a vacuum diagram to figure it out. If you are lucky, there might be one on the smog sticker under the hood. If not, that is going to require a manual, either the factory one, or a Mitchell Vacuum Diagrams manual that covers 1977 domestic cars.

 

It is possible, even likely, that is is supposed to be like that, and there was once a foam filter over the fitting.

[Bloo]

2 hours ago, John348 said:

what I would suggest for you to do first is with the engine cold disconnect the coolant temperature sensor, and take an ohm reading between the two connector pins on the sensor. Reconnect the connector and start the car and get the engine up to operating temperature. Shut the car off and disconnect the  plug and take an ohm reading across the two pins of the sensor. The reading should indicate an increase in resistance, if not there is the problem.

 

Yes!

[Bloo]

1 hour ago, MarkV said:

Also of note that  I just noticed was the fuel economy meter is now always orange and never turns green. Maybe thats linked? 

 

Is that a vacuum gauge?

[MarkV]

Would a disconnected a/c compressor with lines cause a vacuum leak causing the computer to make up and causing a rich condition? 

[Bloo]

If you mean the refrigerant lines from the compressor to the evaporator and condenser, that makes no difference.

 

If any vacuum control hoses for the ac are open those would need to be plugged, and plugged well. No vacuum leaks are acceptable. It really is amazing how small of a leak it takes to screw the driveability (and the emissions) all up.

[John348]

44 minutes ago, MarkV said:

Would a disconnected a/c compressor with lines cause a vacuum leak causing the computer to make up and causing a rich condition? 

 

Good guess but no.

I strongly suggest that you try testing the coolant temperature sensor as both Bloo and I suggested. It could be one or a combination of several different things but  we need to start ruling things out

[sfair]

Take note of post #19.

[Rusty_OToole]

Since I don't have magic eye balls that can inspect a car from 1000 miles away I have to guess some things. First guess is, that your mechanic went over the car and verified that everything is to factory specs. Air filter not plugged up, spark plugs good, ignition in good shape, fresh oil change, fuel injection working properly. Your pic of that gadget with the mission hose makes me suspect that something is not working right. I don't know what that thing is or what it does but it is kind of a tip off that something is not right, when you notice parts missing.

 

Now first you or your mechanic has to go over the engine and make sure it is up to spec. I would inspect and replace spark plugs and air filter as a matter of routine. If the plug wires are original I would test them too or just replace them on general principles.

 

With the engine checked over and tuned to the teeth take it on a few long drives. Fill the tank with hi test and add a can of fuel injector cleaner. You may have some old gas in there that is not what it used to be. Run the tank nearly dry then fill up again and add a can of injector cleaner.

 

Run the second tank down, fill it half full, pour in a quart of methyl hydrate (alcohol, you get it at the hardware store). Then go get the emissions test done, make sure the engine is good and hot when you go in, I mean drive around for half an hour so it is fully warmed up.

 

It should pass, if it doesn't there are additives you can use.

[MarkV]

New plugs and wires as a part of the rebuild and air filter. I will run through a tank of gas with additive. Which is the best to use?

[John348]

17 hours ago, MarkV said:

Also of note that  I just noticed was the fuel economy meter is now always orange and never turns green. Maybe thats linked? 

 

 confirming the emission test results, the engine is running rich!  While it could be a clogged injector not misting the fuel properly, but that would be give very noticeable engine conditions. 

I am sure there are additives that might work, I myself have never found one.

Edited June 11, 2018 by John348 (see edit history)

[pfloro]

As I continue to ponder your rich running situation, two things come to mind:

 

1) As part of the 1977 emission controls package, your engine has the A.I.R. system.  The Air Injection Reaction system pumps

     oxygen (~20% of our air) into the exhaust ports.  This helps the combustion process continue after the hot exhaust gases

     have left the cylinders.  I believe it helps reduce HC more than CO or NOx but I don't know.  Making sure this system is working is simple.

 

   Rising up from each exhaust port on both manifolds is a steel manifold.  There is check valve at the top of each manifold.

   The check valve (one way air flow) protects the air pump in case there is a backfire.  Simply disconnect the hose from one

   of the check valves and verify that air is flowing at idle and part throttle. The only time air should not flow is during a high

   vacuum condition (deceleration).  To prevent this flow, there is a vacuum actuated diverter valve which reroutes the pump's air into

   the air cleaner housing.  Extra oxygen during deceleration (when the mixture tends to be richer) can cause backfiring because a richer mixture

   can explode in the very hot exhaust stream.    See the attached image of YOUR passenger side check valve...

 

2) How do you know that at some time in the past, if leaded gasoline was put into the fuel tank.  Granted, the restrictor in the gas tank

     filler neck would have prevented a leaded gas pump nozzle from being inserted but was the car at some point fueled from a jerry can?

     I don't think it took very long to coat the catalyst honeycomb with lead effectively neutering it.  This might be a long shot but... 

 

I read that with the analog FI system, the engine ran so clean that a catalyst wasn't needed to meet 1977 emission standards.  However,

GM chose to install one anyway...

 

Perhaps others can share their thoughts on these two points...

 

You'll find the problem.

 

Paul

 

 

[mike6024]

You probably also have high HC, hydrocarbons, but that gets reduced by the AIR injection as said above, so by the time it gets out the tailpipe the HC are at a passing level. So the AIR system is doing more to reduce the HC's than it really should have to do.

 

Also if you do have a catalytic converter, install a fresh one. That would help, but ideally you need to get the mixture optimized before putting on a new cat. A cat run on a too-rich engine will go bad much too quickly.

[Bloo]

IMHO fix the car.

 

HC     CO      O2      CO2

50      1.6     0.0      14.1

39      2.0     0.0      14.0

 

WIth HC that low, and 14% CO2, the catalyst is working. Perhaps not as efficiently as it could. The readings ahead of the cat will be higher, and the CO you are looking at AFTER the cat is too high. That is probably 5 or 6 percent (or more) going INTO the cat. That would be a nice full throttle mixture. It is way too rich for idle and cruise in anything remotely modern with reasonable fuel distribution.

 

Kludges like dumping alcohol in the tank never end well, because the original problem still exists. There is no halfway with cars like this. Fix it once and fix it right. It is SO CLOSE.

 

so, IMHO:

 

1) Check the temp sensor the way John348 described.

 

2) Verify it has a 195 degree thermostat in it. Since the work was just done, it should not be too hard to figure out what the part number was. It might even be on the invoice.

 

3) Verify that the vacuum hose to the MAP is not leaky. It needs to fit TIGHT, and not have cracks (look right at the end by the hole for radial cracks) no tape repairs, etc. If it comes off and goes on real easy it is probably too loose.

 

4) The indicator you mentioned that never goes green, Is that like a vacuum gauge? If so, that is indeed a clue. Low vacuum enriches the mixture. It probably does it in a step, and so may not be the problem. If it really should be green normally, then it needs to be sorted. Perhaps a leaky vacuum hose?

 

5) Ignition timing, especially retarded ignition timing, could cause low vacuum. Cars of that era have specific procedures. It can be really easy to get it wrong. The smog sticker might tell how. If not we are going to need the manual.

 

6) Verify that the vacuum hoses are routed correctly, and that there are no loose ones. The amount of vacuum hose I went through back when I was working on cars like this one this is staggering. The cars were only about 10 or 20 years old then. The chances of all of yours being good are about nil.

 

Good luck, you are almost there.

 

Edited June 11, 2018 by Bloo (see edit history)

[Bloo]

43 minutes ago, mike6024 said:

 

Also if you do have a catalytic converter, install a fresh one. That would help, but ideally you need to get the mixture optimized before putting on a new cat. A cat run on a too-rich engine will go bad much too quickly.

 

Don't do it yet though. The rich mixture combined with the AIR might melt it. Also, I am relatively sure from the numbers that your cat is doing something. It isn't going to take much of a cat to get under those limits once the engine is burning right.

 

The real advantage (since the car is low mileage), is that you probably have one of the old GM cats with the pellets. A modern honeycomb style cat would free up the exhaust and probably give you a little more power.

Edited June 11, 2018 by Bloo (see edit history)

[pfloro]

As I mentioned in post #4, consider measuring the fuel pressure at the fuel rail. Won't too high a pressure cause more fuel to be injected during the pulse? I don't know.

[MarkV]

The thermostat is #13008 which is 180 degrees and wrong. Obviously we need a 195degree one. Would that send these readings that far off?

[John348]

14 minutes ago, MarkV said:

The thermostat is #13008 which is 180 degrees and wrong. Obviously we need a 195degree one. Would that send these readings that far off?

Yes it will.

[MarkV]

I am taking it back over there later this week. Should I have them replace the coolant sensor and egr as preventative?

[MarkV]

The green light is vacuum controlled as the car gets hotter it is more prevalent, so the vacuum must be related to operating temperature right?

[John348]

2 hours ago, MarkV said:

I am taking it back over there later this week. Should I have them replace the coolant sensor and egr as preventative?

 

Myself personally I would replace one component at a time, so I am 100% sure what the problem was. Even if you were to test the CTS with the wrong temperature thermostat it might test as being faulty when it is not.  

 

1 hour ago, MarkV said:

The green light is vacuum controlled as the car gets hotter it is more prevalent, so the vacuum must be related to operating temperature right?

 

I am not very familiar with that specific system, but it is very possible that the vacuum is controlled by a temperature controlled ported vacuum switch. Again with an incorrect temperature thermostat it would be affected.