EFI Tuning the 5.0 Windsor

[skullandbones]

I better go back and explain the components better. I have a Summit after market EFI harness that has gotten really good ratings (SUM-890120). I also have the AAW body harness so all wiring is new except the trans harness which is OEM but in excellent condition-retaped and continuity on all circuits). I also did a lot of research on the instruction data and liked many of the features including an inertia switch connection. The MAF I got to match to the 24 lb injectors also Summit is this: SUM-29051B-C. The C is for cold air system. The cold air was planned from the beginning to get cooler passenger side fender air instead of engine bay air.

So it is the cold air version. I'm using a Summit 75 mm throttle body which is port matched to the Holley Systemax folded plenum chamber intake. I have stock fuel rails, regulator, coil, sensors, fuel pump, tank, T-5 trans, EEC-IV PCM all coming from a 92GT. Recently, I found that I had installed the black TFI module instead of the correct gray one but now that has been fixed. My basic timing (without Spout) is 12.5 to 13 degrees which I was finally able to verify (had difficulty reading the timing marks on the balancer but finally got the right combo of paint for color contrast).

I planned to check fuel rail pressure under load but have not been able to get gauge to work this time. I think someone abused it the last time it was used. Taking it back for replacement. Will start again tomorrow. I've been double checking for vacuum leaks, distributor wire arrangement making sure wires are spaced and not touching, firing order, etc. Will probably have some representative numbers by Monday evening. Thanks,

WEK.

[skullandbones]

I've had a problem getting a working fuel pressure tester. I finally was able to take parts from one to get a working unit. So here are the results of my fuel rail pressure test for the old stock 92GT fuel pump. All in psi.

At idle - 40 after a few cycles, 1st gear at 4K - 38, 2nd gear at 4K - 38, 3rd gear at 4K - 38.5, 4th gear at 3K - 40, and 5th gear at 2.5K lugging up hill - 41. The fluctuations were from 37 to 41 but steady when at a certain rpm. I would have expected a lot more variation for it to be a fuel pressure problem. So it is a little disappointing but expected. Tomorrow I am changing out the pump for a new one and redo the test. Then we will go on to the next things on the "list". Thanks to my wife, Trish, for riding and doing the readouts for me. This is actually the fourth time she has been riding along with me in the roadster. Tomorrow will be the fifth. I will have to think of excuses like this to get her in the passenger seat more often.

It wasn't pretty but it worked.

WEK.

[Jeff Kleiner]

Hey, that high tech gauge attachment method looks very familiar! I think I used checkered flag duct tape though; you know to make it more racey! Yeah, it doesn't seem to be a fuel pressure problem. I got a load of bad fuel a few years ago that partially plugged the pickup sock and didn't feel the starvation and hesitation until pressure dropped below 15-18 PSI. I know you plan to put the pump in as preparation for future upgrades anyway.

...I will have to think of excuses like this to get her in the passenger seat more often.

"Want to go get ice cream?" always works for mine

Cheers,
Jeff

[Carlos C]

Very high-tech stuff, indeed...

Carlos

[skullandbones]

Changed out the fuel pump today. I replaced the original one for a 255 l/hr walbro. The fuel pump I took out was also a walbro. I guess that could be stock. Don't know. It was smaller in height and slightly smaller in diameter. Later this evening we took another trip to do the same readings we did on the original pump. So here they are.

1st gear at 4k - 38, 2nd gear at 4k - 39, 3rd gear at 4k - 40, 4th gear at 3k - 40, 5th gear at 2.5 uphill again was 41. The pump cycle sounded quite different and resulted in 41 before start and 41 at idle. The other thing that was nice was the access panel I made for the fuel pump. About five minutes and I had the rivets popped out and ready to change the pump.

I added some injector cleaner called "Seafoam". It's supposed to work well. We will see. It will take a while for it to run through the system. So far, I'm still getting the misfire, stumble, or whatever you want to call it. Also, inspected the gas, tank, sock, etc. Everything was clean like new.

Next I will try the orientation of the MAF to see if that has any effect. I have a question: does the plug wire orientation have to be so that the number one plug is at 12 o'clock when the basic timing is at 10 degrees BTDC? I was wondering if the TFI picks up the signal at a certain place on the base of the distributor and if I have moved the plug wires over one place for distributor clearance, it may be picking up the signal to soon or too late. I never had to contend with that issue before as long as the dizzy was in the right position relative to the cam. If that is the case, I do remember changing all the wires over one position to get the necessary space to rotate the distributor. I did not see any marks on the distributor for that purpose. Maybe I missed something.

Thank you,

WEK

[Jeff Kleiner]

Bill,
Look at the top of your distributor cap. See the numeral "1" molded into it at what would be at or near the 12 o'clock position? That's where #1 needs to be, reason being that the ECU is looking at the signal from the PIP (Profile Ignition Pickup) sensor to sense crankshaft position and fire off the injectors. Although you can set ignition timing after reclocking the plug wires the injector timing will be out of synch because the ECU doesn't know that when the PIP armature in the distributor passes by #1 position it isn't really firing the ignition in that cylinder---in other words by manipulating the wires you have been able to make the spark occur at the right time but the fuel will be either early or late. Simple enough fix; pull the distributor and reinstall it so that when the rotor points at #1 it really is #1.

Jeff

[Carlos C]

WEK:

Jeff is absolutely right about why you need to clock the distributor the described way, but keep in mind that if you have an after-market unit, you need to follow their instructions. I have an Accel distributor on my 351, and its instructions specified to place the #1 cylinder wire on the 9 o' clock position (90 degrees from the molded "1" on the distributor cap, and then follow the typical firing order). I've never had a problem with my distributor, running it as instructed.

On another note, I still think you should test your O2 sensors. Good luck.

Carlos

[skullandbones]

The MSD distributor cap I have does not have a 1 molded on the top. The stock Motorcraft distributor cap does have one. So comparing them side by side, I have my number one in the same place as on the stock distributor. I'll be working around the engine bay now that I've gotten finished with the fuel supply side. Carlos, I will get to your suggestions but I am slow. They are on the list. If I don't eventually figure this out, it's going to be embarrassing to say the least. I'm already getting ribbed by my friends about when are you going to get that piece ........... running right? Oh well. Getting back to the distributor: I am wondering if the car could be running as well as it does and still be off one gear position on the cam. I will start from scratch and do TDC and make sure everything is lining up the way it should. Thanks for your help and suggestions, WEK.

[HCP 65 COUPE]

Skull

Have you installed a wide band o2 sensor? Its a great way to spot a lean miss in the fuel curve just before the miss happens it will go
lean and then get really lean for a second because the fuel did not burn in the cylinder.

This to me sounds like a lean miss make sure you dont have an injector wire that is grounding out intermittently and check your grounds

ECU's are very sensitive to ground loops.

Also is your fuel pressure regulator refrenced to the manifold? A stock ECU wont deal with it well if its not.

[HCP 65 COUPE]

Wek

Are you running a CDI type ignition box? Or have you modified the ignition in anyway?

The surging idle is most likely a lean mixture. Does it surge more, less or the same after
Warm up? Pay very carfull attention to the conditions when the engine misses like
During a throttle transition is it at cruise how much load is on the engine up hill or down.

If its totally random it's Likely electrical in nature groundsq would again be where I would start.

How did you delete the EGR system? I read the Instructions for the summit harness you are using and on page 8 it says the ecu will dump into limp mode if the EGR system is deleted without A chip burn/reprogram.


Try and pull codes off the ecu it may point you in the right direction a accutron code scanner is a cheap simple way. They make them specific for eec4s

Can you describe to us the way the harness and battery are grounded

[skullandbones]

I have a lot to do it appears. I don't know how to check the O2 sensors but I will research that.

HCP 65 Coupe: I do not have a wide band yet that's on the wish list along with the PiMP from Stinger, the front Wilwood 6 pot brakes, and a few other upgrades. But I will consider that as a high priority item since I have had that suggested by several forum members. I had not thought about using it to diagnose that issue though. I need the wide band as a component of the new computer so it is an add on that will help in the transition. Do you have a preference? I have heard of the Advent, I think it is called. According to Stinger, all of the popular devices are compatible with their unit. I have been very diligent in grounding things as I have read in the Ford Racing articles that it is very important for EFI. I ran a ground from the bell housing bolt to the computer frame (braided OEM ground wire). I ran ground from battery all the way to the block. I have dedicated ground blocks for grounds in the dash area and in the rear of the car. The EFI harness is new. I'm not sure how I would introduce a ground problem there. I am going to check the injectors for impedance sometime down the line. I do not understand "referencing FR with manifold" which is aftermarket. You will have to explain that one. I don't have a CD system or any other aftermarket ignition items. The surging idle is worse after the engine warms up. The stumbling is when the engine is loaded but continues intermittently even afterwards with a distinct change in exhaust note that can be recognized as it changes from normal to stumbling mode. No particular single rpm range. I dced my EGR by removing the device but left the valve connected to the harness and depressed the plunger to mimic an active EGR so the computer would "think" it was operating. There is no CEL due to that but until I adjusted it to the mid range of it's travel there was a CEL. I will eventually change that valve out for a resistor to clean up that look on the harness. I plan to have the codes pulled as soon as I get a chance. It might also tell me if there is a stuck or grounded injector if I find a misfire code of some kind. I have a couple of pics of the engine to see if the rotor and #1 look right. The balancer is on 0 degrees and I checked to make sure that the cylinder is at TDC with my inspection camera. In the second pic the rotor is pointing at about 11:30 from the front of the engine. The cap in the next pic is showing #1 at the same relative position closest to the D in MSD. The last pic may show the PIP sensor. Is that it: the little black thing under the rotor?

Thanks,

WEK.

[CHOTIS BILL]

A lot of my posts don’t get posted so I am not sue this one will make it but I will give it a try.

HCP are you sure the sensor will read lean? The way I remember it on mine every time I had a miss I would get a rich reading because the oxygen level was high because it wasn’t being used for combustion. It been a long time since I looked at this and as always I may be wrong.

Bill Lomencik

Skull

Have you installed a wide band o2 sensor? Its a great way to spot a lean miss in the fuel curve just before the miss happens it will go
lean and then get really lean for a second because the fuel did not burn in the cylinder.

This to me sounds like a lean miss make sure you dont have an injector wire that is grounding out intermittently and check your grounds

ECU's are very sensitive to ground loops.

Also is your fuel pressure regulator refrenced to the manifold? A stock ECU wont deal with it well if its not.

[Carlos C]

WEK:

The best way to troubleshoot the O2 sensors, or any other issues for that matter, is to invest on a code scanner. Some out there are as cheap as $50 and still reliable. Also, companies such as Sunpro make sensor testers, which have come very handy for me multiple times.

The following are the steps to testing an O2 sensor with a volt meter and a propane torch. Parameters to keep in mind: LEAN = too much oxygen in the exhaust (0.4 volts DC or less); RICH = not enough oxygen in the exhaust (0.6 - 1.1 volts DC); optimal is about 0.5 VDC, but the reading will fluctuate depending on how much heat you're applying to the sensor, and how close the heat source is to it. More on that shortly. The important thing is that the sensor voltage reacts the proper way during the steps, and that you're supplying enough heat.

1. Remove sensor from vehicle, and mount on a vise by its body, leaving its ceramic tip pointing up.
2. Attach the volt meter leads to the two sensor wires (polarity does not matter). If you happen to have a one-wire sensor, then connect the positive lead to the sensor wire, and the negative lead to the sensor body, as far away from the tip as possible. At this time, the voltage reading should be very low, as there's plenty of oxygen around the ceramic tip and no heat (lean).
3. Using a propane torch, apply heat to the sensor tip, heating it to about 660*F. Do not allow the tip to glow. At this point, you should see the voltage rise to a point below 1.0 VDC and then stop.
4. Then, bring the flame close enough to the tip to completely smother it. This will re-create a "rich" (low oxygen) condition, and the voltage should jump up to about the 1.0 VDC mark. This step should only take a few seconds. Do not leave the flame on the sensor for too long, or it will get damaged.
5. Completely remove the heat from the sensor. Within 5 seconds, the voltage should drop showing a "lean" (too much oxygen) condition.
6. Repeat the above steps on the second sensor.

If during any part of this test, the voltage does not react the way I described above, the sensor is bad. If you're not sure if it failed the test, you can always let the sensor naturally cool down, and the repeat the test. If you have any questions let me know. The reason I strongly believe that your culprit is an O2 sensor is because you keep describing a stumble and symptoms which I had before, and was caused by a faulty O2 sensor. Hope this helps.

Carlos

[HCP 65 COUPE]

Bill,

An o2 sensor measures o2 % in the exhaust gas if there is a miss it will read lean because the cylinder did not burn off any of the air.

Fuel or no Fuel.

Sometimes what will happen after a miss is the ECU will see lean and then compensate and add to much fuel then it gets rich.

Wek,

what I was trying to ask was, is the fuel pressure regulator vacuum line connected into the manifold and I think you said it was on the other page.

Ref is short for Vacuum reference. I don't type so good or fast.

With you saying that the surge gets worse after warm up. I'm almost certain its lean @ idle.

The warm up enrichment is helping the surge before the engine gets to temp.

Is there a bend even close to your maf? If there is the air sometimes sneaks past the sensor hanging in the middle, because the air
will kinda hug the walls of the tube in the sensor and more air will get by causing a lean mix.

As far as widebands go I have only used Innovate motorsports so I cant say about any others.

As far as the grounds are concerned even with a brand new harness they can cause problems because a bad ground will cause
a voltage drop (ground loop) across a sensor giving false readings, another thing to think about is EMI causing sensor noise.

Make sure your nosiy wires(spark plug wires and alternator feeds) are well away from any sensor wires when ever possible

The black thing under your rotor is the hall sensor and reluctor that triggers the hall sensor its the modern day equivelent of points.

I think the TFI pip signal does come from there with a wider or narrower gap in the reluctor but Im not certain. Some one else may
know that one better than me.

[Carlos C]

WEK:

I had previously addressed Bill on my last post (#53). I've already edited it, as it was meant for you. Sorry to both of you for the confusion.

Carlos

[DaleG]

A lot of my posts don’t get posted so I am not sue this one will make it but I will give it a try.

Bill Lomencik

Bill, happens to me all the time; I write, then forget to hit "Post Reply".

[skullandbones]

I took the roadster to the parts store to have the codes checked. I had an opportunity to go on the freeway for a little bit and got to pull it thru the gears which surprised me. It ran perfectly as it has before this issue. Ran thru from 2 to 5K rpms with no hesitation or stumble. It was quick, crisp, and loud. So I got to the store and had these codes pulled.
DTC:34 insuff EGR flow or EVP voltage
DTC:67 neutral switch error
DTC:81 air divert solenoid
DTC:82 air divert super charger'
DTC:85 canister purge error
DTC:84 vacuum regulator
DTC:29 insuff input vss
DTC:33 EGR valve not open
DTC:96 fuel pump secondary circuit fail

I was expecting some misfire codes but nothing found. The only thing I did different from yesterday is take the distributor out and check placement. Replaced it in the same position. Moved timing back to where it was originally at 13 degrees (without spout). I also moved some SP wires to better locations so not to cross as much and reattached one at a time. I put some injector cleaner/ fuel conditioner in the tank which had about a half tank full at that time. So I am a little perplexed at this. I suppose it could have been a loose plug wire but there were no codes to support that. If it was a ground loop short what happened to it? I am planning a trip to the dyno shop to discuss with the guy to see what I should do about the codes if anything. I don't know if he will see something the reader did not detect. But at this time, I'm happy that it's running as well as it is. If it continues, I will have to assume I may have had a dirty injector that was cleaned sufficiently to quit acting up. I looked at the other plugs yesterday, too. They all looked pretty much perfect. It's kind of a head scratcher for me. I will drive it again tomorrow to see if I have any more of the bad symptoms.

Note: I may have thrown some of those codes during removal of the EGR as I did get CELs while adjusting the valve to override the CEL. I may need to delete the codes and start over to see if they re-occur or not. Suggestions?

Thanks, WEK.

[Carlos C]

WEK:

That's a good collection of what's called "soft codes". Besides maybe #96, they won't affect drivability. Most of them are actually emissions-related, which are omitted from these builds. The way to erase these soft codes is by disconnecting the fuse leading to the ECM, or disconnecting the battery.

Keep in mind that there are separate tests that need to be done with a code scanner: key on, engine off; and key on, engine on. There are also portions of the tests where you need to conduct a wide-open throttle (WOT) exercise. This is the way to get the full spectrum of potential codes.

Carlos

[skullandbones]

That's what I was thinking too. We did the key on and engine running but did not pull any more codes while the engine was running. Didn't do the "WOT" thing. First, I've heard of that. What might be expected from that exercise? Anyway, I will delete the codes and start fresh. I'm hoping that it continues to run well. I just always like to know why and that's not happening here. Wow; it sure is a completely different animal when it is running right! Thanks, WEK.

[HCP 65 COUPE]

Im glad things seem to be working better maybe it was just a pluged up injector hopfully it does not return.

A pluged injector would explain a lean miss and the surging idle(lean idle mix)

[skullandbones]

Im glad things seem to be working better maybe it was just a pluged up injector hopfully it does not return.

A pluged injector would explain a lean miss and the surging idle(lean idle mix)

Me too! How come they didn't translate to some sort of code? I had a similar problem with my DD EFI and it threw specific cylinder misfire codes as well as random misfire code and a CEL. All hell broke loose. But on this one, there is no trace of the computer seeing a misfire. That's weird! WEK.

[Jeff Kleiner]

EEC-IV does not have the capability to detect or report misfires.

Jeff

[Carlos C]

WEK:

Each code scanner has a different way of letting you know when you need to do the WOT test during the key on, engine on procedure. The instructions will specify this. The WOT test checks for timing advance, engine knock, injectors timing and duration, exhaust mixture, throttle position, etc.

The soft codes will re-appear, but you shouldn't concern yourself with them, as long as they pertain to items that have been purposely eliminated from the system (i.e. Emissions).

Carlos

[skullandbones]

EEC-IV does not have the capability to detect or report misfires.

Jeff

OK, Jeff. Is there an equivalent or any thing that you might recognize as that type of code?

WEK:


Each code scanner has a different way of letting you know when you need to do the WOT test during the key on, engine on procedure. The instructions will specify this. The WOT test checks for timing advance, engine knock, injectors timing and duration, exhaust mixture, throttle position, etc.

The soft codes will re-appear, but you shouldn't concern yourself with them, as long as they pertain to items that have been purposely eliminated from the system (i.e. Emissions).

Carlos

I will have to research that. The guy at the parts store didn't even mention it. It would be interesting to see some of those codes looking back on what I have experienced the last 600 miles or so.

I'm not sure about all the "soft" codes coming back. I think I generated a few of them when I was doing the un-installs. We will see though.

Thanks guys. WEK.

[Jeff Kleiner]

OK, Jeff. Is there an equivalent or any thing that you might recognize as that type of code?

Probably the closest you would come would be a code 42 indicating a rich condition, but I think the problem would have to be chronic rather than just a random misfire for it to log it.

By the way, a simpler way to clear code memory rather than disconnecting the battery is by going into KOEO (key on engine off) self test mode then if you remove your jumper between the "self test input" and "signal return" terminals while it is displaying faults it will erase any that are stored.



Cheers,
Jeff

[Carlos C]

WEK:

Auto part store clerks are not technicians by any stretch of the imagination. I have personally come across some that are completely clueless. They just get some training on how to use the store equipment, and some grasp the information better than others. You'd be better off getting a scanner, read the instructions, and do the tests yourself. You'd still spend a lot less money than taking it to a real shop.

Any codes, soft or hard, not resolved before you reset the computer, will come back the next time you turn the key on and/or start the engine. The term "soft" is referred to codes which do not affect drivability. Hard codes wreak havoc on your engine's functionality.

Carlos

[CHOTIS BILL]

HCP,

You are correct. I went back and checked my data from the last time I had a misfire in ’09 and the O2 defiantly goes lean on with a misfire.

Bill Lomenick

[skullandbones]

OK. I have looked into the codes on several sites. It does appear that a lean or rich event could be logged from the O2 sensors but I don't know if those would only be found with the engine running (history). It's not consistent for these to be absent at this time with all the trouble I have had over 600 miles of driving. However, if there is a basic difference in the two types of code reading systems so that that is possible, then I guess it could happen. I have never dealt with OBD 1 before this car. If that's the case, I will go back for another session to see if the codes show up.

Question: does the engine have to be running to catch those codes?

Thanks,

WEK.

[skullandbones]

Thought I would bump it up to see if I get an answer on the question on Post 68.

Also, I've put this little problem in perspective after some time thinking about it. Sometimes we have problems that just defy resolution. In QC we called it a "flier" in the statistical sense. We never had time to give it the time and resources to dig to the bottom of the problem. It's a good thing it didn't happen more to become statistically significant. Well, that's the way I'm treating this situation. I think I experienced a statistical flier for lack of a better explanation. I was watching the drag races this weekend and three different teams talked about their "gremlins" that were inhabiting their clutch or black box or fill in the blank. I guess it's OK to have a gremlin now and then. I just hope mine stays away fro a while. I'm having too much fun driving this thing when it is right! After several drives this week, it looks like things are OK at least for now.

Note: I wanted to get my fuel pump changed out and have the engine running as good as I could get it with the basic timing and such. So it's there now and I will be taking it to the shop to do a baseline on the engine to see if my 300+hp crate engine has improved with the upgrades I did during the install.


Thanks, WEK.