Intermittently runs poor at low TP

[portmod7]

I am finally working on this again. I thought maybe distributor endplay was causing the timing to jump around, but I pulled the distributor out and shimmed the endplay as tight as I felt comfortable and it didn't seem to make any difference.

At idle, the timing doesn't jump around at all. But it seems to be dependent on RPM. Around 1500-2000 RPM it starts to jump. If I continue on to say 3000 RPM, sometimes it seems like it will advance another 8 degrees or so for a total of around 16. When it jumps, the engine definitely runs differently, so I don't think it's my light, the timing is definitely changing. Again, all of this is with the B pin disconnected.

One other thing I thought to try is to see if I could see the timing jump around with the B pin connected. I connected the B pin and the timing seemed to be very smooth and correlated with what TunerStudio showed. However, one thing I noticed, as the RPM's increased the flashes of the timing light got closer together to the point where it seemed like the light was on constantly, which is what I would expect. However, at 2500 RPM with the B pin connected, the flashes of the timing light got noticeably further apart, the light didn't seem to be on "constantly" any more. I didn't notice that with the B pin disconnected. Maybe it has something to do with the "base timing" jumping around and the ECU trying to keep up with it?

Thanks for any help!

[Matt Cramer]

Was the timing actually jumping with the B pin connected?

Do you have a data log of what happens at 2500 RPM with the B pin connected?

[portmod7]

With the B pin connected, it didn't jump around like it does with the B pin disconnected. However, I did notice it jumping erratically for a split second a couple of times at 2500 with the B pin connected. It's nothing like what it does with the B pin disconnected though. With the B pin disconnected it will jump the 8 degrees and stay there until the RPM or whatever is causing it changes. But with the B pin connected, at 2500, the timing seems to hold steady for the most part with the timing light flashing slow, but once in a while the timing will bounce around for an instant, too fast to even tell how much it changes.

I will get a data log together for both situations.

[portmod7]

I got a couple datalogs tonight, but it didn't go as expected. I still have the same issue with the B pin disconnected, but today it doesn't act the same way it did yesterday with the B pin connected. Now, with the B pin connected, when the ECU advances the timing, it advances so far the motor doesn't even want to run. It's WAY different than yesterday and I didn't change a thing.

Tonight, If I set the "base timing" with the B-pin disconnected and then connect the B-pin, to get my timing light to match TunerStudio I have to manually retard the timing at the distributor so far that the engine won't even run with the B-pin disconnected. TunerStudio indicates that the ECU is calling for appropriate timing, and at idle I can match it with my timing light, but as soon as I open the throttle and the engine gains RPM and the timing advances drastically with the timing light and no longer matches TunerStudio.

With the B-pin disconnected, toward the end of the datalog, hopefully you will be able to tell that the throttle was staying pretty consistent at 8%, but the RPM was varying a lot. This variation was when my timing light would indicate the jump from "base timing"(lower RPM) to advanced an extra 8 degrees(higher RPM). Holding the throttle at 8% seemed to be the spot where the timing wanted to alternate.

With the B-pin connected, you can see that I try to pick up the throttle and then the motor starts falling on it's face. The timing light indicated that the timing was advancing FAR beyond normal, so much so that the engine would barely run.

Take a look at my tune, and the datalog, let me know if you see anything in there that looks like a problem. Thank you!

[portmod7]

After I slept on it, I realized what actually changed. When I said "I didn't change anything" in the post above, I was wrong. The day before, when I was running my "tests" and didn't get datalogs, I had changed the "spark output" setting to "going low(normal)" before I started the engine. However, I did not cycle power to the ECU before I started the engine and ran my "tests". So, I am pretty sure the ECU was operating with the "spark output" setting of "going high(inverted)" even though I thought I had changed it to "going low(normal)". I didn't even cycle ECU power when I disconnected the B-pin. So, when I came back to the project to get datalogs, the ECU power had been cycled at that point, and the "spark output" setting of "going low(normal)" went into effect and caused my screwy results.

Today I confirmed my mistake. I set "spark output" to "going high(inverted)" and then cycled power to the ECU. After I did that, the engine was performing like I described the day before. I got two new datalogs today with the engine behaving like I described in my post dated 2/4/15. My timing light didn't show the timing to bounce around abnormally at all with the B-pin connected, only with it disconnected. With the B-pin connected, the timing shown by my timing light correlated with the TS dashboard.

ALSO, today, I tried to play around with the "trigger offset" and the trigger wizard. However, manipulating the offset didn't change anything. It didn't change what my timing light was showing at all. Correct me if I'm wrong, but changing the offset should change what I see with my timing light, right?

Sorry for the confusion, thanks for the help! Should I repost this issue in the "Ignition Troubleshooting" section?

[portmod7]

I figured out what I was doing wrong with the trigger wizard, I wasn't hitting "burn" after I made changes. I used it today and it was working fine.

I went ahead and set my trigger offset and drove the truck a little. It still has the same low throttle position bog, or whatever you want to call it. And, as described before, sometimes it has it, and sometimes it doesn't, it seems like it's all dependent on how the ECU boots up each time. I got a datalog today, but it's evidently too big of a file to upload, 5.5MB. Seems like 70% TP is the magic number, once the TP hits that, it runs fine, but below that it seems like the timing is off, or the fuel mixture is real lean, or something. I can cycle the power to the ECU until it boots up in whatever way that makes the engine run decent at lower TP's.

Is it possible to send my ECU to DIYAutoTune, or someone else to have them check it out, or try to run a motor with it or something? I've also recently relocated to Mooresville, NC, I see that the "2015 Mega Meet" is going to be here this spring, is there somebody close to the area that might be able to take a look at my project in person?

Thanks!

[trakkies]

I'd say if any software is being corrupted, it would show as a difference report when you connect TS - assuming you have saved the tune before quitting.
IMHO, it's far more likely to be a sensor 'funny'. But again, TS can give a clue. Do both the ATS and CTS show ambient temp on a cold engine? Does the TPS show the correct throttle position on a cold engine, before you start it? Is the barometric sensor showing air pressure? Etc.

[portmod7]

So, I figured maybe the best way to check my sensor readings was to get a datalog during a series of ECU power cycles without starting the engine. Does that sounds reasonable? Since it seems like my problem comes and goes with ECU power cycles I thought I might see an inconsistent sensor reading through several cycles. I've attached the datalog and the tune. I don't see anything that catches my eye, but maybe one of you who knows a little more about it will see something?

If there is a better method to go about this, please let me know.


Thanks!

[portmod7]

Ok, I just wanted to update this. It seems like 95% of the threads I read just drop off with no solution, so I wanted to say that I found the problem.

It seems to have been that I had the four injectors paired up side to side on my throttle body. I switched them so that they are paired up corner to corner and it has solved my issue. I was/am running 4 squirts per cycle and alternating.

I think the problem stemmed from running the "unique" FAST throttle body on top of a dual plane intake. I say "unique" because it is my understanding that most throttle bodies run low fuel pressure and the injectors inject fuel into the barrels of the body. The FAST throttle body runs at 43psi and the fuel is injected below the baseplate.

All I can figure is that with the original configuration of my injectors, each side of my dual plane intake only received fuel every other "event". For whatever reason, at large throttle openings the air was able to scavenge fuel from the opposite side of the intake, but not at lighter throttle openings. At first I thought my TPS reading was somehow altering fueling, but I ended up unplugging it, and it changed nothing, so that put me on the path of blaming the air/fuel flow in the intake at different throttle positions.

Now, both sides of the intake are getting fuel every time there is an injector event and the engine is happy. Or at least that's the way I see it. I've already been told by a very knowledgeable source that TBI's are basically fool proof, dual plane or not, and what I've done really shouldn't have changed anything. Again, I figure that maybe this FAST throttle body is a different animal than the majority of the EFI community is used to dealing with. But I could be wrong about all of it. But my project seems to be running like it should now, I changed A LOT of stuff in the process, and an injector pairing change is what did it.

Hope this helps somebody down the road, thanks to everyone that offered me help with it!

[portmod7]

I forgot to add the reason I thought the problem was intermittent and depended on rebooting the ECU... Since in my configuration the ECU does not know where the engine is at in its cycle, the timing of the injection events is started from the first time the ECU picks up the tach signal. I think that randomly, this would happen at such a point in the engine cycle that the alternating fuel distribution to either side of the intake coincidently happened in sync with a majority of the left and right banks' need for fuel, so it would run decently when the injection events happened just right. That's the only solution that made sense to me.

[portmod7]

Back again, one more time. I was still having some issues, although different, and was pointed toward "injector data". So I rigged up a test rail on my engine, and tested my injectors. Come to find out, 3 of them were only flowing about 2/3 of what they were supposed to be. So, I sent them off and got them cleaned. When I got them back, the cleaning report backed up what I saw when I started trying to test them. These are FAST injectors, #80527(65lb/hr)

I got them back on the truck today and tested them. They all flow much better now, just not as identical as I had hoped. Is that the quality of the injectors I have, the quality of the cleaning they got, or is that just the way injectors are unless you have them professionally matched? When I tested them, I ran them for 7 seconds and ended up with flow rates of 64, 66, 68, and 77lb/hr @43psi in no specific order. I had deadtimes of .571, .458, .724, and .766ms respectively. Battery voltage corrections of .178, .169, .145, and .152ms respectively.

Since I had two pairs with similar dead times and voltage corrections, I paired them up that way on each of my two banks and entered two sets of deadtimes and voltage corrections in TunerStudio. I also put the higher flowing injectors diagonal from each other on my throttle body. I am running 4 squirts alternating and can idle it down to 500rpm if I choose to. Hopefully through all of that I have averaged out the difference as good as possible without having them professionally flow matched.

It runs better than it ever has, as you would probably expect. And with the good injector data, I hope working on my tune at all temperatures and pressures will go a little better now. I've had this thing running for 3 years or so now and have fought strange issues the whole time. I've never really solved any of them. But this exercise has made a world of difference for the better. I am optimistic that this is the main issue I've been fighting the whole time. Maybe I missed it, but in putting this system together and learning about this stuff, I never found one place in the "instructions" that told me to flow my injectors to get data on them. I sure wish that would have been in there somewhere, probably would have saved me a lot of headache. But maybe it goes without saying, and this is just another example of me being out of my element. But I am learning!