EFI Puzzler -- higher pressure, leaner burn

[parsonsj]

Mike Norris and I were messing around with II Much yesterday, and one of the things I wanted to do was to increase the fuel pressure in the fuel rails. I've been running at 42 psi (or 38 depending on which gauge you believe).

I wanted to run at 58-60 and see if there was any more power or if the idle would be smoother.

I bumped the pressure to 60 (Aeromotive rails, regulator, A1000 pump) with no problem. However, the motor went lean. Mike started tossing in fuel and it made no difference. Thinking the o2 sensor might be the culprit, we swapped it over to Mike's dyno sensor, and saw the same results.

Next, we reset the pressure to 42, and reloaded the original fuel map that Mike had saved (he's a smart one). Everything went back to normal. Then we increased the fuel pressure with the engine running. The engine went progressively leaner as we did that.

Not knowing what else to do, we put everything back and I drove the car home.

The only thing we could think of is that the injectors (Holley 42 lb) aren't rated at pressure higher than 42 (45?) and the high pressure prevents them from allowing more fuel by. Does that make any sense at all? Is there a different brand injector that we should try? Or is it even worth trying to get to 60 psi?

thanks!

jp

[Fuelie Fan]

This is in interesting one, I'll ponder it. However, I really don't understand the rationale of blindly changing pressure. First and foremost, smoothing idle and making more fuel available at WOT would require OPPOSITE adjustments in fuel pressure. Assuming your idle pulsewidths are too small and this creating your idle instability, you would need to lower pressure, not increase it.

Secondly, what actual data do you have the suggests ANY need for fuel pressure adjustment? Waht are your pulsewidths at idle? What is your max duty cycle? What type of injector are you using?

[parsonsj]

We weren't "blindly" changing pressure. We were conducting highly scientific experiments. :)

Seriously, we were trying to emulate late model EFI paramaters, in particular GM LSx engines over the past decade. They run 58 psi, in part, I believe, to reduce vaporization in the fuel rails due to engine heat, but also maybe to improve fuel atomization.

My car is just a bit sticky starting (I have to crank for 6-8 seconds) on a cold engine. Other changes (IAC, fuel enrichment, etc.) have made little if any difference. Warm engine starts in 2-3 seconds. Still not OEM by any means, but I'm using FAST's Crank Sensor convertor, and it takes two revolutions before the ECM finds the crank. So idle wasn't the primary issue, cold start was. (I mis-spoke above).

Also, the car's power is lower than the engine dyno would suggest it be (620 at flywheel, 460 at rear tires). So we thought it worth a try. The engine shop dyno number came from a Pantera EFI setup that stayed at the engine builder. I've changed cams too, to something more streetable, so that might explain the power difference as well.

Interestingly enough, lowering the fuel pressure below 42psi also leaned the mixture at idle. It seems that the ideal fuel pressure for my injectors (Holley 42 lb) has a bell curve to it, and the best point for them is around 40-42. We didn't know that until we were tinkering last night. I found that surprising, and wondered if there was a reason for it. Will other injectors have a different sweet spot?

Good questions about the pulsewidth ... and I don't have those answers. My copy of the current program got sent in with my busted laptop, and I've never saved the program since. I'll see if Mike will post those numbers up later.

thanks,
jp

[68Formula]

Sounds like the idea was to improve the atomization by increasing the pressure. Probably either the increase in pressure was too much for the injectors slowing them to open, or the ECU was pulling out fuel to compensate and got into an pulsewidth outside the injector linearity range.

[parsonsj]

Sounds like the idea was to improve the atomization by increasing the pressure. Exactly. We thought there might be some power there, with a bonus of helping cold start.


either the increase in pressure was too much for the injectors slowing them to openThat occurred to us (eventually). Anybody know if the Holley 42 lb injectors are not rated for 60 psi?


the ECU was pulling out fuel to compensate and got into an pulsewidth outside the injector lineariy range.I don't think so. At one point we told the injectors to use as much fuel as at WOT. It made no difference. I don't *think* (but what do I know??) the ECM would behave like that.

jp

[Fuelie Fan]

I never tested the bosch style injectors above a DP of 45 psi, so i'm not sure. But, in general, no it's not a bell curve. I'd be quite surprised if they couldn't handle that level of pressure, but it's certainly possible.

The pressure helps with atomization, but you'd probably see bigger gains in that arena by going to a newer injector with a better spray pattern. The newer ones also have a larger linear response range, ie they can operate repeatably at lower pulsewidths. The pressure and injector design were probably developed in unison; I couldn't tell you what the reduction in droplet size would be going from 45 to 60 psi with your current injectors. Obviously pressure is the cheap solution, but better injectors would be a better solution.

Does cracking the throttle open affect crank time at all? Sometimes you can use that as a quick check to see if you're flooding with too much prime pulse.

[andrewb70]

The only thing we could think of is that the injectors (Holley 42 lb) aren't rated at pressure higher than 42 (45?) and the high pressure prevents them from allowing more fuel by.

thanks!

jp

Ding...Ding...Ding....

The Holley (Delphi) injectors are not happy above 50psi. Stock LS plastic, pencil style (Bosch I believe) were designed to operate at 58.5psi and seem to have no issues going higher.

Andrew

[parsonsj]

Geez Andrew, I tried to call you first. I figured if anybody knew about that you would.

So where were you last Friday? :)

jp

[parsonsj]

Does cracking the throttle open affect crank time at all? Sometimes you can use that as a quick check to see if you're flooding with too much prime pulse.Good question. I'll check that today. I'd really love to fix this annoying little problem.

thanks!

jp

[68Formula]

I never tested the bosch style injectors above a DP of 45 psi, so i'm not sure. But, in general, no it's not a bell curve. I'd be quite surprised if they couldn't handle that level of pressure, but it's certainly possible.

The LSX engines that he mentioned running at 58 psi are Bosch injectors so clearly they can handle it. Clearly they like 58 psi. He is talking about Holley injectors which mimic the old Bendix design.

Another aspect to consider is the injector drivers.

[Mike Norris]

Hey All,

Thanks for all the input for sure and I will try to shed a little more light on the subject.

I do not know what the pulse widths were at idle as I was initally stunned at what was happening. As far as WOT at 60psi, we never went there, but I would think the injectors were at 70-75% duty cycle at 42psi and 6500 RPM. I ned to go back to past scans and check.

Of course the day we have issues, I could not get anyone on the line from Holley, FAST or other local tuners to bounce things off.

As John mentioned we were trying a test to see if (1) the long start would be better and (2) if there was any power gains with the higher pressure/better atomization. I had used the Holley/MSD, RP style injectors in LSx cars before with mixed results.

John raised the pressure to 60 and I dropped the complete VE table 30% to start. Show lean lean lean. Kept rasing the VE and finally about tripled the VE numbers at idle from 38 to 100 and she idles nice at 14.0-1. We go back to 40 psi and the original tune and she runs great. Raise the pressure from 40-45 PSI, she leans out.

Needless to say that all we learned was do not try this with with these style injectors. I do have a set of 42lb Ford SVO injectors at the shop I always use in high fuel pressure cars (60-70 PSI) along with a set of Precision Turbos 42's that always work great. We will swap them at some time and see how things go. I can also pay a little more attention the the starting and driveability at that time.

Later All.

Mike Norris

[parsonsj]

Cranking pulsewidth is about 20ms at 70 degrees, and drops to 12.8 at about 135.

jp

[parsonsj]

Today I bumped the IAC start position up about 20%, and configured the IAC to allow faster and bigger movement.

The car started much faster, and there wasn't any "sag" after initial fire. A very nice improvement. Tomorrow, I'll bump the cranking fuel up just a bit (say 10%), and see if that has a positive effect as well.

[Fuelie Fan]

Raise the pressure from 40-45 PSI, she leans out.



Wait are you saying that just changing from 40 to 45 caused it to lean out, or is that an incomplete statement that would finish as "...to 60 psi" ??

[Fuelie Fan]

John, I don't think I'd start adding fuel. Raising IAC base is the same as cracking the throttle; you are rich at startup.

[parsonsj]

I don't think I'd start adding fuel.No? I'm listening... but I think it can start a bit faster. What would you suggest?

jp

[parsonsj]

Today's update: I sorted out the sync signal from the cam. The firing order is 18726543, with the cam sync signal preceding cylinder 5.

Now when I retard timing on cylinder 1, it actually retards timing on cylinder 1, lol.

Today's cold start was more of a sag, even though I didn't make any changes to the cranking fuel or IAC. Today's change was to change the run/crank rpm number to 150 (from 400), and to have it start the fuel enrichment decay at 6 revolutions rather than 12.

I'll have another go at it tomorrow.

jp

[Mike Norris]

Oh the tinkering we do :)

Fuelie Fan, it was not a mistatement. When we raised the pressure from 40 PSI to 45 PSI, the car leaned out. Went to 50 PSI and it leaned out more. Hope this helps.

Mike

[parsonsj]

Mike,

You'll be happy to know I've fixed the brake lights, and fuel sending unit. :)

jp

[Fuelie Fan]

Well then this is more than what Andrew was suggesting in terms of it just being too much pressure. They're designed to operate at 43.5 psi, and given that there are manufacturing tolerances on EVERYTHING it would be completely unacceptable for them to be non-linear beginning at (or actually before!!!) their designed operating point.

[terryr]

The old 1980s Dodge 4 cyl Turbos bumped fuel pressure up to 70 psi from the factory. I don't think they were special injectors in any way. [bosch?]

[68Formula]

The old 1980s Dodge 4 cyl Turbos bumped fuel pressure up to 70 psi from the factory. I don't think they were special injectors in any way. [bosch?]

Between manufacturers not all have the same linearity range even for the same injector type.

Also, I am almost certain the Chysler Turbo cars used "peak and hold" injectors and drivers. So yes, they were special.

[Fuelie Fan]

Also, how much boost was it running? You have to remember that it's the pressure difference across the injector that's important. If the manifold has 20 psi in it, then even though the gauge shows 70 psi the injector is exposed to a differential of only 50 psi.