MustangWorks.com : Ford Forums - View Single Post - How low can I safely set my fuel pressure?

jimberg · 06-06-2001, 02:03 PM

I think we're almost there, red82gt.

First, I'm not disputing what you said about lowering pressure to make a 24#er act like a 21#er. My point is that there is no benefit from doing that. In fact, by lowering pressure too much you could cause the injector to start dribbling instead of spraying. At what point that would occur, I have no idea. Since the MAF sensor is calibrated for the injector size, the pulse width will be correct without much adaptation.

Your quote of the article supports what I've been saying. The computer has no clue what the fuel pressure is and definitely has no way of controlling it. The temporary fix they are referring to is the air/fuel ratio throughout the rpm range. If you raise fuel pressure you will richen the mixture but then after a while the computer will adapt back to what its tables say it should be at. Where it matters and is a permanent fix is at the top end.

Let's say that an injector (24# @ 39psi) can open for a maximum of 10 msecs. That would mean approximately 0.00007 #s of fuel for that pulse length. The computer wants to have the air/fuel ratio at 13:1. That would mean that only as much as 0.00091 lbs of air could be in the cylinder in order for the computer to keep that a/f ratio. If we get 20% more air than that (0.001092#s which means .000084 #s of fuel is needed), we're going to be much leaner than what the computer wants because we passed the physical limits of the injectors at 39psi. We'll also be losing a considerable amount of HP. By increasing fuel pressure, we can increase the amount of fuel able to be delivered for that 10 msec pulse. To compensate, we can bump it up to the point that it can deliver 0.00009 #s of fuel. Now the computer can keep it at 13:1.

The other option is to use bigger injectors, 36#ers for instance. They can deliver 0.00010 #s in 10 msec. We only still need .000084#s though. The computer just opens the injector for (You should like this since it may explain part of what you've been claiming to some degree) 8.4 msecs. But, if the computer can only control the injector width in 1 msec increments (probably not that coarse a timer, but should have some resolution limits), it will either choose 8 or 9 msecs. 8 delivering a little less fuel than needed and 9 delivering a little more. In closed loop operation the computer would alternate between the two times since it's going to base its decisions on what the O2 sensors are reporting. In open loop it will probably just pick one or the other. This is were it's possible that small tweaks could make a difference since you could reduce or increase fuel flow to fine tune the margin of error caused by timer resolution. This would all depend on the resolution of the timer, of course. If it can control at a resolution of 10 thousandths of a second than it probably wouldn't help much.

Dan's calculator simply tells you what size injector you need and what fuel pressure would give you that throughput. I don't think it's there for recommendation of fuel pressures. Maybe he should add a link to the EEC article on the results page, though. We can suggest that.

You're right about the computer being able to protect from being overly lean at WOT. Code 43 covers that. This means that the O2 sensors are monitoring at WOT, just not for the purpose of adjusting a/f ratio.

------------------
351W 89 Mustang GT Convertible

06-06-2001, 02:03 PM	#20
jimberg Registered Member Join Date: Oct 1998 Location: Rogers, MN Posts: 2,089	I think we're almost there, red82gt. First, I'm not disputing what you said about lowering pressure to make a 24#er act like a 21#er. My point is that there is no benefit from doing that. In fact, by lowering pressure too much you could cause the injector to start dribbling instead of spraying. At what point that would occur, I have no idea. Since the MAF sensor is calibrated for the injector size, the pulse width will be correct without much adaptation. Your quote of the article supports what I've been saying. The computer has no clue what the fuel pressure is and definitely has no way of controlling it. The temporary fix they are referring to is the air/fuel ratio throughout the rpm range. If you raise fuel pressure you will richen the mixture but then after a while the computer will adapt back to what its tables say it should be at. Where it matters and is a permanent fix is at the top end. Let's say that an injector (24# @ 39psi) can open for a maximum of 10 msecs. That would mean approximately 0.00007 #s of fuel for that pulse length. The computer wants to have the air/fuel ratio at 13:1. That would mean that only as much as 0.00091 lbs of air could be in the cylinder in order for the computer to keep that a/f ratio. If we get 20% more air than that (0.001092#s which means .000084 #s of fuel is needed), we're going to be much leaner than what the computer wants because we passed the physical limits of the injectors at 39psi. We'll also be losing a considerable amount of HP. By increasing fuel pressure, we can increase the amount of fuel able to be delivered for that 10 msec pulse. To compensate, we can bump it up to the point that it can deliver 0.00009 #s of fuel. Now the computer can keep it at 13:1. The other option is to use bigger injectors, 36#ers for instance. They can deliver 0.00010 #s in 10 msec. We only still need .000084#s though. The computer just opens the injector for (You should like this since it may explain part of what you've been claiming to some degree) 8.4 msecs. But, if the computer can only control the injector width in 1 msec increments (probably not that coarse a timer, but should have some resolution limits), it will either choose 8 or 9 msecs. 8 delivering a little less fuel than needed and 9 delivering a little more. In closed loop operation the computer would alternate between the two times since it's going to base its decisions on what the O2 sensors are reporting. In open loop it will probably just pick one or the other. This is were it's possible that small tweaks could make a difference since you could reduce or increase fuel flow to fine tune the margin of error caused by timer resolution. This would all depend on the resolution of the timer, of course. If it can control at a resolution of 10 thousandths of a second than it probably wouldn't help much. Dan's calculator simply tells you what size injector you need and what fuel pressure would give you that throughput. I don't think it's there for recommendation of fuel pressures. Maybe he should add a link to the EEC article on the results page, though. We can suggest that. You're right about the computer being able to protect from being overly lean at WOT. Code 43 covers that. This means that the O2 sensors are monitoring at WOT, just not for the purpose of adjusting a/f ratio. ------------------ 351W 89 Mustang GT Convertible