Really strange idea of the day!!!

[jim_in_dorris]

I was talking with my son tonight, and according to him, all the new toyota's, and dodges pulse width modulate the fuel pump itself to control fuel pressure. If a 747 class ecm had a spare PWM output, is it possible to write a routine to control fuel pressure based on load and rpm? No more adjustable fuel pressure regulators. Just putting this out as a discussion item. What does everyone think. My 355 build will require #80 injectors at 20 psi or #61 injectors at 34 psi. How nice it would be if you just told the pump how much pressure to supply (like a vacuum referenced FPR except much more accurately).

[gregs78cam]

A few years back when I was working on a friends 300ZX (the one in my avatar) we learned that the fuel pump in that was controlled by a module. We ended up bypassing module control so it would put out constant pressure. I don't know if it is similar to what you thinking, but it sounds possible.

Just did a quick search, and it seems like the module runs the pump at 2 speeds depending on load. So yea it would be something to look into. And you meant '7427 right?

You would also probably have to give the PCM feedback on fuel pressure. And you would probably have to come up with another way to bypass the excess flow(like an inline restriction on return), or just dead-head the TBI.

[1project2many]

Your son's correct. The strategy today is to use a returnless fuel system which reduces vapor formation in the fuel tank with pump pressure controlled electronically. Although you could implement a non-feedback based pump control strategy you leave yourself open to problems if actual pump output doesn't match the model used for development. Better approach uses a fuel pressure sensor. Once you have fuel pressure as a variable controlled by ecm you have a lot more dynamic range for a set of injectors so larger cams and boost will become easier to tune from idle to full power.

[EagleMark]

Don't think there is extra PWM circuit on 1227747 which I believe Jim runs. But if he did a 16197427 conversion that does have a PWM circuit for the E trans which could be first part of the puzzle?

[RobertISaar]

keep in mind you wouldn't be able to directly run the pump off of the PCM, way too much current draw. you would need to setup a solid-state relay or some transistors to handle all of the load.

[jim_in_dorris]

Okay, after reading the replies, and thinking about it, I came up with the following;

Robert-- I wonder if the fuel pump relay can handle PWM

I was thinking about setting the fuel pressure regulator at a maximum pressure, then if something goes wrong, it won't go over that pressure. If I was to use say an EP377 pump which IIRC is rated at about 45 psi, I would set the regulator at 35 PSI, then build a table to drive the fuel pump. I would most likely use multiple settings like:

idle -2000 rpm 33%
2001 - 3000 rpm 66%
over 3000 100%

Then include 3 different values for flow rates depending on my injectors.

Having a closed loop feedback system would be the best route, but I am trying for a minimalist approach here at first. If that doesn't work, increase the complexity a little at a time. Also, if the FPR looks stock, it's one less thing for the smog nazis to object too.

[EagleMark]

Robert-- I wonder if the fuel pump relay can handle PWM

From what I read about this in the other $OD fan option no! On this page would be the relay you would need.

http://www.gearhead-efi.com/Fuel-Inj...-control/page3

[RobertISaar]

if you try and drive a relay via PWM, expect bad results, to say the least. those things are meant only for discrete operation, hence the suggestion of a solid-state relay or some transistors., since those don't have any actual moving parts.

[jim_in_dorris]

Just speculating at this point, I really didn't expect any other answer about the relay than what Mark and Robert replied with. So Step 1 would be to develop a circuit to run the pump. That is fairly straightforward. I guess the next step would be to investigate driving the circuit with some software. This is where I really need a test bench. I guess first I will start playing with software. Stay tuned for the next installment of "HMMM will this work?"

[RobertISaar]

i'm just about to finish my 100% solid-state bench for my 16149396.

and yeah, the circuit is the easy part, getting an open loop fuel pump pressure system to work consistently is going to be the interesting part. do you happen to know if there are any spare A/D channels on the PCM you're planning on doing this with(so you can run it closed loop with a pressure sensor)? i got lucky that Ludis documented the 7727/7730/7749 type PCMs long before i knew anything about them and i actually interally modded my 9396 to add 3 extra A/D channels, though there are more that could be opened up if necessary.

[jim_in_dorris]

<quote> i'm just about to finish my 100% solid-state bench for my 16149396.<quote>

Nice. I would love that myself. I understand the problems with open loop, and will probably end up closed loop. However, I do intend to investigate open loop first before abandoning that attempt. I could probably put the pressure sensor somewhere out of sight so it would be unobtrusive.

[jim_in_dorris]

Okay, perusing 98's fan control code gives me the output for the fuel pump, now the question is: How many amps is an EP377 fuel pump? I could use the fan relays for the fuel pump.

Next question: Would an oil pressure sensor work for closing the loop on the fuel system? The one in my truck is a 1 wire, not sure if that would work or not. The range would probably work, just have to figure out what the calibration looks like. I.E. what voltage is what pressure. I think it might not work because of sensor grounding, but I could probably ground the sensor in the T adaptor.

[RobertISaar]

i actually have a similar situation with the 1 wire oil pressure sensor... what i did was grab a spare instrument cluster, powered it up on a bench, connected the oil pressure circuit to ground through various resistances(it's a 0-~80 ohm sensor) and then used my multimeter to check the voltage between the resistor and the cluster. my results? at most, i should see around 8.2 volts or so when the sensor is pegged at 80PSI. so if a run a /2 voltage divider, i can put it into the 5 volt range of the A/D circuit in the PCM, then just tap a wire into the oil pressure sensor circuit(using a high value resistor, to prevent it from skewing the gauge).

anyways, as long as you have 1 open A/D channel that is internally connected to ground via a resistor, you can do this.

[jim_in_dorris]

Okay, now I need to find an open A/D channel. One thing nobody has mentioned is the Oil Pressure Switch that is connected electrically to the fuel pump. I don't think that it would work well if it was wired in parallel to a PWM circuit.

[EagleMark]

You can remove it. It provides power when running if relay fails. If relay fails you will have long crank starts until oil pressure switch powers fuel pump IIRC. They are not used in aftermarket wiring harness and I don't use them when rewiring a harness.