PDA

View Full Version : Electric fuel pump mods and issues


veganxxx
11-06-2017, 09:29 PM
So this is going to be useless for most people, but at least interesting to some-

To run vegetable oil in the 04-06 models you need to have an electric fuel pump on the oil fuel system to provide the 60 or so PSI to the fuel rail. Traditionally, you mount the second pump on the frame before the heated fuel filter housing. Both fuel lines meet up at a valve that selects which fuel system you are drawing from, and the pumps are switched by a relay when the valves are switched.

After going through lots of raptor pumps (high performance diesel pumps) which are both expensive and noisy I decided to try the bosch replacement for the powerstroke fords (they run their unit injector rails at a similar PSI to the sprinter common rail). This pump work fine but during the switch from one pump to the other pump it took a second to build pressure. This second lag made the computer very unhappy and would throw a low fuel rail pressure code and often stall the motor.

I decided that I would try moving the pump under the engine bay and have it draw both fuels and be on all the time. I cut power to the electric pump in the stock diesel tank tank and just ran direct power from the stock fuel pump relay to the new pump.

That worked great except.... it would throw high fuel rail pressure, fuel quantity exceeded codes while running on diesel. This would primarily happen when flooring while driving. I assumed that there were spikes in the fuel pressure and so the solution was a fuel pressure regulator.

I tried a regulator and dialed too low for full load it will throw both over pressure and underpressure codes. Dialed higher it would still cause too much pressure. I have a T in the return line that the regulator bleeds into.

Then I tried using a blower motor resistor to slow the RPM of the motor. But that also still caused the over pressure codes. If you use a big enough resistance the pump is too slow to even start the motor.

My question is, what's the difference between having the pump mounted further back vs. under the hood in regards to fuel pressure at the rail? I know there's some pressure drop from going through a filter, but I should have been able to replicate that with the pressure regulator.

Also why only flooring it up hill? It seems like demand would be the most at that point so too much pressure would more likely happen at idle, when there is less fuel demand, or coasting.

02-03 are a vacuum system, sucking through the fuel filter and pushing direct in the rail so it seems like it could work in theory.

If I can get this to work it would be a really great setup and useful if someone has an in tank pump die in an area where it's hard to get sprinter parts but easy to get ford parts.

Aqua Puttana
11-06-2017, 10:35 PM
...

That worked great except.... it would throw high fuel rail pressure, fuel quantity exceeded codes while running on diesel. This would primarily happen when flooring while driving. I assumed that there were spikes in the fuel pressure and so the solution was a fuel pressure regulator.

...
There is no low pressure pump [the OEM is in tank] monitor/sensor on the OM647 engine. Any fuel pressure codes will relate to the high pressure side.

It is possible for a low pressure condition to result in a high pressure DTC. When the pressure is noticed as going low the ECM attempts to correct which can overshoot and set the high DTC.

There is also a transition area for the fuel control which Dennis mentioned just recently. :thumbup:

The basis of the system works in two stages.
In lower revolutions the fuel system works under control of the qty control valve mounted on the HP pump.
At higher revolutions and higher system pressures the rail & injection system comes under rail pressure control.
Hence a scan tool with live data is the only real way to diagnose the issue .
Sometimes if you rev the engine to about 4000 r/m for a monetary pulse and let the throttle go it will often cut out and as the rail pressure drops out & across the control phases. Its one way to get an an idea of what going on.
Also an ohm check of the injectors doesn't go amiss sometimes.
All the best
Dennis

I'm not saying that the transition stage is affecting your acceleration problem. I just thought that it might be interesting information for you.

As an aside.
It's no surprise to me that some of the system pressures are similar in the Bosch systems regardless of vehicle manufacture. Bosch must standardize somewhat on components.

In your exploring and experiments have you ever measured the return line pressures? I'm thinking the OM647 and OM642 return pressures would be in similar range. Eg. - The injector return fittings are the same.


Test the pressure between the return collection pipe and fuel tank.

* First-generation Common Rails [solenoid injectors], the fuel return pressure is the decisive factor in opening the safety valve in the high-pressure pump should deliver between 0.6 and 0.9 bar (8.7 13.0 psi).

* Third-generation Common Rails with piezo injectors deliver a return pressure of 10 bar (145 psi).

The 13 psi would be at the high pressure pump return outlet. [My thought is that outlet would mostly come into play during the lower revolution stage that Dennis mentioned.] That outlet dumps into a larger pipe which is open ended. The return line pressures to the tank most likely never reach anything near that pressure.

vic

veganxxx
11-06-2017, 10:58 PM
From that analysis on why the computer would throw a high code, maybe my problem isn't enough pressure/quantity when really asking for it.-

Floor it, demands more fuel, not enough is coming from the pump, the computer tries to compensate, created too much pressure, throws the code.

This seems to make more sense. I also wonder if this may be a result of having to pull fuel through the stock lift pump while it's off. Before I switched things around I pulled fuel with a mighty van out of the stock diesel filter and it seems to flow without much resistance. But there's probably some resistance to pull through that motor.

Maybe a test would to temporarily hook up power to both pumps in tandem and see what happens.

veganxxx
11-07-2017, 12:02 AM
Tandem pump test-

on diesel in tank pump on and pump under the hood on-

No codes and lots of power. Maybe felt the fastest a sprinter has felt floored up hill from a dead stop.

So this proves that the second pump when run by itself is not causing codes because of too much pressure/flow but probably because of too little.

The next test would be to see if with just the one pump it runs on vegetable oil without problems. If so, I would guess the restriction of pulling through the in tank pump while that pump was off was causing the codes.

The next question is- Can you run two electric pumps in tandem without hurting either one (most likely hurting the second one that is being pressurized)? If you can then problem solved. Just have both on for diesel and cut the in tank one with a relay when switching to vegetable oil.

Aqua Puttana
11-07-2017, 12:20 AM
To my knowledge, centrifugal pumps in tandem cannot affect each other to the point of doing damage as long as the pump/pumps are never starved for inlet/supply liquid. Centrifugal pumps are properly throttled on the output, never the inlet or they can be starved.

Positive displacement pumps react differently, but that is not what you need to worry about. (Except when messing with the high side pump.)

vic

veganxxx
11-07-2017, 05:01 PM
Ok full test, everything works. In retrospect I should have known there was some restriction pulling through a non running fuel pump. This is because the pump did sound like it was struggling and became quiet once both pumps were running. If only it had said “low fuel pressure” instead of high, I would have been more likely to suspect that was the problem. At least we learned that high can mean low, just like bad really means good sometimes.

Ciprian
11-07-2017, 05:34 PM
Just a translation issue from german. It does not mean that the pressure is too high. It means that the computer asks for high pressure and the high pressure pump gives all its got and for whatever reason cannot supply the pressure that the computer asks for.

Sent from my LG-H918 using Tapatalk

veganxxx
11-07-2017, 10:41 PM
I did actually encounter a high pressure code where the pressure in the rail was in fact too high because of a looped return creating excess pressure instead of being able to bleed to the fuel tank as is documented here-

https://sprinter-source.com/forum/showthread.php?t=10515&page=2