Fuel dampener

[raygreenwood]

Ray, my pressure fluctuation happens with the engine off (pump on to test) so I don't think it is influenced by injector flow. It is more like the regulator opens/closes rapidly resulting in the pressure never settling.

Ah...yes.....that will happen with virtually any....if not all.....regulators. What you are seeing is the phenomena that any/every spring has a specific "cracking" or unseating pressure required. I order springs for dies and injection mold work by the unseating pressure as its critical.

But.....once unseated and being held suspended off of its "seated" position.....the spring should then regulate or stabilize at a linear rate.

If you are using a regulator from another vehicle designed to run at either a much higher fuel volume or a much higher pressure range....the spring that the engineers installed my have an unseating pressure that is dangerously high.
If yours is only fluctuating +/- 2 psi when clicking on the pump....and only happens once when the pump is energized but the engine is off....its not an issue.

If with the engine off and the pump energized...it keeps fluctuating/cycling....you have either too large of a spring in the regulator.....or more likely....you are having a fuel volume issue....meaning difficulty maintaining steady fuel volume to maintain stable pressure from the pump.

If as you noted....you have gone through several different...known quality pumps and still have this problem......I would surmise that with the engine not running. ....you have low voltage to the pump causing it to output lower volume.

In other cases if you are using fuel pressure with that range of pumps that is over about 30-35 psi....I would suspedt a weak or missing check valve on the outlet side of the pump. Ray

[andy198712]

Thats a dead head system right? most modern cars and bikes go that way now (gsxr's for one and thats the ti's i'll be using) i don't see why we don't go that way as much

Ahh just re read and i see it still uses a reg at the rail too....

If you were asking that of me......no.....the WC vws are not a dead head system. Not necessary. It has a return line. You can acomplish the same level of stability and pressure change dexterity with a Standard ring/loop with retturn to the tank......simply by having far more volume at a higher pressure.....delivered smoothly ......than the engine requires.

The dead head system while nifty.....requires a much more expensive pump....logic to control it....and suffers catastrophically if any part is less than perfect.
Ray

Got it!
i noticed sports bikes seem to dead head, and there fuel pumps are cheap but intank...

[Steve Arndt]

I don't have a check valve in the system. Should I add one? I thought that the check valve was only used to hold and maintain pressure after shut down (for quicker startups, and to keep the fuel in liquid state and not vapor in the hot fuel rails).

Also -----
This is the canister I was talking about using to make a fuel volume vessel. We replace about 1 a week on late model Benz cars at our shop.

[raygreenwood]

YES!...you need a check valve.

What the check valve does..especially on two port roller cell pumps as compared to the three port D-jet pump.....is....now think about this and wrap your head around it........its just like an air compressor. You have to have an inlet and outlet flap valve.

As each roller cell in the pump makes its way through the compression cycle to the outlet port...it forces its charge of compressed fuel through the port. If there is no check valve....the back pressure causes equilibrium against the next charge of fuel coming out of the port. It also causes stalling of fuel flow across the roller cells.

The higher the pump outlet pressure...the more pronounced this effect will be....which is why ALL CIS pumps like the Bosch 044...running high pressure MUST have a check valve or the fuel pressure will be highly erratic.

To easily demonstrate this....put an 044 pump on a CIS engine at idle with no check valve. At about 75 psi.....you will see the needle bounce from between 70 and 80+ psi at a high rate of speed.

What is happening is that as fuel charge exits the roller cell.... into the face of line pressure only marginally lower than its own......output volume stalls. Since the engine is also using fuel...you get a dip in fuel pressure. But, line pressure is still relatively high. The pump rotor must actually turn several rpm dumping several roller cells worth of fuel at the port to gain enough volume to overcome the pressure head. You then see a needle spike to high pressure for a split second.

Simply put....the pressure head creates "push back" on the roller cells without a check valve.

The D-jet pump fed from a reservoir of semi pressurized fuel at the return port. Also connected to this return port is excess fuel from the roller cells from "push back" from line pressure...because D-jet had no real check valve either......but D-jet also was low pressure...max 28-32 (reliably). Start running higher pressures with a D-jet pump and it gets erratic as well.

You dont need much of a check valve spring pressure. Just enough to keep main line back flow from impeding pump egress. Ray

[MinamiKotaro]

Where in the system would be best to plumb the check valve?

[Piledriver]

As close to the pump as possible.
The check valve element needs to be low mass so it can cycle at pump output-pulse speed, at least on a roller cell.

The factory BB sized sprung ball check valve in the output banjo bolt works just peachy, and I strongly suspect at least the CIS roller cell pumps pumps are flow rated with it installed (flow rated @~90 PSI/7 BAR too)

Recommended at ANY pressure.
Maintaining back pressure after shutdown is also a feature, but not the main one.

Gerotors like the walbros do not require it, but it also hurts ~nothing and will help hot starts.
You should probably use a less restrictive swagelok unit in that case, as those pumps are usually flow rated w/no backpressure at all.
(The E85 Godzilla Walbro pump may be an exception)

I suspect the lack of an output check valve was ~half of what was wrong with the 044 I got from ToBy.
(The other half was also no feeder/sump on his install, roller cells do not suck, but in a bad way)

It works fine used as Bosch intended, with a feeder/sump and check valve.

It would probably work OK as an in-tank, or perhaps sucking on the bottom of the tank via a ~13mm ID hose and fitting ala early CIS mk1s, but vw moved to the feeder and sump for good reason, it even cost them more, and you KNOW it isn't easy to overide the bean counters.

[raygreenwood]

X2 on what Piledriver said. On many of the factory pumps....the check valve is built into the fitting on the pump outlet. Many people dont even know its a check valve unless you disect the pump.

A simple ball and spring check valve works. There are lots of versions buipt to work with fuel systems. Google keywords Bosch fuel pump check valve and you will find pages of them with pictures. Prices range from $15 for a CIS pump check valve for an 044....to about $40 for those listed for BMW and Volvo ....we know how that one works.....Ray