Fuel Supply Pump Failed Flow Test

[PeterPortugal]

Hi All,

I finally got round to testing my newly rebuilt early (long/tall) style fuel pump on the car and it didn't meet the specifications given in the manual. Not sure how much difference it makes but my car is a W111 with a two plunger fuel injection pump.

Firstly I recorded the pressure at the fuel filter and it was low, maybe 0.3atm when it should be 0.6 to 0.8atm. As many learned folks have said flow is possibly more important I tested this (looking for at least 1 litre in 15 secs). Initially I tried at the return to the fuel tank and I got less, maybe 600 or 700ml.

I tested the supply side and easily got the litre in under 15 secs at the inlet to the fuel filter in the engine bay. I then took the flow at the outlet of the fuel return damper (as per the manual) and was back down to about 700ml in 15 secs. The fuel filter is brand new.

Any suggestions? I have just rebuilt the pump and the components appeared to be in good condition. Voltage is OK too, above the minimum 10V the manual says.

Regards

Peter

[Shvegel]

There is a sreen in the fuel tank as well as a screen in the inlet of the pump.  Have these been checked?

[Pawel66]

I am afraid I will not help resolve Peter's issue (even though if I would have a flow after supply lune and lower flow after return line I would think the return line is clogged a bit), but for mine and others knowledge: when you measure the fuel pump pressure - I understand there is a port on fuel filter where you can screw in the sensor - but then - do you block the return line?

[PeterPortugal]

Guys,

The fuel pump is just rebuilt so the inlet screen was clean. A new filter went into the tank too.

When I did the flow test I had already removed the pressure gauge in the filter.

I am trying to work out does the pump have a problem or something else in the system?

I blew out the supply line (it was clean and unobstructed) but I have not done the return line. But on my last test I measured flow at the outlet of the fuel damper (so the return line was not being used) and it was still low.

Regards

Peter

[Benz Dr.]

Have you even tried it out yet? Although it's a bit on the low side it still may be enough to make your engine run OK.

[BaronYoungman]

Was the armature re-wound?  As the armature ages the windings degrade and the commutators lose connection causing the armature to not turn as fast hence the flow decreases.  Just a FYI also might check to make sure the brushes move back and forth smoothly so they keep a good connection with the Commutator.  Bob

[PeterPortugal]

Bob / Dan,

You both make good points. The engine does run and the car isn't in use so I can live with it for now.

Bob's suggestion makes sense because a vane pump is positive displacement which means it must be slowing down if the flow decreases. It can do a litre in 15 secs with no back pressure but when it has to pump through the system the motor just hasn't got enough in it to maintain the correct flow.

I am going to check the brushes first (which were replaced) and if they look ok the I will get the windings overhauled.

On the positive side the seal is holding up and it doesn't leak so the rebuild wasn't entirely unsuccesful !

Thanks

Peter

[Kevkeller]

If this is the same pump as on the early 113’s it is not a positive displacement pump.

[PeterPortugal]

Hi Kev,

Somebody else told me that via personal message and I am looking at it again.

It is an odd design. It certainly isn't a typical centrifugal type which does have a capacity which varies with differential pressure.

I need to sleep on it.

Regards

Peter

[Kevkeller]

You’re right it isn’t typical of the centrifugal pumps I’ve seen either. I really don’t understand the theory behind this pump.  Doesn’t seem very efficient to me.

Kevin

[PeterPortugal]

Gents,

I am eating my words here. There is a channel in the pump body underneath the vanes of the impeller. This channel is completely open between suction and discharge chambers so the pump certainly is not positive displacement because liquid cannot be trapped between the vanes.

There is no significant increase in cross sectional area (volute in a centrifugal pump) between suction and discharge (just different sized apertures to the two chambers) so very little head will be generated. As I was informed by personal message it is a high flow, low head design.

For anybody who understands centrifugal pump operation this will give a very gentle curve of differential pressure vs flow so large variations in flow can occur with small changes in differential pressure.

I have checked the brushes which seemed to be working fine and I am going to try it out again. If there is no change then it's an overhaul of the motor windings.

Does anybody know of a specialist in Europe for this?

Regards

Peter

[Kevkeller]

I bought an armature from another member thinking it was for my short pump but I misread the ad.  It is for the earlier long pump.  It looks to be in real nice shape if you want to PM me.  I’m in Paris fir another week before I get home.  I can send pictures of it when I return if you’d like.
Kevin