Return fuel via my CAV filter clarification sought

P

pcatterall

Well-Known Member
Joined
2 Aug 2004
Messages
5,507
Location
Home East Lancashire boat Spain
Visit site
Sorry to return to this but I still have some uncertainty which other Perkins engine owners may share.
The existing system works fine but I would like to simplify it.

The engine filter is currently configured as follows

1)One Pipe in from front with fuel from 'lift pump'
2)One pipe out from front to the heat start.
3)One pipe out the back to the injector pump
4)One pipe in from the back which goes to the injector pump

then 5) a pipe into the top ( where the bleed screw usually fits) with return from the injectors. This is via a banjo type connection which then goes via another banjo back to the tank

1,2, and 3 are ok and I have no problems with them
4 and 5 I guess are both returns but would like clarification

5 is the one that really niggles as it is not easy to bleed and the banjo fittings cover the filter fixing bolt making it difficult to access.

I guess ( and seek clarification please) that there are two returns one ( 4 )from the 'start' of the injector pump ( fuel not required by the pump) and then the one (5)from the injectors ( surplus fuel)

To simplify the system I guess I would have to take these two returns directly back to the tank ( either separately or by joining them) ?

It is worth noting that 4) would be a swine to get at as the heat exchanger and other stuff would have to be removed but at least when blanked off there is less chance of having to get at that fitting again.

So guys.... can you clarify if my guesses on the flows are correct and ( finally) can someone explain why the original design might have routed the fuel returns in this way/ what were the considered merits?

The old maxim ( if it works don't try to fix it ) may apply of course!!View attachment 34176
 
I was pondering exactly the same issue over the weekend! My Perkins 4108 is set up exactly like yours, and despite having fitted new washers I have a very slight leak coming from the bleed bolt. I too would prefer to connect both returns to the return pipe and miss out the CAV altogether. Failing that, could I just take the return that goes to the top fitting and feed that back directly and leave the return that connects to the side fitting alone?

Neil
 
pcatterall said:
Sorry to return to this but I still have some uncertainty which other Perkins engine owners may share.
The existing system works fine but I would like to simplify it.

The engine filter is currently configured as follows

1)One Pipe in from front with fuel from 'lift pump'
2)One pipe out from front to the heat start.
3)One pipe out the back to the injector pump
4)One pipe in from the back which goes to the injector pump

then 5) a pipe into the top ( where the bleed screw usually fits) with return from the injectors. This is via a banjo type connection which then goes via another banjo back to the tank

1,2, and 3 are ok and I have no problems with them
4 and 5 I guess are both returns but would like clarification

5 is the one that really niggles as it is not easy to bleed and the banjo fittings cover the filter fixing bolt making it difficult to access.

I guess ( and seek clarification please) that there are two returns one ( 4 )from the 'start' of the injector pump ( fuel not required by the pump) and then the one (5)from the injectors ( surplus fuel)

To simplify the system I guess I would have to take these two returns directly back to the tank ( either separately or by joining them) ?

It is worth noting that 4) would be a swine to get at as the heat exchanger and other stuff would have to be removed but at least when blanked off there is less chance of having to get at that fitting again.

So guys.... can you clarify if my guesses on the flows are correct and ( finally) can someone explain why the original design might have routed the fuel returns in this way/ what were the considered merits?

The old maxim ( if it works don't try to fix it )
Click to expand...


I think 5 is there a self bleeding air release. Your pump and injector spills go back into the filter inlet and this can result in air bubbles just recirculating. This air will rise to the top of the CAV and cause some of the problems you see here. Your set up means that continuous bleeding is taking place. I suspect if you inspect the banjo bolt it is not a normal one but has a smaller than normal port in it to maintain the lift pump pressure...
 
So why not take the spills directly back to the diesel tank, where the air will disapate? Perhaps the designers thought it was easier to use the CAV as a joint, rather than having a separate T piece joining the two spill pipes to the common return to the tank? Would welcome any thoughts!

Neil
 
Bilgediver said:
I think 5 is there a self bleeding air release. Your pump and injector spills go back into the filter inlet and this can result in air bubbles just recirculating. This air will rise to the top of the CAV and cause some of the problems you see here. Your set up means that continuous bleeding is taking place. I suspect if you inspect the banjo bolt it is not a normal one but has a smaller than normal port in it to maintain the lift pump pressure...
Click to expand...

Thanks bilge diver. Sorry, did you mean that 5 is there as a self bleeding air release? Just to clarify, the pump excess is returned to the filter inlet and the injector excess is returned to 5 and also from 5 there is a pipe to the tank.
I would also ask the same question as Neil, why would they design it like that? there must be a reason ( mustn't there ??)
 
connected that way because the engine was sold as a self contained unit ie. only requires fuel supply and electrics to fit, cutting down cost for the installer of static plant/
 
I can see what Pampas means - with the CAV you have a ready made connection point for the fuel return that is already there by virtue of being a CAV. Plus it is readily identifable - all you need is a hose with a suitable connector and you're off. If you did not use the CAV you would have either to fit two returns, or fit a separate T piece, which would then have to be terminated in some way so that the installer could readily connect to it. Simpler all round for the installer to use the CAV, but possibly not the best for the maintainer, as we are finding - or am I clutching at straws?

Neil
 
neil1967 said:
So why not take the spills directly back to the diesel tank, where the air will disapate? Perhaps the designers thought it was easier to use the CAV as a joint, rather than having a separate T piece joining the two spill pipes to the common return to the tank? Would welcome any thoughts!

Neil
Click to expand...

I have memories of conversations in the distant past regarding problems getting some CAV rotary fuel pumps primed. Maybe this is
an attempt to get rid of all air asap I think the connection between the filter and the banjo on top will probably have a restrictor oriface to
help maintain lift pump pressure to the fuel pump and this must not be replaced by a standard banjo bolt.
 
Under no 5 there should be a non return valve (a ball in a socket), these sometimes jam and stop exess air / fuel returning to the tank.
I would check this is ok and leave it as it is (it has worked like that for 40 years).
If you alter the system you will find all sort of odd problems with pressue in the wrong places.
 
neil1967 said:
I can see what Pampas means - with the CAV you have a ready made connection point for the fuel return that is already there by virtue of being a CAV. Plus it is readily identifable - all you need is a hose with a suitable connector and you're off. If you did not use the CAV you would have either to fit two returns, or fit a separate T piece, which would then have to be terminated in some way so that the installer could readily connect to it. Simpler all round for the installer to use the CAV, but possibly not the best for the maintainer, as we are finding - or am I clutching at straws?

Neil
Click to expand...

I must be loosing the plot!! If you just feed the return to the CAV then ( surely??) you still have to get it back to the tank You cant just keep feeding returned and newly pumped fuel into the CAV (??) so why feed it into the CAV when you have then to get it out again.
Just as an aside ......I see in the manual that all the fuel flows in and out of the CAV are shown as a solid line with direction arrows. Except the return from the injector pump which is shown as a dotted line with no direction. Is this significant?

PS sorry I didn't spot the 2 last responses, some sort of sensible explanation seems to be emerging. The design was a bit clever in that it uses return fuel flows to carry any air in the CAV away using the pipe mounted at the bleed point. I will check 5 for this ball bearing thingy.
I wonder........................ I have had a long standing problem with air ingress after long periods on idle..............something like a sticking valve would be a neat explanation.......... sticks when pressure is low but ok on normal revs.... hmmm?

Thanks all
 
Last edited:
Sorry, I wasn't suggesting that you didn't need a return, clearly you do. What I was suggesting was that rather than using a T piece betwwen the pump and injector returns and the return to the tank, the designers have eliminated the T piece by using the CAV, which provides a convenient point for the installer to connect the return pipe. Given the other post, I suspect the designers may have been cleverer than that anyway!

Now to try and solve my minor leak - blue hylomar I think!

Neil
 
Hylomar can be dissolved by fuel, I would use PTFE tape or silicone RTF if you have to use a sealant, try new copper washers first.
The one way valve is in the adaptor that the injector return banjo fits onto.
 
macnorton said:
Hylomar can be dissolved by fuel, I would use PTFE tape or silicone RTF if you have to use a sealant, try new copper washers first.
The one way valve is in the adaptor that the injector return banjo fits onto.
Click to expand...

This has been quite interesting and a bit of an eye opener ( the one way valve aspect) I guess that the fuel return from the injectors ( which seems to above the one way valve) will just flow direct to the tank and that is the fuel not required by the injector pump which passes through the CAV ( a second filtration?) through the one way valve ( carrying any stray air bubbles?) , joins the return from the injectors and flows back to the tank.

So we old CAV types have actually got a bit of sophistication!!

Up until this debate I have always been urged to 'scrap the return to the CAV and take it straight back to the tank'. I think this sort of well meant advice often comes from forumites whose own system of 'direct to tank' works well and assume that it would work well in all cases.

I hope others with the same set up will be able to find this thread.

Thanks again
 
pcatterall said:
This has been quite interesting and a bit of an eye opener ( the one way valve aspect) I guess that the fuel return from the injectors ( which seems to above the one way valve) will just flow direct to the tank and that is the fuel not required by the injector pump which passes through the CAV ( a second filtration?) through the one way valve ( carrying any stray air bubbles?) , joins the return from the injectors and flows back to the tank.

So we old CAV types have actually got a bit of sophistication!!

Up until this debate I have always been urged to 'scrap the return to the CAV and take it straight back to the tank'. I think this sort of well meant advice often comes from forumites whose own system of 'direct to tank' works well and assume that it would work well in all cases.

I hope others with the same set up will be able to find this thread.

Thanks again
Click to expand...

Basically that's the way it works, although the fuel returned from the pump is not filtered a second time it uses the filter head to restrict flow and stop air getting into the pump via the return line when not running.
 
You must log in or register to reply here.

Share this page

Top