Upgrading fuel filtration on an IDI VW

[TurboJ]

what does AKI already do?

Aki runs 10mm lines and a twin filter setup. No catch tank but no transfer pump either.

I just had a good conversation with Aki. Like I suspected, when discussing pump inlet pressure, the VE pump cannot be compared to the VP pumps of TDI engines.
So the feed pressures that are succesfully used on VP pump engines do not apply to the VW IDI's VE pump.
Aki has extensive experience on modifying and flow testing the Bosch VE pump.
All his pumps have undergone a serious testing regime on a pump bench, and my pump is no exception.

All the IDI VE pumps Aki has tested have reacted very aggressively to a pressurized fuel feed. Big timing curve variations.
In some cases the pump advance has been retarded to zero with a pressure feed.

Aki did note that all VE pumps are not the same. OEM pumps might act differently than the large core pumps we are using.
Perhaps this is also why Giles and others have seen no problems with pressurizing VE pumps.

Where a VE pump's timing can be affected by pressure, a VP pump's cannot - within reason of course - because of the electronic timing control.

So as the flow bench test indicate that even small inlet pressures mess with the timing on the pumps that we are using, pressure feed is right out of the question.

But in any question, one should ask why you should even contemplate pressure feeding a pump if it isn't necessary?
My catch tank design is meant to optimize the original working principle of the VE pump, but adding a pressure pump would alter the principle.

[R.O.R-2.0]

what does AKI already do?

Aki runs 10mm lines and a twin filter setup. No catch tank but no transfer pump either.

I just had a good conversation with Aki. Like I suspected, when discussing pump inlet pressure, the VE pump cannot be compared to the VP pumps of TDI engines.
So the feed pressures that are succesfully used on VP pump engines do not apply to the VW IDI's VE pump.
Aki has extensive experience on modifying and flow testing the Bosch VE pump.
All his pumps have undergone a serious testing regime on a pump bench, and my pump is no exception.

All the IDI VE pumps Aki has tested have reacted very aggressively to a pressurized fuel feed. Big timing curve variations.
In some cases the pump advance has been retarded to zero with a pressure feed.

Aki did note that all VE pumps are not the same. OEM pumps might act differently than the large core pumps we are using.
Perhaps this is also why Giles and others have seen no problems with pressurizing VE pumps.

Where a VE pump's timing can be affected by pressure, a VP pump's cannot - within reason of course - because of the electronic timing control.

So as the flow bench test indicate that even small inlet pressures mess with the timing on the pumps that we are using, pressure feed is right out of the question.

But in any question, one should ask why you should even contemplate pressure feeding a pump if it isn't necessary?
My catch tank design is meant to optimize the original working principle of the VE pump, but adding a pressure pump would alter the principle.

if the pump timing was retarded to zero, then the pressure regulator is kaput..

the pump operates off of pressure DIFFERENTIAL.. the pressure at the pump inlet must simply be less than the internal pressure.. the vane pump raises the pressure so there IS a difference..

tyler and giles said there is NO DIFFERENCE IN TIMING CURVES when pressure feeding...

Aki says you can COMPLETELY NEGATE the timing advance?!

those are 2 100% opposite answers...

[Alcaid]

I've experienced VE pumps struggling with high RPM high output without feed pump (sort of misfiring on top) and being totally cured with a feed pump (18psi) and then dynoing over 170whp, not going to happen with timing retarded to 0!

That the VE and VP pumps can be totally different sounds to me like crazy talk, internally they are very much the same, but knowing it comes from AKI who has the know-how it just confuses me 

Even my pump builder (pumps proven in Bosch approved test bench) recommends a feed pump...

How many CC does your pump put out now?

[TurboJ]

It's set to 90 CC for a basic setup to get the car running. We'll start adjusting it from there.
It was tested at 125 cc on the pump bench, but about the maximum cc I don't know.

It is confusing that there are so different experiences for different people.

There has to be something interesting in the difference of the VE and VP pump.

Aki runs no pressure feed on his own IDI engines.
However on the 340-hp ALH Polo build that Aki has been closely involved in, they are running ~0.5 bar pressure feed.
Apparently the TDI's electronic control can correct the internal pressure accordingly.

Here's a direct quote from Aki, translated by me:
" On these mechanical distributor pumps you should under no circumstances increase the feed pressure coming to the pump, because it directly affects the advance,
actually in a decreasing way. Even so, that in effect, a one bar pre-pressure reduces the advance to near zero, which leads to serious problems. This problem the TDI
electric pumps do not have."

Think what you like, but Aki did not make 260 WHP with a slipping clutch on his AAZ by accident.

I am inclined to think that both sides must be right - but it would be interesting to know why and how.

I guess it has to have something to do with variations within the VE pump family.
We are using the 11 mm 'small truck' pump and they are all quite extensively modified.
Not just for max flow, but for controlled, linear flow and throttle response. And wide adjustability.
On my pump for example, even the internal feed pump has been modified to provide approximately 30% more flow than standard,
and the internal pressure and advance curve adapted accordingly.

[Alcaid]

The pumps I have experience with is based on VW pumps with parts from other pumps. If you base your pumps on something else, maybe that is the reason for the different results?

[R.O.R-2.0]

The pumps I have experience with is based on VW pumps with parts from other pumps. If you base your pumps on something else, maybe that is the reason for the different results?

still tho, there is a pressure regulator in ALL bosch VE pumps..

it regulates the internal pressure based on inlet pressure..

how do we have such different responses?

Giles says that even feeding the pumps with VERY HIGH pressure has no effect on timing advance curve..

Aki says that the timing advance can be COMPLETELY NEGATED by feed pressure..

honestly tho, if the pressure regulator is working PROPERLY, then pressure feeding the pump should have NO EFFECT on timing advance..

[theman53]

My thinking is have AKI design you deal as it is over my head and right in the realm of his.

The reason I asked you to ask him was to see what the deal was, but it seems we have differing ideas. Also, my thinking with the earlier posts were to run no catch can and a filter before the IP...let you filter be the catch can. You could probably put a pressure regulator on it and get it down to 1psi so the pump has to suck out of the filter if you are worried about the pressure. If you have ever filled a fuel filter you will find it takes quite a bit to fill it so it should work for what you want it to. If you were really concerned you could add 2 in tandem or if you have to have your tank, run it pre IP and don't dump the return into it. So it would be filter, tank, IP.

If I have time I will send this to Giles and see if he could comment, or someone else could as I am getting offline now and for the rest of the night, probably.

[TurboJ]

I thank you people for all your feedback!

Discussion can be like this sometimes, it takes its own route and the original questions may be lost on any of those sidetracks...
It's good that this discussion took place however.
I still don't know how the different pumps can act so differently.
I have no reason to doubt Giles' data, but neither I have Aki's.
Maybe there's some piece of important information that hasn't yet been mentioned.

I am 100% sure the pumps Aki has built, tested, and run, are not faulty in any way - so the pressure regulator being bust isn't really an option.
Besides, how would the regulator have exactly the same fault on every pump Aki has built?

I am going with the design pictured on the last of my two diagrams.
I'll fit the catch tank above the inlet port on the IP and run a high-flow, low pressure feed pump (considering a Carter 4600HP at the moment).

theman53, I get your idea about using the filter as a catch tank of sorts. That's basically what the OEM design is.
However, the filter could never be a true catch tank, because the line from the filter would then have to start from below the filter unit - you'd gave to fit
the filter upside down. And it would have to have an extra return line to the tank to separate all the air.

I will be fitting a rough filter to the injection pump's return line to the catch tank. Then every cycle through the IP will be filtered by the rough filter.
And if I use the 4600HP transfer pump, that will flow 100 GPH ( 380 lph ) which would mean the main filter would be getting around 130-150 percent of the
flow going through the IP.

Now if only I would find micron ratings for any of the European filter manufacturers the products of which I have readily available...

[theman53]

I get you are doing it your way. It may work for you just fine, I just don't get how or why. I haven't thought as much as you about it I guess.

My thought is with the clear line I have, I have never seen air in the lines of mine. The nature of a pusher pump should fill the filter, if you were that worried you could mount the filter on its side so that the out was always lower. As long as you were not using more than 1/2 the filter worth of fuel at any given time it would seem impossible to me for there to ever get air in the line once the filter was full.

For fun you should get AKI to put a vacuum gauge on his stuff that has nothing pushing it. Then when you are done put it on your setup to see if all your idea worked perfectly, or if there is little to no difference. I am all for learning and if there are gains to be had I would add your setup to my car. I just can't get my head around how it works.

[TurboJ]

Yes, that's a good idea...

I think I will measure my heart out of my own car in any case ( if it ever runs...)

[745 turbogreasel]

These might theoretically be better than a catch tank...
http://www.fassride.com/
I know people who use them and don't have problems.

[R.O.R-2.0]

These might theoretically be better than a catch tank...
http://www.fassride.com/
I know people who use them and don't have problems.

thats still pressure feeding the pump tho..

the OP feels that pressure feeding the pump is not the way to accomplish things..

[TurboJ]

Aki had a discussion today with Pasi of PP-diesel.
He is one of the best known professional diesel pump builders in Finland.
This guy has built several hundreds of high-performance pumps, and is considered to be one of the very best anywhere.

His comment on the subject of pressure feeding the VE pump was that you should never do it, at least never over 0.2 bar (4 psi). Go over that and it will screw with the
advance adjuster. Inline pumps are a different matter, but he didn't see any exceptions in the VE pumps that would allow running feed pressure into them.

This leaves me in even greater wonder about what Giles and the others are talking about.

----

I know of the Fass and Airdog and the likes, but nice though they are, they aren't really the optimal solution for me. And there is the pressure issue as well.
Besides, the cost of either is multiple compared to the setup that I'm building for myself.

Thank you for the suggestions anyway!

[theman53]

http://www.jegs.com/p/Holley/Holley-4-Port-Fuel-Pressure-Regulator/746390/10002/-1

what about something like this?

[TurboJ]

Thanks for the link!

I will surely look into those things if a regulator is needed.
My goal is to not need one though..

I guess I will only know when I get to test my pump and fuel system in action.

Hmm. There is one more thing that you guys could definitely help me with!

I would need some sort of a warning system in case the feed pump fails and the catch tank runs dry.

Is there a quickly-acting sensor that would sense if the liquid level drops?
A coolant overflow tank sensor comes to mind, but would that work similarly with diesel fuel as it works with water/glycol?
Depending on the size of my catch tank-to-be, it would, at full load, run dry within 30-60 seconds of the transfer pump dying.