TDI in a Toyota pickup

Whitedog- I've never put a gauge on a ve pump, but it may be a good troubleshooting tool.

I was thinking fuel starvation as engine seems to pull well, then gradually lose power. Case pressure valve and air entrainment would probably prevent it pulling well right away.

I usually pressurize the IC plumbing from the turbo outlet with my air compressor. You can usually hear the leaks right away. I spray all unions, etc with a spray bottle of water w/alittle soap and look for bubbles. That must take forever with a mityvac.

Greg's got a good point on the pressure relief valve. On my current TDI, that I found this same problem when I resealed the pump.

It's not usually drastic, but the symptoms of cpr valve failing/falling apart means no pressure to push the advance piston in the pump so the pump can't advance correctly at high rpm - it doesn't affect idle/low speed so you can't tell right away it's a problem, also it's progressive so unlike running out of fuel/air where you get lurching, this just results in poor power.

I have a inlet pressure gauge on a tdi right now for a similar issue, it came to me running very pooor at higher rpm, it was -20mmhg on the pump inlet. After re-routing the non-standard filter arrangement and fitting a lift pump outside a 'looped' fuel return under the bonnet, it now has 2-4psi positive pressure on the pump inlet at WOT. I'm pretty happy it's getting fuel now!

Greg.

greg123 said:

Use a facet 40106 or similar lift pump (you can T the return line back into the inlet then, reducing flow from the tank drastically) to keep a good supply. Have you taken out and pressed back together (presuming it's falling or fallen apart, like 90% are) the injection pump case pressure relief valve? This causes timing/power to be off at high rpm.

Greg.

Click to expand...

Greg, I have talked with the VP37 rebuild specialist here in Oregon about that valve. Kyle is Da Man on these things and I learned alot about what is going on. HERE are the results of my conversation with him.

That's very interesting. I don't want to drag this o/t too much, so suffice to say it won't hurt if he check it.

I do have a few q's re the assumptions on the thread you linked to, for example if the plunger stays in the bore, then why the loss of pressure at low rpm just because the circlip is out a bit?

Has Kyle actually measured case pressure on a pump with a disintergrated relief valve to test the theory? The theory elsewhere and that I have (so far) gone on is that when the circlip goes down enough, the top of the piston goes lower than the outlet hole, as the piston moves down further it's no longer in a tight bore, diesel can flow round it and straight out the uncovered port thus leading to lack of pressure, as you say the advance piston is spring loaded retarded so low rpm function is okay but max advance/high rpm function is not.

HOWEVER, I'll argue this out if I get a chance to run a vehicle back to back with the case pressure valve dismantled/normal and monitor timing and N108 duty cycle. Case pressure would be good too, only I don't know where I'd tap into that!

Anyhow I'll post this on the other thread more specifically.

Greg.

whitedog said:

Greg, I have talked with the VP37 rebuild specialist here in Oregon about that valve. Kyle is Da Man on these things and I learned alot about what is going on. HERE are the results of my conversation with him.

Click to expand...

debugging 1

First: I would like to thank you all for your input into this problem. I honestly can't express how much help you have been. I love this place.

Regarding pressure: I think I miscommunicated in the pressure drop statement.

I have a 1/8" pressure line plumbed off the intake manifold. It t's and one part goes to the boost gauge, the other to the MAP sensor in the ECU.

To do the pressure test, I removed the line from the intake manifold and pressurized it with a small pump (my mityvac). The 15psi, 10psi, and 5psi numbers were read off the boost gauge, the corresponding numbers were read off my computer. The 15sec and 10sec drop times refer to how fast this system leaks. It could partially be my mityvac leaking or line leaks.

I am going to test for intake system leaks, per Spulen's suggestion.

It seems it would be good to figure out if the lack of boost were a leak problem or a "requested" problem. IE, it seems if there were a fueling problem that caused the ECU to request less boost, this should show up in VAG-COM logs as requested vs actual boost.

If the problem is fueling, there would be a low requested boost and a close actual boost. If the problem were system leaks there might be a high requested boost and a low actual boost.

Wouldn't the fault 00575 "manifold pressure" be a result of a large difference between the actual and requested boost? Or is there some other means for the ECU to compare pressures?

Ideas?

debugging 2

Thanks for all the suggestions on fueling.

Maybe we can start with the simple stuff for this noob

Regarding general fueling issues: I may have fueling problems because of large pressure drops in the system. For example, if the fuel line is too small or clogged or the fuel filter is clogged. (to start)

-Or- perhaps the fuel line length/tank elevation is enough different from a stock Passat to cause problems.

Here is exactly what I did regarding fueling up to the IP:
Stock toyota tank, taken apart and thoroughly cleaned. Fuel pump removed and in its place a viton tube extending to near the tank bottom.

Stock toyota fuel lines, except for Viton flex sections.

Fuel filter: New spin-on diesel filter (see post 162)

Do any of these pieces sound like a problem?

Spulen: did you use the stock fuel lines or increase their size?

The clear lines in between the filter and the IP seem to indicate reasonable flow. When I was first starting the engine and there were bubbles, they moved through quickly. Now there is just a small one that seems to stay in the line.

I am currently reading Whitedog's post about the injection pump.
http://forums.tdiclub.com/showthread.php?t=238872
I may have an intelligent question later...

Two things that bother me about mine: Injector balance and the start of injection. Those #'s seem to be way out of spec.

Injector Balance:
@ idle 1: -0.16 2: -0.21 3:-0.05 4: 0.42
@1900RPM 1: -0.14 2: -0.31 3: 0.05 4: 0.42

Click to expand...

Nothing wrong there.

I'm working on getting the tooling to do some testing on the pump case pressures. I should have something tomorrow and I'll let you know.

e*clipse said:

Thanks for all the suggestions on fueling.

Maybe we can start with the simple stuff for this noob

Regarding general fueling issues: I may have fueling problems because of large pressure drops in the system. For example, if the fuel line is too small or clogged or the fuel filter is clogged. (to start)

-Or- perhaps the fuel line length/tank elevation is enough different from a stock Passat to cause problems.

Here is exactly what I did regarding fueling up to the IP:
Stock toyota tank, taken apart and thoroughly cleaned. Fuel pump removed and in its place a viton tube extending to near the tank bottom.

Stock toyota fuel lines, except for Viton flex sections.

Fuel filter: New spin-on diesel filter (see post 162)

Do any of these pieces sound like a problem?

Spulen: did you use the stock fuel lines or increase their size?

The clear lines in between the filter and the IP seem to indicate reasonable flow. When I was first starting the engine and there were bubbles, they moved through quickly. Now there is just a small one that seems to stay in the line.

I am currently reading Whitedog's post about the injection pump.
http://forums.tdiclub.com/showthread.php?t=238872
I may have an intelligent question later...

Two things that bother me about mine: Injector balance and the start of injection. Those #'s seem to be way out of spec.

Click to expand...

I'm concerned about 2 things. 1, I think you should measure pressure at the IP inlet rather than guess, preferably put in a lift pump to get a couple psi or more positive so you can quit worrying about it (rear mounted is best, self priming then). Seconds, that pipe to the bottom of the fuel tank - when you brake/go up a hill, what stops it dragging a load of air in as the contents of the tank slosh out the way?

I once had a Rover SDi come to me with that issue, behaved well other than if less than 3/4 tank of fuel and you booted it up a hill. It's fine mesh had ripped which keeps surface tension and air out. VW's use a clever 'catch tank' which keeps a reserve of fuel. Not saying this is your problem, just something which bothered me a touch.

Greg.

Quote:
Injector Balance:
@ idle 1: -0.16 2: -0.21 3:-0.05 4: 0.42
@1900RPM 1: -0.14 2: -0.31 3: 0.05 4: 0.42

Nothing wrong there.

Click to expand...

Just trying to wrap my head around this: If these numbers are ok (for example, a difference of -0.21 to 0.42, what should a person be concerned about?

Regarding the IP: If I read your post correctly, the relief valve can effect the low speed idle (mine wanders and is too low) Also, my timing seems to be too advanced 6.6 degrees BTDC. Could this also be a symptom of a broken relief valve?

Greg,

I was concerned about my inlet system. It feel a bit inadequate. Basically the fuel line is connected to the fuel pump support (so it doesn't flop around) and ends approximately 1 cm above the bottom of the tank. This is in the middle of the tank.

Do you have some link for VW's system?

Thanks,

E*clipse

Greg, I'm picturing something like a Coffee can with a bunch of small holes around it and the pick-up tube in the center of that. The holes would allow fuel in, but as the fuel sloshes around, the fuel would be slow to leave the can, always providing a supply of fuel at the pick-up. Is this what you had in mind?

Rick, if those balance numbers get beyond +/- 2, you have a problem. You are well under that so everything is fine. Take those positive numbers and negative numbers and add them together. You will get close to Zero. This is what gets you to the smooth idle. If any of those numbers get beyond +/- 2, that means that the ECU is working extra hard to keep a smooth idle.

Yes the toyota tank has baffles in it to prevent pulling air instead of fuel. Unless the tank is really low on fuel, it shouldn't suck air even at higher g's. If the pickup tube in the tank is 1 cm approx, above the bottom of the tank then that should be okay.

You can test this quite easily using your vaccum pump as a gauge by T'ing in just before your filter on the firewall. Your vaccum reading will only be accurate when your pulling that steep hill in fourth at wot. Then install the gauge between the filter and the injection pump and repeat.

I'm wondering if your 5 micron filter might be too restrictive to be on the suction side of a fuel system. If you find that your getting low vaccum on the filter inlet side under load, then everything is fine in the tank. If your getting high vaccum on the filter outlet side, your filter is too restrictive. If that's the case, I see two options. Bosch does make a fuel filter that was used on 85 model year vw diesel's with only an inlet and an oulet(I can get you the part # if you need it). Or put a lift pump in before your current filter.

I vote for the electric lift pump. Many people are running a 2 micron filter from Cat on the suction side of their VWs so 5 won't be a problem.

Ok, the consensus is to go w/ a lift pump. I was trying to avoid the extra complexity, but I can see good reasons for one.

Is there a recommendation for pump pressure/flow rate?

Will a fuel pressure regulator be necessary if a low enough pressure is selected?

Hi, Your numbers at ilde look good, the maf is correct for egr0. Timing is a bit high but only at idle and tip it. It will always follow the mapping thruout the rpm range regardless where you set it in vagcom at idle. I've seen fuel lines with pin holes in the tube that runs along the top of the tank inside. Try a fuel can with a electric pump. Make sure you run the return back into the can. You only need a gallon, and take a blast. Call me if you like to discuss this more.
Jeff

e*clipse said:

Now, some more tuning questions:

I looked into the OBD scanner thread and compared some numbers to the A4 measuring groups stickey.

Here are my numbers, w/ the truck warmed up:

engine idle speed: 840 RPM
start of injection: 6.6 degrees BTDC
Injection quantity: 4.4 mg/stroke
Engine coolent temp: 64.8C > 84C (warmed up during test)
Fuel Temp: 30.6C
Air Temp: 16.2C
N75 duty cycle: 24.7% > 33.9% @ 3000 rpm
Cold Start Injector duty cycle: 66.9%
Modulating displacement sensor voltage: 1.620V
MAF spec: 850mg/H
MAF act: 475mg/H
Injector Balance:
@ idle 1: -0.16 2: -0.21 3:-0.05 4: 0.42
@1900RPM 1: -0.14 2: -0.31 3: 0.05 4: 0.42

My main concerns are:
idle speed (too slow)
MAF numbers (seems much larger than it should be)
timing (seems way early)
injector balance (way different from each other)

I also got this error code:
fault code 00575 (intake manifold pressure) >> can result in low power

It can be the result of a faulty N75 valve or a leak in the plumbing.

I tested the boost lines between the manifold and the boost gauge and ECU by removing the vacuum line from my manifold and using the pressure side of my mityVac to pressurize the system.

Boost gauge Vag-Com
15psi 2000 mbar
10psi 1700 mbar
5psi 1375 mbar

It takes about 5 seconds to drop from 15psi > 10psi
and about 10 seconds to drop from 10psi > 5psi


Ideas??

Click to expand...

Would the original Toyota gas pump be suitable?

Karmann-diesel posted this pic which is excellent. You can't see all the workings, but basically it has a float at the bottom which isolates it from the tank when diesel is not present (sloshed away) and the diesel in this 'mini tank' stays. It's a fuel trap that inlet and return run into, a mini fuel tank filled by the big fuel tank.



Fully agree re the balance numbers being okay.

I don't know what the toyota tank is like, providing you are happy running air in during a slosh/hill isn't going to happen, fine. You can always run a loop of clear pipe out the bonnet by the window, have done this for fault finding before, to check while driving for bubbles - do it just before the pump inlet that way you cover the filter and plumbing.

Lift pump - if your toyota is efi NO the pressure will be way high. If carb, should be fine. A facet 40106 gives me about 3psi pressure when used pre filter in a looped return system, they are cheap and readily available, durable and dead easy to fit. I use them all the time in vegoil conversions where we have to loop the return and therefore the natural purging of air doesn't happen, so a system above atmospheric pressure to eliminate air ingress is both needed and a superior setup to stock (impossible to have an air ingress leak, only a fuel leak which can be found a thousand times easier).

Any standard lift pump 5-8psi of a half decent flow should do the job. If flow isn't enough loop the return. To check it's all working t a pressure gauge into the fuel inlet on the IP, then you can eliminte the guesswork.

Greg.

whitedog said:

Greg, I'm picturing something like a Coffee can with a bunch of small holes around it and the pick-up tube in the center of that. The holes would allow fuel in, but as the fuel sloshes around, the fuel would be slow to leave the can, always providing a supply of fuel at the pick-up. Is this what you had in mind?

Rick, if those balance numbers get beyond +/- 2, you have a problem. You are well under that so everything is fine. Take those positive numbers and negative numbers and add them together. You will get close to Zero. This is what gets you to the smooth idle. If any of those numbers get beyond +/- 2, that means that the ECU is working extra hard to keep a smooth idle.

Click to expand...

In my truck I ran an inline carter pump that put out 3-4psi IIRC. I made a 3/8" pickup tube for the tank. I never had any issues with it.
On my current GTI/TDI I'm using a MK2 intank transfer pump. This pump puts out about 6 psi through a 2 micron filter unregulated. The pump itself could be used on a toyota efi pickup if you wanted with some work.

The air thing is a non-issue. I was thinking about this and yesterday I installed about a foot of clear hose to my inlet on the Beetle pump. It was empty along with whatever drained back to tank in the stock hose. I started the car and there was never a problem, even while I was fiddling with the hose at the tee to stop the air leak. This means that the small amount of air that may get in there during cornering on't be an issue. Remember that htis is a truck, not a car and it won't be cornering as fast. The only problem could be when off-roading at extreme angles with a very low fuel tank. If you are off-roading at extreme angles with low fuel, you are creating your own problems and need to see a proctologist to find your head.

e*clipse isn't that guy, ergo, not an issue.

you are creating your own problems and need to see a proctologist to find your head.

Click to expand...

Can I use your quote in my sig??

Thank you very much for the suggestions. Here's the plan:
Small lift pump ~ 5psi ordered.

I'll see what that does, because it's install will be easy - out of the tank w/ available wiring.

I'll also get mobile vag come working so I can post some reasonable logs.

- Rick

more testing - and - leak

I got some logs with 0-60 pulls, flats and uphills.

Generally, the boost worked well if the truck was cooled down a bit; but I honestly don't know if the issue is temperature related or merely resetting things by turning the motor off.

Specifically: I did a couple of tests, and there was a point where the boost stopped working. It would stay in that mode until the engine was turned off.

Upon re-starting the boost seemed to work ok.

I'll post the logs tomorrow.

Also, under conditions of no boost (up a hill) I caught these numbers:
requested boost: 2200mbar
actual boost 1200 mbar
n75 duty cycle: 92%

Could this be the problem? - A sticky vnt actuator?
I tried putting the mityvac on it and it would only move ~ 1/4" under max vaccum.

Also - the leak: I have a BIG oil leak coming from the general area of the oil outlet of the turbo. It's definitely NOT the inlet - I wrapped paper towels around the inlet & outlet to test this.
Has anyone else had problems with this? I thought it was a no pressure drain.

I don't know how far the vnt rod should stroke, but I'm sure it should be more than 1/4". I'd take the actuator loose and try to stroke the vanes by hand. Repeated motion may break loose crud internally and get it to stroke fully. Be patient and keep at it.

The requested vs actual boost deviation can definitely be due to a vnt not stroking fully.

Net pressure in the crankcase should be negative, but pressure does pulse, with peak of each pulse being positive pressure. So if there is a fault in the drain hose plumbing, it can and will leak.

Hello NC!

When I took this apart last time for this issue, I disconnected the VNT actuator. Out of curiosity, I wiggled the VNT vanes w/ the actuator free to move. Result: The VNT vanes move freely.

Regarding the drain: When I took apart the drain this last time, I tried checking some of the (obvious)
1) The drain hose is clear. I can blow through it easily.
2) The connector to the crankcase is clear; I can poke a screwdriver into the crankcase no problem.

It seems the ONLY reason it should leak is that there is some back pressure.

I made a CCV, and all that part of the system seems to be working freely. There definetely is no leakage coming from the crank case vent connection to the block.

How about warped flanges or bad gaskets? I made a new paper gasket and torqued the 6mm bolts as tightly as I dared.

Do you have the stock turbo actuator on that turbo?

Ok, here are the logs I promised. First the good news

Ok, it's no speed run - it is a pickup truck...

1) Note the requested MAP and the actual MAP. They seem pretty much on target. The actual MAP is just a bit lower.
2) Note the requested IQ and the actual IQ. They also track nicely. This may rule out some of the fueling hypothesis.
3) Also note the N75 duty cycle. It appears to be working fine.

Now for the bad news.

I don't have the requested boost or the duty cycle, but this log shows exactly what happens at about 25 seconds.

The test was done accellerating onto a hill.

At about 23 seconds, the MAP actual fell form 2000mbar to 1200mbar. Is this the dreaded "limp mode?" Or is is the "warp drive collapse?"

OK, stupid question from someone that does not know early tdi's: This is a 1Z engine right? Does it use a waste gate or a vnt turbo??

The standard 1Z had a wastegate.

Mine has a VNT-15, with a RC 2 controller mod, new N75, new MAF

My guess about the boost issue is that some limp mode safety is kicking in.

Looks like the stock actuator to me. So why only 1/4 inch travel?