85 Vanagon Digijet

[Lanval]

It was a chilly 68 today. Howz 'bout you come back here. I could fill your time... (don't know about your pockets, tho).

Stupid crossover pipe. I should switch to a real 2.1 set up.

L.

[satchmo]

Stupid crossover pipe. I should switch to a real 2.1 set up.

L.

I agree, the 1.9 cooling plumbing is a little weird, plus now many 1.9 hoses are NLA.

Did you ever determine the actual size of your engine?

Tim

[Lanval]

Not really ~ We checked the case number, and it's a 2.1 case. Given what the PO told me about the job that was done, it's self-evident that they just dropped a 2.1 bottom and hooked all the 1.9 stuff back up.

I couldn't prove it though.

On the plus side, our McGyver fix on the cross-over pipe holds ~ that's not the leaky spot. The engine is leaking from the little hose that crosses from the thermostat housing to the 3/4 head. I double checked it by loosening one clamp at a time and watching to see if more fluid came out when I did so. The first one I loosened (next to then thermostat) changed nothing. Then I tightened that one and loosened the clamp next to the head. I got some fluid from underneath, and when loosened, began to drip more quickly.

I retightened the clamp, and returned the A/C bracket and compressor. I tried to fix a couple of bad wires, wrapping them with electrical tape, etc.

When I went to start it, though, no joy. Interestingly, the fuel pump doesn't run when I turn on the ignition. Colin says that it shouldn't run with the ignition on anyway, but it did before, so clearly something is changed. Grabbed the wife's car and I'm at work now. I'll cross post this into the electrical section...

[Lanval]

noting the comment about the hoses...

Apparently putting an end to the substantial leak at the heads has caused a smaller leak in one of the down pipes coming off the radiator. Makes me wonder if the fix of one thing is going to lead to a series of failures in other weak areas....


What's involved in the switch? It seems like it could only be the hoses/oil cooler/thermostat and water pump (also, the coolant reservoir is different ~ I know because I ordered the wrong one from Van Cafe!)

[satchmo]

noting the comment about the hoses...

Apparently putting an end to the substantial leak at the heads has caused a smaller leak in one of the down pipes coming off the radiator. Makes me wonder if the fix of one thing is going to lead to a series of failures in other weak areas....)

That seems to be usual, and all too frequent, experience on older cars, especially those that have been sitting unattended for several years.

I don't know what all is involved in the switch to the 2.1 system, but besides the difference in coolant plumbing, there are electrical differences in the ECU and on up to the dash. Talking with Peter at VolksCafe might help in sorting it all out.

Tim

PS: The fuel pump on my 86 certainly runs with key on.

[Amskeptic]

PS: The fuel pump on my 86 certainly runs with key on.

Constantly? Ignition on but engine not running?
No bueno.
Colin

[satchmo]

PS: The fuel pump on my 86 certainly runs with key on.

Constantly? Ignition on but engine not running?
No bueno.
Colin

Si. Ignition on, idiot lights on, coolant warning light goes off in a few seconds, and fuel pump is running the whole time. The tone emitted by the fuel pump changes when the system is pressurized, but it keeps running. Turn the key a little more and, presto, the engine starts.

Yours not like that? Too bad, so sad.

I haven't checked to see what happens to the fuel pump when the engine dies with the ignition in the on position.....yet.

Tim

[Lanval]

PS: The fuel pump on my 86 certainly runs with key on.

Constantly? Ignition on but engine not running?
No bueno.
Colin

So you say; what should it be doing? As you noted, the internals of the AFM are different, and I don't have the pin to stop the flow of the fuel pump. So...

How does the water-cooled ignition system work?

With the ignition in the "on" position, the pump is now getting no power ~ I checked using the voltmeter. I've obviously changed something, but not sure what.

I'm going to check a couple of things ~ the first is my bentley wiring diagram which I have here. I'll look at the pix of my pre-work engine. I'm wondering if I changed something while fixing the leak today.

L.

[Amskeptic]

Satchmo,

Do us a favor. Start your engine. Let out the clutch in 1st gear and stall it.
Fuel pump now silent?

Lanval, you have a pump that is not running?

Check your grounds, seriously.

Disconnect ECU plug.
Find the lower right end terminal (the one furthest from the harness)which is #25.
Bridge it with a jumper wire to #20 (that is five terminals to the left on the lower row still).
Turn on ignition and pump should run.

Check your grounds [you should have 5 (five) grounds out of the fuel injection harness] and any stray wires that might be critical grounds. Please note that the temp sensor value relies on the integrity of the ground between the left head and the body.

You can also check the fuel pump power supply at the relay. #30 is the power supply. #87 is the destination wire to the pump itself. Do not apply a test light to #86 with the ECU plugged in.

Helpful?
Colin

[Lanval]

C'mon, why don't you come out here and help me do it?

It's going to be 71 tomorrow....

Also, Costco sells these cool heat lamps. They're cheap, and I could get one or two... heh, heh.

You know you want to do it. You miss the taste of rat turd, admit it!

Oh, rats.

L

[Lanval]

Man, so weird. I'm going to post up a pic of what I found in the relay box. Gawd.

Pic one:

http://picasaweb.google.com/lathropMA/N ... 1552199154

Pic two:

http://picasaweb.google.com/lathropMA/N ... 1572474290

Since no sane person could imagine doing this, I'll spare you the mystery. What you're looking at is a little light that was wired into on of the wires coming off the relay for the fuel pump. The positive side of the wire was taking power from the 86 output, which goes to the number 1 fuel injector. The negative side was stuck through a hole in the relay box, and grounded on the screw that held in the relay box.



OK; here's what (and how) I know.

I have a spare battery, so I disconnected both power and ground to the pump and connected the battery to the pump directly. The pump ran, the pump is reasonably functional (Colin would probably have verified pressure and flow here, but I've got my limits...)

I cut out the little light ~ I can only imagine what they were trying to test...

Both relays are getting power in. I can't easily verify power out to the 87 on the ECU relay because of the layout, but I was able to check on the Fuel Pump relay. The Fuel Pump relay is not putting anything out at the 87.

I didn't have time to create a little jumper cable from the 30 to the 87 position, so I can't verify that the relays are bad at this point. That's tomorrow's job.

HOWEVER:

I'm fairly certain that I DID short the 30 to the 86 on the ECU relay. This shouldn't be a huge problem, I would think, since those two points go from power (30) to ground (86). Perhaps someone could point out if the fact that connecting those two should or should not make sparks...?

I was using an allen wrench to pry the relay off, and it sparked. I assume that means it is correctly grounded, but am not sure. There is something about this that is very, VERY strange, to me at least.

The Wire which runs to the relays from the battery (this is the red wire which comes into the box on the right side ~ you can see it in the pix above ~ there are two red wires which come into a connector on the right side, and you can see the red jumper that moves from relay to relay inside the relay box) has power EVEN WHEN THE BATTERY IS NOT GROUNDED!

How the hell is that possible? Why does it show 13.75 volts, when the battery only measures 12.6 and ISN'T HOOKED UP?!

Anyway, that's why it shorted ~ I didn't know the damn relays still had power even with the battery ground disconnected.

Two problems that amaze me appeared as well (following careful study of the vanagon Digijet Electrical Wiring Diagram in the Bentley):

1. The ground for the idle switch is not connected.
2. The temperature sensor on the thermostat (Temp Sensor II) is not, and has never been connected to the ECU since I've owned this van.

Guess I'll be correcting this tomorrow too.



Addendum ~ Does anyone know offhand how to use a multimeter to check continuity? Is that the "ohms" function?

I'll read up on it, but I was wondering if there's an easy way to check if the pump is grounded with 12v going to the positive side...

L.

[Lanval]

You can also check the fuel pump power supply at the relay. #30 is the power supply. #87 is the destination wire to the pump itself. Do not apply a test light to #86 with the ECU plugged in.

Colin

Done. By jumping the Fuel Pump relay from #30 to #87 the pump runs, and the van will start (albeit run unholy rich).

I verified that the other relay (ECU Fuel relay) does indeed seem to be doing something. With both relays removed, and the fuel pump relay jumped, the pump will run, but the van will NOT start.

Add the ECU relay, and the van will start.

**********************

Temp fixed most of the broken wires, though several remain unconnected. On Sunday I'll try to run those down, and then verify the ECU tip you've given in the above quote.

**********************

I'm wondering if Amskeptic could give us a run-down on how the whole thing is supposed to work. The reason I ask is because both Satchmo's, mine and others have pumps that run all the time. I suspect this is some kind of fix that is put in to resolve a no-start problem (relay issue?) or something that happens to the system when something sticks open (relay?).

In my case, I'm fairly certain that it's something in the engine compartment, since that's where I was working, but I'm also wondering, if the pump is supposed to come on briefly when the key moves to start, if the ignition/key system isn't part of the problem...?

L.

[Amskeptic]

I'm wondering if Amskeptic could give us a run-down on how the whole thing is supposed to work.

Read this Baywindow-oriented article.

When you run into the AFM circuit (36/39 contacts) subsitute for ECU in the Vanagon. I do not know the specific parameters for ECU power to the fuel pump. I am waiting to see if satchmo's experimental stall kills the pump after the engine has been running.

I'm fairly certain that I DID short the 30 to the 86 on the ECU relay. This shouldn't be a huge problem, I would think, since those two points go from power (30) to ground (86). Perhaps someone could point out if the fact that connecting those two should or should not make sparks...?

You have got to type carefully when we are trying to figure things out.
Those two points go from power 30 and 86 to ground? So in the same sentence, you're now calling 86 a ground? That's a damn dead short. You should never ever have sparks in the control side of a relay like the 86 > 85 , electronics get fried instantly. You can expect a little spark if you are firing up the switched load (the fuel pump in this case) between 30 and 87.
Colin

[Lanval]

Read this Baywindow-oriented article.

No link here. What's the URL?"

I'm fairly certain that I DID short the 30 to the 86 on the ECU relay. This shouldn't be a huge problem, I would think, since those two points go from power (30) to ground (86). Perhaps someone could point out if the fact that connecting those two should or should not make sparks...?

You have got to type carefully when we are trying to figure things out.
Those two points go from power 30 and 86 to ground? So in the same sentence, you're now calling 86 a ground? That's a damn dead short. You should never ever have sparks in the control side of a relay like the 86 > 85 , electronics get fried instantly. You can expect a little spark if you are firing up the switched load (the fuel pump in this case) between 30 and 87.
Colin

I'm not saying I didn't blow the thing, but here's what I said in clearer terms:

While trying to remove the ECU relay, I shorted between #30 which is power in to the relay, and #86 which is ground to the body.

Given the situation with the wiring, it's totally possible that it's not grounded, but I'm assuming that the fact that the connection sparked is what you'd expect if I was completing a circuit (however dangerously).

I still don't understand how the system is powered... why is there power even when I disconnect the power at the battery?

I'll head over to the Samba and see if I can find the Digijet manual.

[satchmo]

Amskeptic wrote:

I am waiting to see if satchmo's experimental stall kills the pump after the engine has been running.

This morning's report directly from the garage:

The pump runs at key on (accessory position) prior to engaging starter. After stalling the engine, with the key in the same position, the pump runs for a fraction of a second, then stops. The only thing I hear running after that is the idle control valve, which I don't believe the Digijet system has.

Helpful?

Tim