Prevent engine runaway with the strap on mod?

[Vince Waldon]

My point, perhaps poorly made, is that when you turn off the key on the highway you are turning off *both* the fuel and (for some unknown period of time) the air supply. The engine will stop... it has no fuel and "potentially" no air. So not a valid way, IMHO, to confirm that the ASV stays closed for as long as the key is off and would therefore choke an engine running on oil being injected into the intake manifold from a blown turbo.

(As an aside, I know that the topic of ASVs and runaways can take on religious-level fervor on TDIclub and generally I stay away from such strident topics. In this case, having a runaway myself peeked my interest in all the "no problemo just turn off the key" threads because it might be potentially dangerous for folks to have false confidence in the ASV as an anti-runaway device.

Or perhaps such confidence is warranted. Science will tell...and it beats arguing about which oil is best?! )

 

[VeeDubTDI]

I understand what you're getting at, Vince. My point was that turning the key off on the highway (while in gear) is a suitable method for testing the ASV's effectiveness at blocking airflow. The engine will keep turning because it's in gear (although with a significant amount of engine braking happening), so as long as you have a good head of speed or you're going down hill, you can test how long the ASV stays closed as long as you have forward momentum.

 

[whitedog]

... as long as you have a good head of speed or you're going down hill, you can test how long the ASV stays closed as long as you have forward momentum.

How will you know that it's closed?

 

[VeeDubTDI]

Monitor manifold vacuum with a gauge?

It also makes a noticeable change in engine tone when it's closed. Cycle between key off and key on while experimenting and you'll see what I mean.

 

[robnitro]

Monitor manifold vacuum with a gauge?

It also makes a noticeable change in engine tone when it's closed. Cycle between key off and key on while experimenting and you'll see what I mean.

I thought of that too and tried it out a while back.
I realized the engine tone also changes because the vanes are set open when key is off, changing the exhaust backpressure.

The best way is to use a vac gauge T'ed off the feed to the ASV.

But what good is a vac operated device during runaway. In runaway I would think blowby would be so high that the vac pump can't create vacuum (as it pumps air into the crankcase).

 

[NarfBLAST]

Possible Solution: Manual Operated ASV from the drivers seat

I had an idea to use a bicycle hand brake with a long thin cable connected to the ASV. When you want to stop the engine just squeeze the bicycle brake. I have the EGR/ASV off the car now while I am waiting for an oil feed line for the turbo to arrive in the mail this sounds like a fun project. I already have the chunk of bicycle handlebar, the hand control, the cable, just need to mount everything and run the cable thru the firewall hole above the accelerator pedal and fabricate a bracket to run the cable along side the vacuum actuator for the ASV (I like my shudder free stops) or maybe do away with the the vacuum actuator and simplify the vacuum lines even more... I could just shut down the engine every time with the hand control.

 

[jasonTDI]

This is so rare it's silly. So much work for an uncommon issue. Just saying....

Most of the time if its running away, its damage done long before you could ever stop it.

 

[whitedog]

But what good is a vac operated device during runaway. In runaway I would think blowby would be so high that the vac pump can't create vacuum (as it pumps air into the crankcase).

Well, there is a reservoir for the vacuum with check valves so vacuum would be available.

And blowby is what goes past the rings during combustion. You wanted to say, "Crankcase pressure".

And since the crankcase pressure is routed to the vacuum side of the intake, since the engine is running away, vacuum there will increase, thereby moving more crankcase pressure out, so the vacuum pump will still be able to create vacuum.

 

[NarfBLAST]

<snip>...so the vacuum pump will still be able to create vacuum.

So would it be as simple as having a big red emergency shut down button to trigger the solenoid for the ASV? Or does that solenoid only work when pulse width modulated? Oh wait, I can answer this in a few minutes here... (all the parts are out of the car, and now nice and clean, on a workbench)


Meanwhile, this was my Sunday Afternoon Project, a bike brake grab handle to shut the ASV:

It worked except that it binds and doesn't close completely. I decided to back out of this project but at least I got a nice clean EGR valve and a new ASV arm out of it:

Anyone need to repair their broken ASV arm, I bagged up the old parts that came off:

 

[NarfBLAST]

So would it be as simple as having a big red emergency shut down button to trigger the solenoid for the ASV?

YES! Click thumbnail for video:

• 1:13 ALH TDI ASV test with 12V source Now to find a big red emergency shutdown button for the dashboard and wire it to the anti shudder valve (ASV) to turn it into an emergency shutdown valve (ESV)

 

[VeeDubTDI]

Nice video. Looks like there's a pretty good air gap around that butterfly valve, though. Possibly enough for the engine to continue running with a fuel source. Looks to be about the same gap as a gasser would have for idling.

 

[NarfBLAST]

Looks to be about the same gap as a gasser would have for idling.

But doesn't the diesel need enough air to compress and generate enough heat to ignite the fuel?

 

[Genesis]

I will tell you that these engines will not fire if the valve is stuck closed

 

[robnitro]

Well, there is a reservoir for the vacuum with check valves so vacuum would be available.

And blowby is what goes past the rings during combustion. You wanted to say, "Crankcase pressure".

And since the crankcase pressure is routed to the vacuum side of the intake, since the engine is running away, vacuum there will increase, thereby moving more crankcase pressure out, so the vacuum pump will still be able to create vacuum.

Yes, sorry, I meant that the blowby will cause pressure in the crankcase which will make the vacuum pump weak.

I'm not sure the stock CCV vent on our valve cover can handle the huge flow of gasses that would come from extreme overfueling, such as during a runaway. In fact, some have run a second CCV vent, an oil cap upgrade that has it's own CCV.

As for the vacuum reservoir, what good is it, if the solenoid for the ASV (N239) leaks, as mine did? Check yours and see, unplug the vac supply, attach a hose, suck on it and hold vacuum. See if it leaks down. That also may be why the ASV fails to stop runaways?

 

[whitedog]

Rob, I got it now. Because the engine is running away, blowby increases.

This also makes me wonder about what happens when an actuator nipple has broken off and is "fixed" with a WD-40 plastic tube. Will it still pull the rod when it's needed?

 

[robnitro]

Simple test... take the vac hose off the solenoid and suck on it and use your tongue to hold vacuum (or use a mityvac). See if it holds the ASV closed for some time (leak down test). My solenoid was leaking down, but thats a little harder to test.

 

[akafred]

man i got a sweet minus 90hp co2 set-up on my car.. lol

You could make rig a CO2 extinguisher which would be hooked into your MAF, run a wire with a handle to the drivers side. In case of a runaway, pull the cord and the CO2 will kill the engine without any permanent damage.

Kind of like a NOS setup, but with CO2 instead.

 

[VeeDubTDI]

Do you have those awesome purge valves in your wiper squirters???

 

[TDIJetta99]

I did a quick test with a light bulb a while back.. I shut the ket off going down the mountain in 5th gear.. The ECU had power and ground to the ASV solenoid all the way down the hill, and continued to have power there until a second or two after the engine stopped.. The vacuum source will be there as long as the camshaft is still turning.. The vacuum pump will produce more than enough even with a serious amount of crankcase pressure since it's a mechanical pump and there's a crankcase vent..

 

[Birdman]

.

How would you differentiate between that and over-run or downhill acceleration?

If I can tell that my old Rabbit diesel (maybe 80 HP with the turbo ) was starting to runaway by the guick acceleration (faster even then a full throttle run)or at least something was wrong, then anyone that knows how their car runs normally will be able to tell the same thing.
Mine did it not once but twice on the same 8000 mile trip. Both times I just stood on the brakes till it stalled out.
Now someone who just drives these to work and back, taking it to the dealer when the check engine light comes on or for oil changes might not know what is happening but anyone that has has worked on and knows how the car should run will be able to tell by the acceleration of the car will know something is wrong. Anyone reading this thread should by now know if or when the car starts to run-away to turn it off. Normally a blown seal on the turbo will put enough oil into the engine at the beginning of the seal failure to cause the engine to accelerate, if shut down at that time with a working ASV which will stop all air from the engine will shut it down. An engine can not run with out air period. this is why making the race pipe using the EGR valve with the ASV is the safest way to go.

 

[VeeDubTDI]

I meant the computer, regarding TDIMeister's comment about adding code to the computer that would recognize unintended engine acceleration. How does the computer identify unintended acceleration?

 

[TDIMeister]

I guess I should clarify. The simplest way I can see it working is that the ECU computes a runaway using a Boolean AND logic on two plausibility criteria.

The first is the gear/engine speed relationship. All gear ratios, i.e. 1st, 2nd, ..., can be simultaneously calculated to correspond to certain engine RPMs and vehicle speed. If this gives an implausible result, the only interpreted outcome are a defect in either the VSS or engine RPM counter, or the vehicle is in neutral/clutch disengaged.

The second and more important plausibility criteria is that every ECU has maps of injected quantity related to engine speed and TPS %. If there is a mismatch in relationship of the injection quantity and/or TPS% to the engine RPM (or more specifically the rising rate thereof), then it triggers an implausible signal.

Only when both implausibility #1 and implausibility #2 are simultaneously detected, this is interpreted as an engine runaway and the ECU can shut off fuel and also activate and hold the ASV shut.

Hope that clarifies things.

 

[TDIJetta99]

just turn the key off... If you have a properly functioning ASV, it'll shut it off...

 

[turbocharged798]

I did a quick test with a light bulb a while back.. I shut the ket off going down the mountain in 5th gear.. The ECU had power and ground to the ASV solenoid all the way down the hill, and continued to have power there until a second or two after the engine stopped.. The vacuum source will be there as long as the camshaft is still turning.. The vacuum pump will produce more than enough even with a serious amount of crankcase pressure since it's a mechanical pump and there's a crankcase vent..

Was that on your 99.5? I have to wonder if all years function the same.

 

[slamhouse]

I guess I should clarify. The simplest way I can see it working is that the ECU computes a runaway using a Boolean AND logic on two plausibility criteria.

The first is the gear/engine speed relationship. All gear ratios, i.e. 1st, 2nd, ..., can be simultaneously calculated to correspond to certain engine RPMs and vehicle speed. If this gives an implausible result, the only interpreted outcome are a defect in either the VSS or engine RPM counter, or the vehicle is in neutral/clutch disengaged.

The second and more important plausibility criteria is that every ECU has maps of injected quantity related to engine speed and TPS %. If there is a mismatch in relationship of the injection quantity and/or TPS% to the engine RPM (or more specifically the rising rate thereof), then it triggers an implausible signal.

Only when both implausibility #1 and implausibility #2 are simultaneously detected, this is interpreted as an engine runaway and the ECU can shut off fuel and also activate and hold the ASV shut.

Hope that clarifies things.

Its people like you that i would love to have designing the power trains in my vehicles...

 

[40X40]

Its people like you that i would love to have designing the power trains in my vehicles...

Coincidentally, I think that is what he does, did, or will do.

Bill

 

[TDIJetta99]

Was that on your 99.5? I have to wonder if all years function the same.

Ya know, I can't remember now if it was the 99.5 or the 03.. It was probably back when the 99.5 still had an ASV.. it hasn't for over 100k now..

 

[TDIMeister]

Coincidentally, I think that is what he does, did, or will do.

All of the above, but somehow I don't think Slamhouse meant what he said as a compliment...

Edit: Plausibility criteria #1 is not really even needed. If the ECU detects that there's no TPS input and the engine RPM is increasing much more rapidly than should be the case given the amount of injected fuel, BINGO it's a runaway and should do an orderly shutdown with the necessary steps to fueling and ASV.

 

[TDIJetta99]

The issue I see happening is with a manual trans downshifting.. revs can snap from 1000-3500rpm in a fraction of a second.. You would need it to be able to determine weather it's a runaway or just a simple downshift without rev-matching..

 

[TDIMeister]

That's very true... Then criteria #1 is also needed after all and probably even more complexity...