Date: Fri, 17 Aug 2012 17:44:57 -0400
Reply-To: David Beierl <dbeierl@ATTGLOBAL.NET>
Sender: Vanagon Mailing List <vanagon@gerry.vanagon.com>
From: David Beierl <dbeierl@ATTGLOBAL.NET>
Subject: Re: Start-Stop Syndrome Returns
In-Reply-To: <502E91F3.60509@gmail.com>
Content-Type: text/plain; charset="us-ascii"; format=flowed
Dear Mr. Squirrel,
Query - if you stop partway up a hill, a) how long must you stop
before it makes a difference; and b) once it has made a difference,
how long before the nasty behavior comes back?
At 02:48 PM 8/17/2012, Rocket J Squirrel wrote:
>1. This is not a random cut-out, but quite periodic. The hotter the day,
>the steeper the climb, the more frequently the engine cuts out. On a
>moderate grade, the engine will cut out every 45 seconds; as the grade
>steepens the engine dies every 15 seconds.
Very suggestive. Ambient temp- and high load-related, cyclic. My
immediate thought is something electrical that works harder at high loads:
Fuel pump - no. Constant-load/constant output operation pretty much
regardless of circumstances.
Ignition coil?
Ignition control module?
Idle stab module (low-power but in the engine room where temperature
changes with load)? Not likely?
Distributor - same as idle stab.
Distributor connector - general bad actor but I don't buy it.
Injector driver?
Injectors - no, there are four of them; they wouldn't be doing
Olympic synchronized dying.
System control relays - no, they're big fat power relays and load
changes don't go through the control coils, just the load contacts.
ECU general?
Wiring?
>2. Duration of no-power lasts about five seconds.
I'm inclined to look less on the ignition coil because I think it
would take longer to cool down and start working again
>3. The fact that the tach gives a signal when the engine is not
>providing power indicates that there are ignition pulses, yes?
On the primary side, yes. Not necessarily perfect ones, but they
exist. That gives a qualified pass to the ignition primary, since if
it goes open there will be zero voltage on the negative (pulsing)
side of the coil.
Also qualified pass to the ignition driver module on the wall fwd of
the idle stab module, and probably a clear pass to the
distributor. Idle stab messes with the timing, so in principle it
could be offsetting it out of the firing range.
> So this
>is possibly a fuel delivery issue?
Injection, easily. Supply, much less likely. I can't see anything
in the supply system inclined to cyclic failure - ah! *except* for
the ECU control to the fuel pump relay. I don't see that as
inclined; rather a remote possibility. If you don't have sufficient
fuel in the tank to keep the inlet covered at your angle of climb of
course you're out of gas, but that doesn't meet other constraints.
>4. When engine is running, all instruments indicate normal operation,
>the O2 measuring gizmo shows normal bounce-bounce action.
The gizmo is Ken Lewis's little LM3914 ten-bar LED voltmeter, which
he sets to a range of a bit over zero to a bit under one volt. Very
handy device, and someone who wants to could make the voltage range
switchable to anything desired.
>5. When the engine is dead, the O2 gizmo is dark -- neither rich or lean
>condition indicated.
This either means "not operating" or lean (or off-scale rich, but
unless the calibration or connections are off the maximum sensor
output is still visible, so not likely). I think this nails the
diagnosis pretty hard to fuel not getting into the cylinders. I
think you can set the scale down so a slight negative voltage will
still display one bar - Ken Lewis or the (National Semi) LM3914
datasheet will tell you. I used to know, but there are a lot of
things now that I used to know.
>6. These grades are steep enough that I run either in (auto trans)
>second gear at about 2500 rpm, or 1st gear at about 3000 rpm. Thinking
>that maybe the AFM was sitting on a flat spot I tried wiggling the gas
>pedal but it did not change the behavior.
You're driving on the tach, correct? So you use increasing throttle
as the grade increases?
>7. It's fairly nerve-wracking to be a couple miles from the summit and
>have the engine die one or more times a minute.
There, there. Think if you were a couple miles above the summit.
>8. Once the road flattens, the engine runs fine. If the day is cool, the
>same grades do not cause the problem.
And this will either stay the same or more likely get worse.
>So what I'd appreciate from
>the group would be a list of candidate parts I can replace over the next
>couple months and then, on our next big ol' camping trip up the sides of
>mountains in summer, we can see whether the problem re-occurs.
I think on balance the injector driver is by far the likeliest candidate. So:
ECU first.
Also get your tech to install a DB25M in the ECU case and bring out
low-current LED monitors to the dash from the distributor timing
pulse where it enters the ECU, the injector pulse coming out, and
the ECU control signal to the fuel pump relay. Use some ferrite at
the ECU end of your cable. If changing the ECU doesn't fix your
problem this will give you much additional info about wiring
troubles, fuel supply and so forth, and it's a good start toward a
full Darrel Boehler-type Digitool. You know that you have an
ignition timing pulse because it's getting to the ignition driver
module, but it also has to get inside the ECU to give it timing info
and to keep the fuel pump running. If you're feeling rich just talk
to Van-Cafe about putting one of their Digi-thingies in, like
Darrel's but slightly fancier.
>FOR YOUR AMUSEMENT, MY BEST BUT PROBABLY SILLY IDEA: Bubbles in the fuel
>line. The regular, nearly clock-like periodicity of the cut-out makes me
>think of those old school Christmas tree ornaments that have bubbles
>rising in a tube of water.
Don't drink that water! Now if you need some methyl chloride to put
down your budgie or whatever, Bob Marley's your uncle.
> The hotter the day, the more bubbles; the
>steeper the climb, the more bubbles.
The pump shoves a liter per minute into the fuel ring, and most of it
simply flushes out the ring and runs immediately back into the tank
(it pulls out a full tank every hour, minimum). A major reason for
this is to keep the fuel in the ring nice and cool and non-bubbly.
Yours,
David