88 5kq - Intermittant Hard Starting

[SJ]

From: "Huw Powell" <audi at humanspeakers.com>


> > 88 5kq - no turbo - NF engine
> > Symptom: Intermittant Hard Starting under cold and hot conditions.
> > Majority of the time, the car starts up fine . .cold or hot.
>

> > I'm suspecting the MFS temperature sensor, thats the bottom sender
> I think the MFTS only runs the temp gauge and feeds info to the AC
> system - the top, two pin black bodied sender is for the CISE3 warm up
> control (and ignition, too).
**** Yup, my bad. I had to de-confuse myself.
The bottom sensor is the MFS (Multi-function) sensor. It contains 2 switches
and 1 temperature resistor. Switch #1 grounds out when temp gets hot enough
to warrant shutting off the A/C compressor. Switch #2 grounds out and lights
up alarms when the engine is really hot. The resistor changes value
according to temp and this runs your Instrument Cluster coolant gauge.
When checking the sensors, I noticed a broken wire in the rubber cover for
the MFS sensor. This was the B+ wire. I cut out the sensor connector,
brought it to the workbench, found another connector with similar contacts
and put in another contact with wire attached. I added about 4" of
additional wire to each contact wire so things wouldnt be so tight at the
sensor where the connector gets plugged in.
Back at the car, I somehow got the wires screwed up. I reversed the B+ and
temperature resistor wires. Started engine, and instrument gauges went dead.
Checked the fuse, blown. Chit!
I figured out what I did wrong. Turns out that the MFS is the newer 3 wire
type and not 4 wire type. B+ is not needed. Started up engine . . .temp
gauge dont work. I swapped in a spare MFS, and that worked.
After having 3 beers to calm myself, I took apart the MFS sensor to see what
made it tick. As outlined above, its pretty simple inside - 2 switches, and
1 temp reisistor which was nicely fried . . just carbon remaining.
While examining the guts, I think that I found the failure mode for these
MFS sensors. The temp resistor is grounded by spring pressure only. Maybe
they use a conductive grease or something during assembly. Also, the ground
return for the 2 switches and resistor is a little brass tab on the
connector plastic that makes contact with the brass threaded body. This will
corrode over time and will stop conducting, and the first indication is a
non-reading temp gauge.

I took a bunch of close-up pix of the gizmo and will post them when I get
the roll finished.


> NTC temp sender (top one) - each terminal, resistance to ground should
> be around 500 ohms cold, drops to about 70 when hot, even lower at, say
> boiling.
**** The resistances checked out both cold and hot. I swapped this sensor
with the one from my 100Q. The 100Q has been riding around fine for couple
of days now.

> Resistance (open/closed) of idle and wot switches.
**** Yup, I checked resistance . . .its a solid Zero when closed. I rebuilt
the idle switch a couple of years ago.

> Resistance (should be about 16-22 ohms) of control pressure reg. (DFR)
**** I got 19 ohms


> Cleanliness of fuel path through DFR.
**** I took out the gray DPR (Differential Pressure Regulator). You cant
take this one apart. I didnt see any dirt in the in/out holes.


> Air leaks anywhere... can be hidden.
**** I took out all the rubber stuff and checked it. Its all airtight. No
cracks. The major items were replaced about 3 years ago. I cant do the
"dipstick stall test" as crankcase air going to the intake manifold is
controlled by a .070 orifice.
I did find a leak in the vaccum line going to the A/C controller. I
disconnected the line and plugged it with a golf ball "T".

> Cleanliness, function of ISV.
**** It works. Unplug it and RPMs increase. I gave it a good cleaning about
4 years ago . . .and cleaned it again lightly about a year ago.


> Spark quality - wire resistance, plug quality, cap-rotor age, hall
> sender problems.
**** I replaced plugs, cap/rotor when I first detected this problem. Wires
are couple years old.
I dont think problem is the hall sender - there is no abrupt stop like
cutting out ignition, or misfiring. When the problem appears, the RPMs start
dropping . .playing with gas pedal brings them back up.


> Grounds to intake manifold & all other electrical connections.
**** Grounds are tight. I redid them couple years back  . . .removed,
cleaned, new hardware, put back. I loosened and tightened them again.

> Ignition switch intermittence in "run" position...
**** Dont think so . .no sudden stop to engine. Playing with gas will
maintain RPM.s. Jiggling the switch wont stop the engine.

> There is more, but only maybe a dozen things to check all together.
**** Yeah, I decided to do a DPR current and Oxygen sensor voltage check
when starting up.
DPR starts out at about 90 ma, and drops to about 13 ma in 3 minutes. The O2
voltage starts at 0.5 volts and drops pretty fast to almost zero. And after
a 5 minute warmup, the O2 voltage is still at zero.
I did the same test on the 100Q. DPR started at about 90, dropped to about
10, then started swinging between 7 and 12 ma. The O2 sensor started at .5
and in about 1.5 minutes started swinging between .2 and .9 volts.
I ordered a new O2 sensor. Will try and put it in tomorrow. Then I'll check
DPR current, and O2 voltages.
Still, I dont think the O2 sensor causes the hard starting symptoms as the
problem occurs right at startup when engine is running in open-loop mode.
I'll do some more testing after I put in the new O2 sensor.
I'll also check the fuel system pressure as soon as I make an adapter to
connect to the cold start injector line.
Maybe that leaky A/C vacuum line is the culprit. This line terminates inside
the A/C controller head. It connects to other lines and valves via a 'cone
and socket' rubber contact connector. Just a bit of pressure is what holds
the connector together. I can see temperature affecting the rubber and
making/stopping air leaks.
I'll do some testing after the O2 sensor gets installed. Will try starting
the engine with this vacuum line open and see what happens.


SJ



-- 
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.394 / Virus Database: 268.10.4/399 - Release Date: 7/25/2006