'90 3.1 sick engine repair success story

Reply Subscribe

Thread Tools

Search this Thread

11-04-2006, 12:40 PM

sixgun

Junior Member

Thread Starter

Join Date: Sep 2006

Location: Ft. Worth, TX

Posts: 30

Likes: 0

Received 0 Likes on 0 Posts

Car: '86 TPI T/A, 90 Firebird 3.1L

Engine: 3.1, 5.0 TPI

Transmission: 700R4

Axle/Gears: 3.23

'90 3.1 sick engine repair success story

Well, late this summer I got a great deal
on a '90 firebird. Not needing it right away I
slowly started going through the car, seeing
what needed what.

The most glaring problem was a rough idle and
poor acceleration. Steady speed cruise wasn't
really smooth either.

After checking the obvious, I decided to look
at the engine controls; i.e. the ECM.
Asking the computer to do a self check; it gave
me the familiar 1-2, 1-2, 1-2 indicating the
self diagnostics found no faults, no codes.

Since the thing ran so bad, I decided to look
at the sensor ouputs.

Many tools both old and new are capable of this,
but I use Diacom - it's software that links any
PC through a cable to the 1981-1992 GM ECMs.

I watched all of the engine sensors and parameters,
they looked fine but... The ECM never went into
closed loop operation. The computer was never
controlling the engine.

From what I read in the Pontiac shop manual,
the computer has a back-up basic set of tables
that run the engine if there's a failure.

But as I shut down the engine, I suddenly got
half the trouble codes in the book. Including
ECM and mem-cal failure codes.

So, I re-grouped (cold beverage). I had a ECM
computer that on one hand said it's diagnostics
were fine, never went to closed loop, AND said
it and the mem-cal had failed. I replaced this
ECM with a re-built unit; ($70) after returning
the core.

This ECM checked, went into closed loop, and held
no codes at all. The idle was now a whole lot better, but pickup and steady speed cruise weren't right.

So, I re-grouped (another cold beverage).
I had found a bad ECM that was so broken it couldn't
even correctly diagnose itself as broken.

I went back and again starting watching everthing
the ECM was looking at; some times I had to research
to understand what I was seeing. And still no codes.

So, I re-grouped (yet another cold beverage).
My ECM is now working fine and controlling the engine,
but no pick-up. That's squarely the domain of fuel
control I figured... Fuel control; well that's O2
sensor, Throttle position sensor and injector interval.

I read up on what a healthy O2 should perform like,
read up on correct TPS values, and gave up on injector
timing - this was nothing I could effect easily.

As I again watched all of the engine data the culprit
finally became known... TPS values.

The book says a TPS should show something like .5v to
almost 1.0v for idle, and the computer would remember
whatever the value is. (they don't externally adjust)
It also says that at WOT (wide open throttle) you
could see values approaching 5.0V.

Where was mine... .74v at idle and a max of 1.1v with
the throttle over half way down! My TPS sensor told
the ECM that it was NEVER going much above idle. The
ECM was only allowing idle fuel at any physical position
of the throttle! So my TPS worked, just not very well.
Incidently - it worked well enough to NOT set a code.

After a quick trip to the Auto Parts store, I replaced
my TPS with a new one (~ $33) and fired it up.

Walla! engine purrs, revs right up, takes right off.

So, I celebrated... (final cold beverage)

I wrote this because all I usually read is what's wrong
and many times someone's suggestion; either part or
procedure fixes a problem and the rest of us miss out
on the solution.

And while I'm sure everyone else's fix is bound to be
different from case to case, I thought my journey from
beginning to end may help someone else understand what
it may take to find what's keeping you from enjoying
your ride...

Sixgun

Oh, by the way... The Parts people said a '90 3.1 Muliport
doesn't have a TPS listed (I found this amusing) but
the listing for '91 is exactly correct.