"The engine light came on..." says my bride when I walk in the door after work New Years Eve.

The daylight is about gone and a light drizzle helps the cold penetrate as I get the codes from the LH2.4 computer on this 1989 245GL metallic blue-green: 113 and 232 - injector problem; compensating for too lean or too rich mixture. A couple whiffs tell me it's a rich mixture. I pull the AMM hose and open the throttle to a knock-you-over smell of raw gas.

I shiver and recall Smitty's "Fruit of the loom" moment post and shut off the motor before pulling the FPR vacuum hose. Smells iffy. Well, I've never sniffed one with gas running through it from a broken diaphragm, and I still smell the rich fumes from throttle body. It's dark now, I want to check the pressure but I'm a wimp and decide to wait for daylight.

Rains all day yesterday. Good, we don't need the car anyway on New Years Day.

Tonight, home a little earlier, I get my gauge on the fuel rail to make sure the FPR is OK and the injectors aren't leaking. Solid 40 pounds, up to 43 referenced to atmosphere. Walk away for 15 minutes, come back, still 40 lbs.

Thinking ECU... AMM... ECT...

Put meter on the O2 sensor - warms up to 830 mV and sticks there. Stays there even through throttle blips and reducing rail pressure to 35 lbs using a vacuum tester on the FPR. This thing is dumping the gas in. Yes, the oil stick looks a tad over full!

Found novel (to me) lazy way of depressurizing the rail - to disconnect the gauge. The OBD will cycle the main injectors for testing (three 1sec. presses), so in the time it took me to check the sound from each injector the fuel pressure had relieved to about 10 lbs. Not that much gas, but I'll still change the oil this weekend. I didn't try this, but maybe you could, by connecting one injector at a time, identify a grossly clogged unit watching the gauge as the computer cycles the test.

Took the gauge off and pulled the ECU connector to check the ECT. Looked good, about 480 ohms at pin 13 and rising as the engine cooled some. Pretty much left two items, ECU and AMM.

Having an 89, I learned early reading this board to have an ECU spare. Found one in the pick and pull about a year ago in a 92, but it was a -561, supposedly the unreliable model. The 89 has a -933 (yes 933) so I figured I might never need my spare.

Popped out the 933 and put the 561 in. O2 sensor now shows nice even transitions rich/lean and no more codes are being set. This forum and you all have saved my bacon again! Smells better already. My customer cooked dinner for me and that smelled even better.

Looked at the 933 - there's a bit of epoxy charring under R15 - some melted solder at its leads... well that will be another post if I learn anything.

--
Art Benstein near Baltimore

Art, your post gets printed, and saved, to my "Troubleshooting" file, thanks again for it ! Tom in Oregon

Hi Art;

When I saw your post I was quite sure that you wouldn't be asking a question, and just as sure that I would learn something if I read it anyways. I did, and I did. Thanks for the insight.
Rhys

Rhys, that was a pretty nice thing to say, though I'm curious, what part of that cold, wet, dark experience enlightened you, sir? Was it Smitty's post? That was a scary one, although I can only imagine because I've yet to encounter an FPR with the diaphragm leak. I'm afraid just one carbeque could put an end to my DIY career.
--
Art Benstein near Baltimore

Art, your "get at it" attitude towards the repair (no dithering) and logical progression through the problem impressed me - and checking the resistance through the multipin connector - I usually struggle under the hood with such things.
I agree that a bad FPR can ruin your whole day. The worst I ever saw wasn't leaking, but produced over 90 PSI in the rail - customer said it stalled a lot and smelled "funny". Yikes.

Don't usually "get right at it". Normally I dawdle and clean stuff and take the camera, but trying to find a place where my tray of tools would stay dry under the hood was egging me on.

That stuck FPR was one of my first Volvo experiences. When I took the regulator apart, the little spring looked brand new but rolling around in the chamber. Pegged my 100 lb. gauge.
--
Art Benstein near Baltimore

I believe model 933 is the same as 561, with additional circuitry for the EGR. Your 561 probably isn't controlling the EGR since it was intended for cars without EGR. I wouldn't know if that is a bad thing or not (i.e. putting a 561 in your EGR car).

In any case this is how one of my 561's failed. It ran the car OK, but at about 10 mpg... but in my case there were no trouble codes. My dad had one fail where gas milage was only slightly bad, performance was slightly slow, and the car wouldn't idle when cold. Other failures I read about on the board result in not being able to start the car. I have even heard of a couple cases where the car dies while driving. I think the lesson here is that they can fail in a variety of ways.

I enjoyed reading your post.

--
Matt L. -- Cary, NC -- '91 740 wagon

Matt, you hit right on the loose end that bothered me when the Mrs. asked me if I was ready to eat. I remembered finding a listing of applications for Bosch ECUs on a rebuilder's site and that the EGR was mentioned, so the loose end I left was checking the emissions label.

It is a 49 state car according to the plate and there's no EGR valve as I know them to be.

There was a nagging feeling I'd once before suspected the PO may have had that computer changed. Was it he or was it me that left the top screw out...

Do you know how the EGR is done on LH2.4? I have one on my k-jet car. Is it similar except perhaps a solenoid valve driven by the computer to gate vacuum to the valve? Nothing remotely like that on this car.

Thanks for the kind words and getting my loose end tied. And I suppose I should start looking for another spare.
--
Art Benstein near Baltimore

Well, I have seen the Bosch 572 in California cars and I think there is a 556 around for those too. But you want to get a 951 out of a 90s 240 or 700 and hang on to it.

Hi Bob,

I found where it was I saw the EGR reference that left me with an unsettled doubt. It was FCP's injection selection not a rebuilder's site. I see they list some of the LH3.1 parts too. Definitely worth searching the pick'n'pulls - heck those reman prices could turn me into an ebayer, except I wouldn't turn to my wife to give them the break-in test.
--
Art Benstein near Baltimore

Thanks Art, that straightens it out for me too. Now if I can remember those numbers and applications....

What part number(s) for the ECU are you looking for?

Ahh, this is the perfect excuse to wander the pick'n'pull some days when the sun comes out. The ECUs are cheap as compared with the AMMs, and sometimes I find a car before the ebayers do. I'll just remember not to get a -561 from a car with no collision damage;-)
--
Art Benstein near Baltimore

At the risk of re-openning the case, if you have the 561 you want to have the upgrade 951. Direct replacement, better reliability. There are not as many 561s around any more, due to the high failure rate.

Yes, I was looking for the 951 when I found the 561. The 561 was in a 92. I remember the car well, as it had been hit pretty good on the right front corner and for a pick'n'pull was about as late production 240 as one sees.

After tracing the circuitry to that resistor (R315) that cooked, I wonder if the 933 was used in M47 equipped cars as much as EGR cars. The resistor is in a path between an open collector driver on the ECU to the upshift lamp in the cluster, which in my automatic is the OD solenoid circuit- pin 26. Whatever the application, it has nothing to do with all the fuel it let loose.

Art: brilliant, classic diagnosis. Keep us posted as to the fate of the 933.

The "brilliance" was provided by my headband mounted flashlight, that for some reason, I received two identical gifts of recently. Glad to have taken seriously the overwhelming consensus among bricksters the 89's computer was gonna die someday. I've never been completely satisfied with the "replace the prom rebuild" touted as THE fix. I suppose there could have been a batch flaw with the outfit doing the mask programming, but you would think someone by now would have dumped a still good prom and offered a bunch of eprom copies for sale with or without code improvements. I think Matt L has the more realistic answer; the defects are various.

Thanks, Steve, I'll certainly give that resistor a look-see, which by the way is R315 not R15 as I posted.
--
Art Benstein near Baltimore