I had my 91 L98 automatic for about 20k miles (70k total) and when it started to run a little ragged I thought it needed a tuneup. Engine was surging at idle and at low speed/low load. I changed plugs, wires, cap, rotor, cleaned IAC, and reset timing to 8 btdc, but the car ran worse every time I touched it (no surprise there) until it would barely run. Thanks to Sam Lam and jfb on this Forum (and Helms) I ultimately learned 5 of 8 injectors were shorted (5 @ 5-9 ohms, 3 @ 16-18 ohms), and I installed a set of 24# Accels and things are WAY better now. I recently acquired an older OTC scan tool and was running through the diagnostics in my driveway, stopped, in Park, and now have a few questions that you experts may be able to help me with. At any rpm off idle (1200, 1500, 1800...) the engine has considerable misfire, but idles quite smoothly at around 750 rpm. Injector pulse width is about 1.5 ms at idle, goes to about 2.4 ms at 1800 but then goes back to about 1.5 when I hold 1800 rpm steady-state, again - stopped in Park. On the road, pulse width varies from about 1.5 ms at idle to over 6 ms under steady (but not WOT) acceleration, and when I downshift, then go WOT, then back off completely pulse width goes to 00.0 until the engine comes back to approximate idle. It seems normal for the injector pulse width to increase with an increase in engine speed, TPS voltage/angle (and it does just that on the road), but should the PW come back to an idle on-time at 1800 rpm? Am I seeing lean misfire? I've read that shorted injectors will toast the ECM; could that be the cause of the misfire? I may be experiencing a slight misfire on the road too, but the car makes good power (for a stock L98), and seems to run generally smooth (all ignition parts are new). The O2 sensor dithers between rich and lean as it should, but I noticed the BLM values were 118 in the driveway, and 004 on the road with integrator values around 118 under both conditions. Unfortunately, I don't have a clue what that means. I'm open to any and all opinions - thanks in advance, and sorry for all the wind.

 

FWIW, if you want to trouble-shoot the car yourself it's best to make the investment in a real recording
ALDL cable & software package (plus a laptop) so you can see/log more of what's really going on. AKM has
a reasonable price on DataMaster+cable - check the references on this page for his URL.

Since you have recently replaced the injectors, lean mis-fire seems like the least likely problem - at least
as far as the injectors go. In a car of that age the EGR valve will also need to be replaced more often than
not... that could cause some of the symptoms you describe.

If the car ran with faulty injectors for some time, I would worry more about burned valves. I would do
(or have done) a compression/leak-down test first, to see if anything major was damaged. Then I would
check the exhaust headers for back pressure - to see if the cats were melted or plugged. If those tests are
OK then I'd look further for the source of the misfires with the logging software. Some areas to look at
include:

If they are still original, the fuel pump/sock/filter/regulator diaphram all need checking; as well as the
wiring and connectors closest to engine heat.

The ECM injector drivers are current-limited and fuse-protected, so blowing one would not be easy. But,
if the ECM is original in your car it is a 13+ year old circuit board. It couldn't hurt to pick up a spare and
have it on hand - they don't last forever, especially if it spent some portion of its life sitting outdoors...

The pulse-widths you listed sound normal. Looking at the engine in the driveway only gets you info on the
idle timing - you need to log some data on the road to get a better idea of what the whole system is doing.

HTH

 

Thanks for the advice, Doc. I removed the EGR valve when I replaced the injectors, and it was fairly clean. Your point about burned valves is a good one (I was hoping against something serious), and I'll do a leakdown test this weekend if it ever stops raining. The car does spend its life outdoors and the ECM looks original, but I have no way of knowing what the former owners did or didn't do. The in-line fuel filter was replaced recently and didn't have any crud inside it that I could blow out (didn't cut it though). I drove this car from NY to Florida and back this April and it was flawless; then in 3 weeks it had 5 bad injectors. I'm assuming the injectors were failing gradually and finally reached a threshold where driveability was lost. GM had some major problems with injectors of that era when subjected to ethanol with a low pH. The acidic fuel would disolve the insulation on the coils and short the injectors. Knowing NY (and the rest of the country soon) replaced ether with ethanol in motor gasoline last November makes me think this could have been the cause of the problem. If that's so, then the fuel pump may already be in trouble if it's the original. Well, again thanks for the advice - you've given me a lot to chew on.

Joe Valentine

 

It is normal for the injector pulse time to increase during acceleration to a higher speed/rpm and then decrease to the normal pulse time once that speed/rpm is established. The reason the engine is able to maintain a higher than idle speed/rpm with the same injector pulse time is because; it is injecting more often (higher RPM), so it is using more fuel.

It doesn't seem right at first, but it is normal.

MIke

 

PW calculation is hard-coded into the ECM program. A general description was posted here about a year ago:

http://forums.corvetteforum.com/show...post1545625829

That one didn't mention DFCO or DE, but those are the fuel subtractions that lead to the 0.0 PW on deceleration.


Quite a bit of reading material is available if you want to dig deeper into the fuel loops for $8D:

The commented $8D code has been in the public domain for 10 or 12 years. A copy of the complete text file can
be found here, under 'Source Code & Hacks" :

http://www.moates.net/fileman


A short discussion of PW calculation by itself is going on here:

http://www.thirdgen.org/techbb2/show...hreadid=255308


A recent effort at a newly commented hack (with disassembler & compiler info) is here:

http://www.thirdgen.org/techbb2/show...hreadid=245659


Ken Canata posted a very good description of how GM code interpolates a 3-D data table here:

http://forums.corvetteforum.com/showthread.php?t=547494


There is a GM engineering document with part of the programming description for a Turbo Sunbird around somewhere -
but I can't lay my hands on the URL just now. The code is for a 727 ECM and is similar to the 8D. If you want the
file send me an e-mail.

HTH