1994 6.5TD - Code 36 - Injection Pulse Width Error (Response Time Long)

[Iain]

I keep springing a "code 36 - Injection Pulse Width Error (Response Time Long)". This only occurs at 80-100 kph (50-62 mph) at cruising/light throttle conditions. If I let off the throttle or press it down the service engine light goes away.

Truck has about 340,000 km (211,000 miles). About 140,000 km's (about 87,000 miles) on the injection pump which was last changed 3.5 years ago. I am using an FSD Cooler mounted on the intake and have replaced the driver last year.

Truck starts fine, seems a little low on power but will still blow some black out the tail pipe if I stomp on it. Vacuum pump and turbo side of things are o.k.

Is this a precursor to having the pump replaced or would this code 36 be related to the FSD starting to pack it in again?

Thanks in advance for your responses

[Robyn]

If the pump has all those miles on it I would be suspect of the pump but the FSD can cause issues like this to pop up.
Is your FSD/PMD remote mounted.
If the pmd is still onboard the pump, get a fesh one and remote mount it on a heat sink down on the inside of the front skid pan which is a much cooler place.

Let us know

Robyn

[gmctd]

Two things you can do as eliminative -

Remove the FSD module from the heatsink - remove and pitch the plastic covers on the transistor drivers - loosen the four nuts, run each down finger tite - then 1/4-turn more with a nut-driver - this restores electrical continuity amongst the stainless steel hardware - no more than 1/4-turn or you'll shatter the ceramic hermetic seals, ruining the transistor(s) - install the module to the 'sink

Remove the black ground wire from the 'sink, and reattach it to the Ip where it belongs - that ensures Injection pulsewidth can meet correct timing specs.

Clean the module\'sink with lacquer thinner, then run a bead of RTV around the juncture where the module meets the 'sink - that will weather-proof the assembly for the next step.

Remove the 'sink from the intake manifold and remote mount it outside the engine bay - high engine-bay temps and heat-soak is the actual killer.

Well, that's four or five things, so sue me..............;)

[Iain]

Is your FSD/PMD remote mounted.
If the pmd is still onboard the pump, get a fesh one and remote mount it on a heat sink down on the inside of the front skid pan which is a much cooler place.


I have the FSD/PMD mounted on an "FSD Cooler" (black unit with cooler fins on it to describe it without being technical!) bolted to the 2 left side intake bolts. If I mounted it on the inside of the front skid pan, 1. does it matter if the front crank seal leaks oil on it and 2. will it matter if it gets wet down there (rain/snow/ice/mud/etc)?

Iain

[Iain]

Remove the black ground wire from the 'sink, and reattach it to the Ip where it belongs - that ensures Injection pulsewidth can meet correct timing specs.


This sounds interesting.... O.K., when the last injection pump was installed, I removed the FSD from the pump and mounted to the FSD Cooler bolted to the intake. I do not remember disconnecting a ground wire as it is a plug in connection for the FSD. Are you saying I sould be running a ground wire "jumper" from the FSD Cooler to the pump?

Iain

[gmctd]

Usually, the 'sink vendor suggests to unattach the black ground wire from the top of the IP and attach it to the 'sink - which, of course, is anodized, and which is an excellent electrical insulator.

If your IP is the oem unit, you may notice that black ground wire has a generous coating of some pink or purple goop on the screw\terminal - that is sealer to prevent cathodic deterioration due to the steel screw thru the copper terminal on the aluminum housing, aided and abetted by oxidation.

The insulated 'sink is then attached to the semi-insulated intake manifold with little prep for electrical conductivity - the ensuing cathodic deterioration and normal oxidation results in wierd stalling\hesitating\bucking symptoms, usually resulting in the owner\installer joining one or more of the GM Diesel forums, desperately seeking susan.....er, help - some, even after replacing the IP did not cure the symptoms.

If you did not disturb that ground, then try the chromium-oxide burnishing thing with the FSD driver nuts - worked for me.

Try it - you'll like it....................

[moondoggie]

Good Day!

If nothing else works, you might want to read a_topic_I_started_a_long_time_ago (http://www.thedieselpageforums.com/tdpforum/showthread.php?t=22757). DTC35 & 36 are closely related.

Good Luck & Blessings!

[Iain]

Just an update,

-Checked grounds on frame (RHS) and cleaned.
-Ground at back of head, impossible to get at unless you are a monkey, but it appears tight and the strap is good and not broken.
-Ground at cab appears o.k. as well but due to A/C lines, again you have to be a contortionist to get to it.
-Check and cleaned batteries, cleaned terminals, all o.k.
-Ground on top of IP is good.
-FSD connections are good as well.
-As a side note, installed boost and pyro gauges as well. Boost is about 7 psi but will spike to 11 psi till the computer catches it and corrects it back to 7.
-Still same problem of the code 36.

Next steps:
1. Replace the existing FSD with a known spare and see if problem goes away.
2. Try grounding the FSD Cooler to a solid ground as it is not attached to the pump anymore. As mentioned in an earlier post above, the FSD Cooler is aluminum and that metal is a poor ground. Perhaps just dropping a ground wire from the FSD Cooler bolts to the battery ground on the manifold might help.
3. Scratch my head and crack open the piggy bank for a new pump....

Thanks for all the tips and I will keep you all posted.
Iain

[gmctd]

Sorry, neglected to address this - not necessary for the 'sink to be grounded - bolting it to a metal surface is sufficient.

The ground on the IP is for the Fuel Solenoid and the FSD, which is required.

You can run a short ground from the 'sink mounting bolts to a nearby metal surface - won't cause any problems.

DTC36 could be the FS out of adjustment due to mileage, or long due to increased viscosity fuel - are you using good ole #2 ULSD, WVO, or Bio-D, mixture?

Be really helpful to get a scantool on it

BTW - remove the grounds at the back of the intake manifold, make sure they are really clean, not just 'looks like clean' - they are very important, being PCM control grounds

If you resolve these several things and still the 36, it may be IP time - again, a scantool would be most helpful as final decision tool.

[Iain]

DTC36 could be the FS out of adjustment due to mileage, or long due to increased viscosity fuel - are you using good ole #2 ULSD, WVO, or Bio-D, mixture?

Just regular old #2 diesel (summer) right now. In the winter it is changed to #1 (winter) version which has less viscosity than the summer stuff. I have been using a fuel conditioner as well, but there are a lot of miles on the pump.


As an update, I have tried a known good FSD, still throws the code 36.
I will try running a ground wire from the cooler to one of the intake studs and just see what happens. And I will try to get the rear head ground off, I think I will get some bandaids first before I go back there!

Thanks guys!
Iain

[gmctd]

Also, put a DVM on the ESO wires, check for the same voltage as directly across the battery terminals at idle - low voltage causes increased pulsewidth.

If you have a laptop pc, Swedeburb's GMTDScan Tech is a top quality OBD1 scantool, and very inexpensive - check it out at Engh Motors, a vendor here.

[damork]

By all means move the pump driver off the engine to a cooler location. Mines in front of the driver side battery on a Beta cooler.

I've been getting code 36 for the last 2 years when temps are under 35-40F, and from everything I've experienced, it seems to have as much to do with physical pump hardware as it does with good electronic connections - supply power and ground.

Stanadyne acknowledged to me back in 2001 they had returned pumps that were evaluated which indicated some debris in the driver solenoid that absolutely could have contributed to the problem. At the time, I added additional filtration to my truck (pre-lift pump) and now with 135,000 miles on the pump with a huge reduction in codes. Any codes I have now I believe are more to do with wiring than contamination, I have over 250k miles now and it still started fine down to 0F last winter so I’ll live with code 36 until something stops me.

The latest change I made on my truck was a new heater motor where I discovered a very bad ground that melted with the power supply wire and blew the 50A fuse. It also revealed the ground on the heater was a weak link, so I tied the ground to the main instrument panel beam and polished up the heater ground connections. Still too early to tell, but the truck was hiccupping from time to time and this seemed to really smooth things out. This ground also attaches to the same post on the passenger side head that the ECM ground attaches to. The whole truck is getting more relays to help unload the ignition switch and an improved grounding setup.

[moondoggie]

Good Day!

"...put a DVM on the ESO wires, check for the same voltage as directly across the battery terminals..." An alternative is to connect one DVM lead to the battery terminal & the other to the corresponding wire on, in this case, the ESO. This will tell you the DIFFERENCE between the two voltages. This has the advantage of being a live reading. While both B+ & voltage at the ESO at idle are probably stable enough to take separate readings, measuring this way takes out any doubt - you are measuring the difference between, say, B+ & the power to the ESO live.

Be careful with this, however. There will ALWAYS be a difference between the two voltages; 0.1V or so probably isn't a problem, 0.5 or 1.0 is almost certainly a problem.

On devices with a wired ground, this is equally effective for checking the quality of the ground wire & its connections.

Checking a battery ground wire can be done this way, especially if you have top-terminal or top- & side-terminal batteries. One DVM lead goes to the top terminal on the battery (NOT the ground wire clamp) & the other to the block. Then disconnect one end of the other ground wire, so all current is going through the one you're checking. Cycle the glowplugs & take a look at the voltage drop - while the glowplugs are on, there's ~ 150A going through this ground wire. This has an added advantage: without disturbing anything, you can see if you have a problem or not. If you do, you can then verify that it was fixed. This, instead of cleaning connections that may be fine, & guessing as to whether you've actually fixed anything or not.


Blessings!
(signature in previous post)