Ford Powerstroke Diesel Forum banner
1 - 16 of 16 Posts

· Registered
Joined
·
10 Posts
Discussion Starter · #1 ·
I am having an issue with my 2005 F350 with about 150k miles. I had been using it maybe once or twice a week for the past couple of months. Last week I went to use it, it started fine, but after driving for about 100 ft it started making a really loud noise and having a rough idle (similar to a gasoline engine misfiring). So I drove back home and shut it off. About an hour later I started it back and had the rough idle. I let it warm up to see if it would still do it at operating temperature and it still did. I suspected it was a bad injector. I believe the previous owner replaced 3 of them and the rest are the original injectors that came with the truck from the factory. They also replaced the FICM and a few other sensors after the truck was stolen and several parts were removed.

The CEL was not on. I used Forscan to pull codes and it showed some glow plug codes, a couple of ICP codes and a cylinder 4 contribution code. I deleted the codes, and restarted the truck that day and few times in the next few days and the codes that came back are the glow plug codes and 3 ICP related codes. The cylinder 4 contribution code did not come back, but the engine is still having the rough idle. I have also ran the KOEO injector buzz test and I can hear all 8 injectors clicking each time I run it. However, if I turn on the engine, the issue is still there. So that leads me to believe it is not an electrical issue but a mechanical issue with one of the injectors. The FICM voltage stays at 48v when running, the HPOP psi reading is at 800 at idle, the IPR % is at 31 at idle. That leads me to believe it is not a FICM or IPR issue. I disconnected the starter wire and jumped it to the positive terminal and the engine cranks fine and consistently, which leads me to believe it is not an engine compression issue. A couple days ago I started the engine, and unplugged the ICP sensor and still ran rough, I did not see a noticeable difference in the way the engine ran.

These are the ICP related codes, but it did not make a difference when I unplugged it.

P2284 ICP sensor circuit range/performance
P2288 ICP too high
P2623 ICP regulator open

Please take a look at the video of the truck idling fine and then idling rough and let me know if you think it might be an injector (or more than one).


I don’t have $3200 to spend on replacing all injectors so would like to find which one is causing the issue and only replace that one. I have called around and none of the diesel shops offer a mobile service to come do a diagnosis. They have all asked to tow the truck to their shop or drive it there, but I do not want to risk driving it like this. I would like to replace the injector myself (this is my first diesel, but I am “mechanically inclined”) so I would like to know which injectors specifically need replacing so I can do the labor myself. From what I have read the best option is a power balance test using the IDS software. Is there an alternative to the IDS software? I read about AutoEnginuity, but it seems their power balance test is not as reliable as the IDS is. If I cannot find a place with the IDS software, what options do I have?

Is there a way to physically unplug each injector at a time and run the engine to see if it makes a difference, or can they only be unplugged after removing the valve covers?

Is the only other option to remove all 8 injectors and take them to get tested to see which one is bad and then replace? It seems like a lot of labor.

Any feedback would be greatly appreciated as this is my first diesel and am still learning a lot about it.
Here is a screenshot of TorquePro with the engine running rough.
 

Attachments

· Premium Member
Joined
·
4,727 Posts
I’ve owned IDS and AE, but not the ForScan laptop version. What is missing in AE and ForScan is the uncompensated injector balance test.

However without having IDS, you should be able to turn off individual injectors and still see how the engine reacts; the same or worse, which would tell you what injector is bad with either AE or Forscan laptop.

You also could do that manually by lifting the injector connector cir-clip or removing it entirely off each injectors connector and lift the connector off. Myself, I would probably do that without the engine running then the engine started to assess. I personally don’t like to interrupt the live circuit.
 

· Registered
Joined
·
2,272 Posts
I’ve owned IDS and AE, but not the ForScan laptop version. What is missing in AE and ForScan is the uncompensated injector balance test.

However without having IDS, you should be able to turn off individual injectors and still see how the engine reacts; the same or worse, which would tell you what injector is bad with either AE or Forscan laptop.

You also could do that manually by lifting the injector connector cir-clip or removing it entirely off each injectors connector and lift the connector off. Myself, I would probably do that without the engine running then the engine started to assess. I personally don’t like to interrupt the live circuit.
Yeah, only takes one time of doing that and hearing a zap to never do it again lol
 

· Premium Member
Joined
·
10,126 Posts
Have to get the PC version of ForScan to do the cylinder cut out test -- I do wish they would implement the compensation disable command (not sure why it is not there already)
you do not need the license for this test

Even without disabling compensation, you should be able to hear and notice the differences in fuel delivery, RPM, and IPR percentage

I see you posted a screenie of Torque, so use that adapter to connect to ForScan
 

· Premium Member
Joined
·
10,126 Posts
So, ...watched the vid
that is an IPR problem causing that
because the ICP sensor is giving a false reading

Also in your Torque pic
the ICPv showing 4.5 ?
with a pressure reading of 874

The PCM is making the psi and % numbers up
at 4.5 volts the pressure would be more like 3600 psi
 

· Compression Ignition Addict
Joined
·
11,975 Posts
EDIT - never mind, I read more carefully the second time. Plus 1 on possibly being the IPR, but check the ICP connector for oil and the pigtail wires for an electrical short.
 

· Registered
Joined
·
10 Posts
Discussion Starter · #7 ·
How would I test the IPR? Would the only way to discard that be by swapping it out for a new one? That screenshot was taken at idle while it was running rough if it makes any difference.
I assume the ICP is discarded since the engine still ran rough when I disconnected it correct? Otherwise it would have run better when disconnected if it was a bad ICP sensor? Or did I not understand that correctly?
I do have the PC/laptop version of Forscan. I have the license valid for a couple more weeks I think. Where do I need to go to be able to turn off the injectors one by one?
 

· Premium Member
Joined
·
10,126 Posts
You do not need the license to use the tool for troubleshooting, only need the license for programming

The voltage on the ICP pid is what the PCM uses to determine a "pressure reading" -- if that makes sense to you
that 4.5 volts is what the PCM is seeing as a signal from the ICP sensor
in this case the "pressure" reading is invalid -- it must be verified by using the voltage from the sensor -- there is a chart

that does not mean the sensor is bad, but it does mean the signal wire has too much voltage on it

Also the true oil pressure may not be represented by the voltage -- if the wires to the sensor are shorting
so, there may be random pressures depending on the severity of the short , and the PCM trying to "figure out" where to place the IPR valve

use the oscilloscope function of ForScan to watch for the voltage spikes on the ICP sensor -- even that is buffered some so you will not see all

Usually we see this when the harness has short a to the vRef (5 volt) wire -- also usually at the ICP connector or at a splice from a previous repair
 

· Registered
Joined
·
10 Posts
Discussion Starter · #10 · (Edited)
You do not need the license to use the tool for troubleshooting, only need the license for programming

The voltage on the ICP pid is what the PCM uses to determine a "pressure reading" -- if that makes sense to you
that 4.5 volts is what the PCM is seeing as a signal from the ICP sensor
in this case the "pressure" reading is invalid -- it must be verified by using the voltage from the sensor -- there is a chart

that does not mean the sensor is bad, but it does mean the signal wire has too much voltage on it

Also the true oil pressure may not be represented by the voltage -- if the wires to the sensor are shorting
so, there may be random pressures depending on the severity of the short , and the PCM trying to "figure out" where to place the IPR valve

use the oscilloscope function of ForScan to watch for the voltage spikes on the ICP sensor -- even that is buffered some so you will not see all

Usually we see this when the harness has short a to the vRef (5 volt) wire -- also usually at the ICP connector or at a splice from a previous repair
Would the issue be on the ICP connector and harness or the IPR connector and harness?

Here is a screenshot of TorquePro with the key on and engine off. The ICP voltage is at 0.3.
 

Attachments

· Premium Member
Joined
·
10,126 Posts
Start by looking at the icp connector
with it disconnected ForScan should show zero volts on the icp signal
if there is voltage, then you need to look for a short
 

· Compression Ignition Addict
Joined
·
11,975 Posts
Roll up your sleeves man and start checking things. Not like we haven't given you ideas, and over the internet we aren't going to solve it without your eyes, ears, tests etc.
 

· Registered
Joined
·
10 Posts
Discussion Starter · #13 ·
I plugged in the ODB2 plug and used TorquePro to check ICP voltage.
With the key on and engine off (KOEO) it reads 0.2-0.3v
If I then unplug the ICP with the KOEO, the voltage drops to zero.
I then started the truck with the ICP unplugged and it ran rough. I plugged it in and was still running rough at 875 psi on the HPOP. Then for a few seconds the reading went up to 3600 psi and the engine ran smoothly with the ICP voltage around 3.5. then the psi dropped back to 875, with ICP voltage still at 3.5 and it started running rough.
I shut down the engine, and then turned it back on about a min later. The psi was up around 4000 and engine ran pretty smooth, then the PSI dropped back down to 875 psi and the noise and rough idle started.
I posted some screenshots below.
Why would the HPOP psi only go up to 3600-4000 for a few seconds when the ICP voltage stays between 3.5-4.5v. Could it be that more than an issue with the harness, the issue is with the HPOP pressure being low? If that is the case, could it be a bad IPR valve or a bad HPOP? This is a 2005, so from what I remember reading, the HPOP does not fail too commonly.
 

Attachments

· Compression Ignition Addict
Joined
·
11,975 Posts
If all of this is at idle, then you LIKELY have a short in the wiring or the connector or the sensor, OR the IPR valve is sticking.

There is no reason for the pressure to be so high that a working ICP sensor would read voltage high enough for the PCM to report over say 1500 psi at idle (ie ICP output of 2.0-2.5 volts or higher).

In fact, you would have to be at WOT or towing hard to get the HPOP output up to 3600+ psi (or the IPR valve would have to be stuck closed). 3600 psi would require the ICP sensor volts to be around 4.5.

The PCM sees the voltage that the ICP sensor puts out and converts it to a pressure reading according to a calibration curve (see link below). If something isn't right with the voltage, then the PCM will usually sense it and then put out an inferred ICP reading value to the ICP pressure "PID".

Another useful PID to monitor in conjunction with the others is ICP desired. This shows IF the PCM really does want the pressure to be high for some unknown reason.

Attached below is a table and a graph of the relationship between ICP sensor voltage output and pressure:

 

Attachments

· Registered
Joined
·
10 Posts
Discussion Starter · #15 ·
I think the issue was fixed. I was going to replace the IPR and disconnected the connector. It did not look stock so I traced it back and saw where it had been spliced into the harness. (Background: This truck was stolen and the previous owner said he had to buy a few sensors to get it back running before he sold it.)
So I examined the wires and one of them had a small part where the copper was exposed about 3 inches back from the connector. So I covered it back up with electrical tape and struggled for about 20 min until I was able to plug it back in (I had not removed any parts from the engine at this point). I fired up the truck and it idled fine. I let it idle for about 10 min and it did not hesitate at all. I then took it around the block and did not have any issues.
It's amazing that something as simple as an 1/8 in of exposed wire in the right spot could make the engine run so badly. I assume maybe the wire came in contact with either something hot or rubbed against a pointy/sharp part and was affecting the reading it was getting from the ECU to the IPR. I will double check again and insulate the wiring. Here are some screenshots of TorquePro while idling fine. The ICP voltage now seems to be in line with the HPOP psi readings.
Thanks everyone for the suggestions, I would not have though of checking the relationship between the ICP and IPR. I was going straight for a bad injector. The CEL is on for bad voltages at ICP, I will clear them and see if they come back.
 

Attachments

1 - 16 of 16 Posts
Top