Rail Pressure at Idle - Data Log of Demanded vs. Actual Pressure

[eastmentCTD600]

I've been using my Smarty Touch to data log some operating parameters. Of concern is my rail pressure at idle. Truck idles fine and runs good but has a slight knock, as in timing knock sound, at idle. Its minute but present.

When I shift from park to drive the Actual Rail Pressure (red trace) appears to have large fluctuations around my Demanded Rail Pressure (blue trace). I had a spare FCA so I installed it; however, I'm still getting about the same results with regard to engine sound (knock) and actual rail pressure fluctuation.

Demanded rail pressure is about 7,500 psi in drive and the actual pressure appears to vary +/-400 psi (800 psi peak-to-peak) around the demanded value. If I calculate the average rail pressure from the red waveform its close to 7,500 psi, so yes, the average pressure is meeting the demanded but it seems odd there are such large fluctuations...
-----------------------------------------------------------------------------------------------------------------
Excel plot of original FCA. This plot is with the new parts listed below except not a new FCA.

-----------------------------------------------------------------------------------------------------------------
Excel plot with new FCA. This plot is with the new parts listed below.

-----------------------------------------------------------------------------------------------------------------

Questions:

1. Does anyone else w/Smarty Touch have (or could make) a data log their demanded vs. actual rail pressure at idle in Park then shifting into Drive? I would really appreciate some feedback...
2. Does anyone know if the ECM has a 'learning function' with regard to the FCA control loop? If so this may improve over time???
3. If this is abnormal where do I look next?

Notes:

• New FCA (Bosch)
• New rail pressure sensor (Bosch)
• New crank position sensor (Cummins)
• Smarty Program: SW#2, TQ:1, RP:0, Timing: Stock

 

[AH64ID]

I don’t have a Touch/MM3 anymore, but the rail pressure fluctuations you’re seeing are normal.

The ECM is not learning. It operates the FCA with a PID controller. It can be adjusted with custom tuning, but the fluctuations are still there.

As for the “timing” rattle that’s normal too. Only way I cured it was thru custom programming.

 

[cerberusiam]

Looks like a typical smoothed PWM signal. The FCA doesn't have a direct effect on the injection event so it can vary a lot. The ECM controls injection event with duration on the read pressure and it is dynamic. The slight knock could be an injector off just a little due to wear or setup or carbon build up on the nozzles or just the typical injector rattle of the deep spray systems. Do you tow at all? Very often? Usually a good hard run with a heavy load will clean up the cylinders and nozzles while reducing a lot of the injection rattle.

 

[eastmentCTD600]

Looks like a typical smoothed PWM signal. The FCA doesn't have a direct effect on the injection event so it can vary a lot. The ECM controls injection event with duration on the read pressure and it is dynamic. The slight knock could be an injector off just a little due to wear or setup or carbon build up on the nozzles or just the typical injector rattle of the deep spray systems. Do you tow at all? Very often? Usually a good hard run with a heavy load will clean up the cylinders and nozzles while reducing a lot of the injection rattle.

Thanks for your input cerb
Injectors are remanufactured BBI stage 0.5. They are the 03-04 style (143 degree spray angle 8-hole) in a 04.5 application. Installed last month. When I tow it's usually hooked to about 10k periodically in the summer. Haven't towed with this setup yet.

I don’t have a Touch/MM3 anymore, but the rail pressure fluctuations you’re seeing are normal.

The ECM is not learning. It operates the FCA with a PID controller. It can be adjusted with custom tuning, but the fluctuations are still there.

As for the “timing” rattle that’s normal too. Only way I cured it was thru custom programming.

AH64ID,
I wasn't sure if the PID loop has any capacity for dynamic gain adjustment, or learning, depending on engine rpm (or other parameters). That would be rather sophisticated. Looks like too much gain in the control loop for this operating point, too much gain usually leads to instability. If the gain is static then I wouldn't want it adjusted unless there was instability in other operating regions.... If no capacity for dynamic gain control (learning) then it is what it is... Thank you for your response.

Got a source for custom tuning?