Diesel Engine Mechanical Testing Part 1
I have always been motivated to make more efficient and accurate practices part of my daily test routine. Nothing is more motivating than engine mechanical issues, I avoid tearing an engine down for testing purposes at all costs. Over the last year I have been testing various DSO methods that are part of my gasoline test routines on diesel engines. I have found even more motivation in the diesel world – engines can be tougher to tear apart, and parts are more expensive. I hope to share a series of posts and case studies to facilitate further discussion on these methods.
There are a few obstacles to overcome when testing diesel engines with a DSO. An “ignition” sync is harder to come by (especially mechanical injection) – the example below offers a solution. Access to test cylinder compression can greatly vary – some engines have easy to remove glow plugs and testing adapters are cheap. Other engines require disassembly including the removal of injectors and expensive adapters. Accessing wiring can also be difficult – high-voltage injector wires are sealed tight into harnesses with no individual wire access. Service information can also be lacking – DSOs can help understand the operation of a circuit when info is missing.
Diesel engines are behind relative to OBDII cars when it comes to diagnostic capability – both in scan tool features and coding ability. An example is this older 6.5L GM with mechanical injection. The engine had a 0.5 order vibration, which is a single-cylinder misfire the PCM was unable to consistently identify a cylinder in the contribution test. To easily identify misfire cylinders an exhaust waveform is gathered. The main obstacle on the truck was getting a sync. RPM gauges for mechanical diesels have a piezo sensor that clamps around the line. Every time fuel flows the piezo sensor detects a pulse. I have used various homemade sensors with my scope to get a sync. I have also found the Pico TA329 parking sensor tool works well when held against the fuel line.
In the exhaust waveform the red channel is the piezo sync and the blue is the exhaust waveform. Every time the injector fires you will see a disturbance that is close to TDC. When an overlay is used cylinders 1, 2, and 5 are identified as misfiring (vacuum pulses). Compression was checked on the engine, all 8 cylinders were above spec. More info on efficient ways to test compression will be coming up soon. This truck ultimately needed a pump – but I was able to consistently find misfiring cylinders. After swapping injectors the same cylinders were misfiring. The Flir also makes quick work confirming the DSO testing.
As I post the various parts please feel free to chime in with questions or tips.
Wow, nice work using multiple tools to find your problem, 2 questions, Are we looking at the exhaust temp with the flir? and with the NVH ( something i am very interested in ) where did you put the sensor to test for engine vibrations? I am familiar with use of the seat rail but have not had much success diagnosing engine vibrations with NVH. I put this down to not knowing where i should be putting the sensor.
I could have measured the temperature with the Flir, I just used it to look at the cylinders relative to each other. The temp reading on the screenshot isn't anything of significance since I didn't move it onto the manifold. The NVH sensor was at the seat track. It is hard to quantify what is a bad vibration, but 82 mg at the seat track is very significant and would definitely cause a complaint - this customer could not shift (manual trans) sometimes because of the vibration. You can always move the sensor to the engine but every engine will have normal vibrations. When you measure them at the seat track you are also testing the transfer path isolation integrity. HTH, let me know if you have more questions!
Well my mind is blown! I never would have thought to use the parking sensor tool to get a sync. Finding a way to get a sync on a diesel is something I've struggled with for a long time. Thank you for the awesome post!
I can't take credit for thinking about it... I would never have thought about it either. Someone at Pico suggested to me it could work for such applications.
Where can I get some those overlays?
Wow, good stuff Matt. I also have not tried to use the parking sensor on that. I look forward to trying that out for sure. Whats your though about using a piezo element such as this :
directly on the line itself. My friend Hans shared him using a few of those elements in various locations on the suspension of a vehicle connected to a DSO to help identify the location of a clunk similar to a chassis ear.
I would imagine a sync could be used with that as well.
I would like to put one of those piezo's on a long stick, so that I can reach in tighter spots on lines or for noises. I have a whole bag of them. Just haven't got to it yet. It is a pain to clamp a piezo in tight spots. Most everything I work on is common rail and my wyze probe works well to sync quickly off of those.
Yes, that will work too. I have one somewhere in the shop, I will try to find it and post a picture. It is in a plastic housing and I soldered a banana jack to each wire to easily hook up a scope lead. I even had a rubber band set up that would strap around a fuel line.
Wow Matt! Thank you! I too love the piezo sensor idea.
Thanks! I will share some more research using this tool. Beyond just testing for a sync I have started relating fuel rail vibrations to fuel flow. I have a another case study that will demonstrate this on a truck. I see some great possibilities for GDI cars, especially since you cannot swap injectors easily.
Thought this might be a good spot to show a comparison of the Park assist sensor vs Wyze probe vs Clamp vs Piezo.
I also use crank freq math and alternator speed for contribution and high pressure fuel rail sensors (vehicle equipped) to reason fuel injector/pump failures. I have no luck making sense of some of the high pressure sensor readings. I am still learning on others, but am able to make some sense of them.
Thanks! What is your favorite of the 3?
Probably the quickest would be the park assist sensor. So I believe that could be my "go to" for a cylinder ID.