see if anyone has already diagnosed something similar to this with Pico, I have a camon rail diesel engine that when put into operation it makes a noise like a knock, not affecting the operation of the equipment in any way and does not even show any code of faults, I did the real compression test and everything is ok, apparently it doesn't cause rapid damage to the engine as there are some tractors working with this noise for a while, not everyone has the noise some show it others don't, the manufacturer's staff have already dismantled it one and they didn't find anything abnormal visibly, do we have any tests with Pico that can help to understand this beat?
- engine data
- (205.57 KiB) Downloaded 103 times
- ARQUIVO-tractor with problem.psdata
- tractor with problem.psdata
- (5.84 MiB) Downloaded 88 times
- ARQUIVO-good tractor.psdata
- good tractor.psdata
- (6.1 MiB) Downloaded 89 times
- 20230819-0001-tractor with problem.psdata
- tractor with problem.psdata
- (653.99 KiB) Downloaded 83 times
- 230819_CompressionTest_001-tractor with problem.pddata
- tractor with problem.psdata
- (77.01 KiB) Downloaded 84 times
Regarding your psdata files, I can see we have cam and crank on channels A & B, could you clarify the signal on Channel C which I believe we can attribute to injection?
The knock when put into operation suggests Diesel knock when load is applied
Is the knock present from cold under load and do we have any smoke (black, white or blue) via the exhaust?
It would be ideal to have a capture on a single buffer where you have no knock off load, then knock present when load is applied. (500 ms/div with 2Ms/s)
I.e., Channel A Crank, Channel B Cam, Channel C #1 injector current and Channel D an accelerometer attached to the common rail (AC coupled)
If you could also make a recording using our NVH with a mic and accelerometer within the engine bay that would be awesome
If no NVH then a mobile phone recording of knock Off load / On load
Given some tractors exhibit this noise and some do not (assuming this is diesel knock) could we check for subtle differences in injection control/strategy
i.e., are injectors of the same brand, date stamp and software level between good and bad tractors.
Are good and bad tractors at identical emission regulation levels etc.
Could this be a "Market" specific characteristic (E.g. OK in UK but an issue in the US)
Fuel quality comes to mind too (Or fuel contamination)
As we have seen with diesel knock in light vehicle during “tip in” (light throttle load) it is often not a problem but a nuisance characteristic that maybe improved upon but not fully cured (hence no engine damage)
We will sit tight for any feedback
channel A speed Crankshaft channel B speed camshaft
C channel Camon rail rail valve sensor signal and ground signals
same as instructed in the GT904 test
It's a new tractor with only 592 hours, what you can see is that there are machines that have the milkshake and there are machines that don't, machines from the same year, same model, same standard of emissions, here in Brazil we don't use Adblue on these models, today I'm on it again I will collect data from the injectors as well and I will attach it here.
I will also try to make a measurement of it when it presents the noise in operation to see if anything is more evident, my equipment is 4 channels a 4425 I will collect.
Channel A= speed Crankshaft
Channel B speed Control valves
Channel C speed Rail Sensor
Channel Injectors the 6
I made this collection on the tractor that does not have the noise, it is attached in a little while, attached to the noise
- (28.26 MiB) Downloaded 81 times
File 02 tractor operating making noise
- 20230822-0001-file 01 stopped in slow rotation.psdata
- (25.91 MiB) Downloaded 82 times
https://drive.google.com/file/d/1ZSE2c5 ... sp=sharing
Just wanted to add a couple of thoughts. The first one being pump timing. I know these will be geared so there isn't any movement and being a new tractor with low hours, it's unlikely but not impossible.
Following on from the post where we looked at using the derivative maths function to look at the fuel rail pressure sensor - viewtopic.php?p=103861#p103861.
Using the same technique and applying to the files you sent first of the known good and the noisy tractor we have the following results. I've also applied a 10kHz lowpass filter to channel C and the maths channel formula I've used is as follows lowpass(derivative(C),100).
By including the phase rulers and setting out 720 degrees, I've picked out points in the math channel which appear to indicate the change in pressure as a result of the injector firing. Without injector current it's not precise to say that is when injection took place but making relative measurements with the known good and the bad machine we should be able to see if there is a difference between the two machines. I've then used the falling edge from the double pulse on the camshaft output as the signal to reference against and we have 49 degrees between the two signals.
Moving on the bad machine.
Adding in the rulers and making the measurement in the same place, the bad machine as a value of 42 degrees.
In my experience with these types engines is the gearing can't adjust like VVTi we find in automotive engines but I've not worked on something this new so I may be wrong. Please correct me if I am! If everything is fixed then there would appear to be a difference between these signals though which might be something to consider.
Looking at the other files you have added, in particular file 02 where the tractor is operating and making the noise, I've applied the same maths channel as used in the above images and found the following.
What is interesting here is that the maths channel highlights two points where the rate of change in the rail isn't as large as the others. Making an assumption (which we shouldn't really do) that cylinder 1 fires after the double pulse on the camshaft, using the firing order 1,5,3,6,2,4 from the data sheet you included it would appear to highlight fuelling on cylinders 1 and 6. You can also make out that cylinders 1 and 6 don't drop the rail pressure as much during activation.
I think there are a couple of things I would be looking to do next, especially if you have a known good machine to hand. Firstly, run a capture on the known good but the tractor under load as you've done in file 02. This will let us see if the waveform we are seeing in the bad one is normal or not.
I would also be thinking about swapping injectors 1 and 2 on the faulty machine. I know that this isn't always that easy on tractors given the exhaust system is usually mounted above the engine but hopefully without the full aftertreatment system we see with tier 4 and 6 engines it might not be too bad.
As Steve has mentioned, NVH here might also help with some direction especially if it can be done whilst capturing engine speed. From there we could see if it is related to the fuel pump or not. Software, also mentioned by Steve is a great shout and something to definitely consider.
Keep us posted and keep in mind that if, as Steve has said, this is more of a characteristic trying to fix it will not be easy and sometimes not possible.
yesterday I went to the tractor again and did a vibration analysis of the high pump gearing system, this tractor is using a CP3 pump that theoretically does not have a timing point with the engine, in the vibration analysis it showed a little play in the teeth of the gears from the Crankshaft with the high pump, attached to the images, I noticed in the vibration spectrum that there is a beat pattern that is when the pump pistons come into action every 120 degrees of rotation of the high pump, my doubt is that Could this turbulence in the rail system be the result of that? also the tractors that I collected are not identical, they use the same engine but one is a Puma 230 and the one with the noise is a Puma 215 both from Case Ih, now I'm going to try to get another tractor of the same model as the noisy one so that I can collect it from him also in the same situation that I collected from this other, any suggestions for collection? or I can collect the same as I collected from this one.