After the release of our Engine and Hydraulics kit, it seemed apt that I was asked how the kit can be applied to a faulty machine. What we have here is a Fiat Kobelco E135 tracked Excavator, fitted with an Iveco 4-cylinder diesel engine. There is nothing too fancy about the engine: it is an old-type geared industrial engine with a rotary pump and mechanical injectors. In terms of engine electrical equipment, there really isn’t much going on!
Something I’ve learnt is the access to technical information is very limited for plant machinery. You can purchase individual manuals but there is no central hub or datacenter for Plant machinery, which makes looking up anything about this particular machine rather tricky!
As always (but particularly in this case) the customer interview was vitally important. According to the operator, the machine had always been a little temperamental, but it really started acting up four months ago. The engine would stop suddenly, while the machine was being operated!
The trouble with hydraulics is if the prime mover stops for any reason then everything stops. For example if you are tracking across a site entrance and it cuts out halfway, there are only two choices: it starts again or it is pushed or dragged out the way. The machine was visited on a number of occasions by a mobile technician who traced the fault to a stop solenoid but also found some metal fragments in the high pressure pump. This was then sent away for an overhaul with the belief this was the cause of the problem with the stop solenoid. The stop solenoid failure would make sense due to limited electrical components on the vehicle and the fact there is no warning signs of anything playing up. A stop solenoid would definitely cause the symptoms the customer has been complaining of.
Solenoid changed and machine back to work but 4 weeks later the same thing happens again, engine cuts out all of sudden with no warning. After some more investigation work turns out the stop solenoid had failed again. Another one fitted and back up and running. Three to four weeks pass again and the same thing. In total the machine has had four stop solenoids in as many months and has got to a point where the customer is keeping a spare on site to replace the stop solenoid in order to keep the machine working. As far as I am aware stop solenoids are quality parts, plus supply for certain components is limited to OEM only. Due to the lack of information or on-board diagnostics found on plant machinery, I was asked if PicoScope could uncover anything that may be causing this premature failure of the stop solenoid. Whilst I love a challenge I wasn’t quite prepared for some of the difficulties I was going to encounter.
As I mentioned early the lack of information was quite surprising considering the plant industry is as big as it is. What I’ve also learnt is working on these machines is difficult to say the least. I guess the reasoning behind using older style rotary fuel pump and timing gears is because once the engine is installed, you really can’t get anywhere near it!! I had a letter box style gap to attempt to access the engine bay or the alternative was to go in from above. Either way this was not going to be easy.
Luckily in this instance the fuel pump is almost directly in line with the letter box which meant access to the stop solenoid was relatively straightforward. Also worth a mention here is the extra-long leads that are supplied with the Hydraulics kit. These became really handy on a machine of this size and I think those working in this sector would benefit from the extra maneuverability they give.
In order to assess what is going on with the machine I felt best to be as non-intrusive as possible due to the limited amount of information available. For this reason I started with a simple current measurement on the stop solenoid, and the battery voltage going from Key Off to Key On. All I wanted to see at this point was there was a change within the current for the stop solenoid to prove it was working.
- 24 V system requires the voltage range to be adjusted to ±50 V
- Battery voltage signal
- Rulers used to display measured voltage which is 27.06 V for battery
- Time ruler indicates ignition on
- Rise in current at stop solenoid to indicate activity from ignition on
From this I can be confident that current is flowing through the coil in the stop solenoid, and therefore will at least get the engine to start! So, time to start up.
One thing that stood out to me when starting the engine was a sporadic rise in battery voltage which was also seen in the current at the stop solenoid. Applying some scaling to the current meant this increase could be seen easier. The increase in battery voltage and current is seen in the electrical system and in this instance the voltage rises to just over 30 V. Looking at what could increase battery voltage on this particular machine, the only thing that has the ability to do this is the alternator.
With this in mind I felt it necessary to look at the battery voltage AC coupled and observe the alternator ripple for anything out of the ordinary. I also made use of the fact I have the option for four channels (when space is limited gaining as much information as possible in one go can be extremely helpful). I now have battery voltage AC /DC coupled, stop solenoid current and voltage connections. This will give an idea of what is happening at the stop solenoid when there is a slight increase in battery voltage. What I found was extremely interesting and something I’ve not seen before.
- Battery voltage AC coupled
- Battery voltage DC couple, Stop solenoid current and voltage
- Peak DC voltage at battery and at Stop solenoid
The great thing about looking at battery voltage AC coupled is we can see the ripple from the alternator in much greater detail. This provided me the information to determine things such as diode failure and voltage regulator faults. We can clearly see that there is something considerably wrong with the ripple from the alternator. It would appear that for around a second the voltage is almost completely unrectified and we see a sine wave typically associated with AC voltage. We would normally expect the AC ripple to hover around 0 volts but with a ripple effect that might swing from positive to negative, but never exceed 1 V. From the above capture we are set to ±1 V which when this wave occurs actually over-ranges. We also see a lot of sporadic spiking within the ripple as well which is also seen in the rest of the electrical system. When we see this swing in the ripple we also see an increase in the DC voltage of the system which can be seen at the stop solenoid and also in the current. Knowing that this behavior is not to be expected from the alternator it was strongly advised to get the alternator replaced, but what about the customer’s original complaint with the stop solenoid failing over a period of time? Could this overvoltage cause the solenoid to fail? Well, yes it can. Excessive voltage causes excessive coil current which overheats the coil within the solenoid and therefore burnout. There is both an increase in voltage and current when the AC ripple output changes, which would back this theory up, but would this be enough to burn the coil out? I wasn’t able to prove conclusively that the AC ripple was the cause of the failing stop solenoids as typically while I was with the machine it didn’t cut out. What would have been nice was to have seen the battery voltage increase to a point where the current at the stop solenoid was great enough to overheat the coil and burn it out but as always time was against us and the customer required an answer.
With this in mind the one thing we are confident of is the alternator signal is not correct and the alternator itself requires replacing before any further checks could be carried out. The customer agreed and once the repairs had been completed I returned to the machine to take the same readings I had with the old alternator. The results speak for themselves.
We can see that the sporadic spikes have now gone and apart from a little noise which could be down to a number of things, the signal is a lot clearer. Having left the engine running for a period of time, there were no signs of the ‘wave’ I was seeing on the original alternator and the voltage was stable at 28 V.
The machine has now been running for over a month with no reported issues thanks to continued feedback from the operator. I guess with anything intermittent it would be great to capture the exact moment when the fault occurs, but sometimes proving what it isn’t is just as important as proving what it is. I think what I have learnt here is that despite the lack of information, despite not being able to see most of the components , and no access to DTCs, sometimes starting at the very basic of tests can help give you some direction and understanding. In this case it was purely looking at battery voltage.