I’ve been using the MX-5 almost as a daily for the last few months. It’s driving well but a couple of new problems are becoming apparent. Firstly, there’s some smoke on first start of the day and it’s using a lot of oil. Secondly, there’s a slight knock from the engine at certain RPMs but only under light load – when you press the accelerator, the knock disappears, then returns briefly as the revs drop. At idle or at a steady RPM when out of gear, the knock is absent.
The smoke issue could be piston rings or possibly valve stem clearance or seals (or maybe both). I did a compression test to see if there was any obvious problem with the rings and/or valves, results below:
DRY (BAR / PSI)
17.5 / 255
19.2 / 281
19.8 / 288
18.0 / 260
From the service manual:
17 Bar (249.9 psi) @ 300 rpm
12 Bar (174.5 psi) @ 300 rpm
2 Bar (28.5 psi)
Cylinder #1 is slightly outside the maximum difference but all cylinders have decent readings so I’m not concerned that there’s excessive wear or damage to the compression rings although the oil scraper rings could be clogged or excessively worn, allowing too much oil into the combustion chamber.
The spark plugs were all a bit oily and #4 was heavily encrusted in carbon / soot.
I’ve read that some people have had some success in freeing up or de-coking piston rings using Redex injected directly into the cylinders and left overnight. So I put about 5ml in each cylinder and will see how it turns out.
The engine knock could be worn big-end and/or crankshaft main bearings. One way to get a better idea of the issue without dismantling the engine is to check the oil pressure.
I’d already found out that the standard oil pressure gauge on the MX-5 NC is a fraud – it does not read oil pressure. The reading is generated by the ECU and simply varies in sync with RPM. There is an oil pressure switch so presumably the pressure gauge just doesn’t move if there’s no oil pressure. Also, the gauge does not respond until after the engine is running. Hence, there is nothing to catch the eye if you have an oil pressure issue – which may explain why so many NC cars have premature engine failure.
I can’t live with this phony gauge, so as an interim measure decided to fit an oil pressure warning light. From the incomplete wiring diagrams I was able to get from the internet, I was able to trace the oil pressure switch input going into the instrument cluster to the light green wire going to connector pin 1H:
I soldered a “Y” connection into the oil pressure switch wire and confirmed this was correct by cranking the engine and watching the signal change when the oil pressure came on.
The switch is normally closed and the “correct” way to wire it up would be to connect the anode of an LED to a +V supply and the cathode to the switch, hence the LED will stay on until the oil pressure builds up. I was going to take a +12V wire from a convenient ancillary circuit such as the radio or cigarette lighter but noticed that this circuit is switched off during cranking.
With the incomplete wiring diagrams I have, I couldn’t find an ignition-on wire and I didn’t fancy spending hours trying to trace one in the various looms. So ended up inverting the indicator so that it comes on when the oil pressure is on – the LED anode goes to the oil pressure signal and the cathode goes to ground. The internal pull-up in the instrument panel is sufficiently strong to illuminate the LED when the pressure switch opens. With a simple 3D printed housing I mounted the indicator on top of the steering column cover:
The indicator can give a vague idea of whether the oil pressure is “good” or “bad” by measuring the time it takes for the oil pressure to build-up during cranking and drop-off when the engine stops. On a “good” engine the oil pressure should build-up after just a couple of seconds of cranking and, when the engine is cold, should take 20 seconds or more to drop-off after the engine stops.
The MX-5 took more than 10 seconds of cranking before the indicator flickered on and goes off in just a few seconds when the engine stops. To me, this is a fairly good indicator that the oil pressure is poor, almost certainly due to excessively worn engine bearings. This explains the knock and also the clatter from the engine on first start that I mentioned in the first post.
In conclusion, the engine needs to be rebuilt. I suspect it needs, as a minimum, new crank bearings and new piston rings, but I’ll not know until I take it apart.