[Mighty306]

The ECU is actually one of the most robust parts of the car. It should be one of the last things you conclude is at fault.

Have you got anything you can plug in and read live data on? With the sensor pugged in (I assume it's a combined MAF & temperature sensor) it should be possible to see what temperature the air intake sensor is showing (usually via the ECU interpreting a resistance value from a thermistor). When the engine is completely cold this should correspond with the outside air temp.

Btw, measuring resistance on wires is often cited as a means of diagnosis in manuals (including manufacturer workshop manuals) but it can be misleading. A compromised wire/connection can show low resistance on a multimeter but not be capable of carrying the required load. It's much better to load test wires, which either means measuring voltage drop when they are in use (usually by back-probing the connectors) or applying an appropriate load by other means.

Even better is getting live data, & comparing it to a known good. With a fluctuating signal an oscilloscope would the the best way to achieve this. Some single-channel hand-held oscilloscopes are now ridiculously cheap. They won't be able to do everything but I've seen people grab things like injector pulse traces with one of these DollaTek DSO152 Portable Handheld Digital Oscilloscope 200KHz Bandwidth, 2.5MS/s High Sampling Rate with PWM Output : Amazon.co.uk: Business, Industry & Science. For those with a larger budget (£80-£120) you can get dual channel scopes with as much functionality as any home diy-er would need. I only mention this because it sounds like you're some way down the road in terms of learning how to diagnose electronic parts of the system.

 

[Mighty306]

OK, that's good that you've got Diagbox. You should be able to get a lot more information out of that. I'm not that familiar with it (as my Pugs are too old to use Diagbox). Others on here will definitely be able to advise though. It's definitely useful that you've included some more history as well. You've tried fitting a number of good quality sensors so I'd be inclined to agree that's unlikely to be where the problem lies.

I have followed quite a few expert diagnosticians online for a number of years now. They are always urging caution when testing wires & never rely on resistance measurements alone. This was mystery to me in the beginning too so I had to do some more digging to understand why.

A multimeter in resistance mode applies a pretty minimal current to the component being tested. (I don't know the exact value but I know it's probably only a few mA). The wire is analagous to a pipe. When intact it is capable of carrying a specified flow of electrons. Any restriction in that 'pipe' (compromise in the wire) may still be able to carry a minimal flow without showing up an issue. A multistrand wire that was only hanging on by one thread may be able to carry a few mA without any resistance. It probably couldn't carry the normal load expected of it. Equally, that damaged wire could happily show a normal 12.5v (or whatever voltage it's meant to show) without any load applied. I think that makes sense but I'd be more than happy to be corrected (have my understanding enhanced!) by any other contributors.

Here's a page on Auto Diagnostics Fluke using voltage drop testing. Diagnosing Voltage Drops Electrical Automotive Troubleshooting | Fluke .

The other option is to apply a substituted load, such as a lamp, which draws a current suitable for the size of wire. A conventional test light like this will draw a couple of hundred mA & would likely show an issue on a smaller wire by only lighting up dimly. Car Truck Voltage Circuit Tester Electrical Wire Circuit Tester Automotive Circuit Tester for Low Voltage Systems, Fuse, Switch, Wires : Amazon.co.uk: DIY & Tools

You should be able to safely test the harness between the ECU plug (ECU safely unplugged) & any sensor plug (sensor safely unplugged) with substituted loads, assuming there are no intermediate components that could be damaged.

There are times when using a multimeter to test resistance is appropriate. e.g. when measuring a component which inherently has resistance, such as an injector or stepper motor.

 

[Mighty306]

You wrote "The other thing that I'm thinking about trying is, since the power to the ECU is in connector A, and all the MAP sensor wires are connected to the ECU with connector B, maybe I can unplug the connector B from the ECU and measure voltage on pins of a live ECU, although I understand there may be risks of doing damage to the ECU. Let me know if this sounds like an option to you."

I'm guessing the ECU would throw so many codes if you did this that it will probably just confuse matters. I think it would be best in the first instance to see if Diagbox will tell you what what the live reading are from the sensor(s) & go from there if they're out of spec.

 

[Mighty306]

Apologies for the delay - I'm actually away at the moment. I don't really understand the reference relative pressure readings. The relative measured pressure makes more sense but really it is difficult to interpret without knowing what the car was doing at the time. With the ignition on and the engine off I think you would expect it to be zero (no difference from atmospheric pressure). I'm guessing in the first snapshot the 750mBar was constant idle (slight vacuum compared to atmospheric pressure as at idle you would expect about ~320mBar of vacuum to be created in the manifold).

Where it goes significantly negative I'm less sure about. Hopefully someone else can chip in?