Talk Morgan Forums Thread Like Summary - Temperature readings: Diagnostic help please!

Bit of a long one this chaps, but hope you can stick with it and offer any words of advice?

A few days ago, I popped out in the Morgan for a short but vigorous drive, only a few miles (less than 4 to be specific). When I got home I noticed that the engine fan had kicked in, something that heretofore I had observed only very rarely. The water temp gauge on the dash hadn't moved at all.

So I let the engine go cold overnight, and the following day started it up with the computer plugged in to see what was going on. Remember my car is thoroughly non-standard. The engine fan is controlled by the ECU, the water temperature gauge in the cockpit is fed from a temp sensor (taken from the original CVH engine, where it sits in the block) which is sited in the water return feed pipe from the top of the radiator, immediately before the pipe enters the thermostat block on the engine.

Prior to starting the engine I removed this sensor, it's an old-fashioned copper stubby thing that just protrudes into the water flow and changes resistance as it warms up. I stuck it in a pan of water on the stove with an ohm-meter attached, and observed that, as expected, the resistance dropped as it got hotter. So I assumed it was working ok and put it back.

The engine performed entirely normally, the water circulated round the block until the ECU temp passed 100C, then the thermostat opened, water started circulating through the radiator, the dash gauge started measuring the temperature of the now-circulating water as it heated up, the engine fan slowed down as the head temperature stayed steady on 100C.

So I assumed that everything was fine and that what I'd observed on my short drive earlier was the engine not getting hot enough to open the thermostat, and worried no more about it - until today.

Today, I went for a 30-mile jaunt. The engine performed entirely normally - if anything, it was going better than usual. But the temperature gauge rose relatively quickly to indicate around 100 to 110C and stayed there. When I got home I got the infra-red thermometer out and did a bit of measuring.

Temperature of the head itself was as expected, around 110C. Temperature of the water passing the sensor was around 80C. Everything checked out fine - apart from the flipping temperature gauge. The car's been in this configuration since 2008, I've never had any issues with it at all, and typically the temp gauge reads much lower than this, I wouldn't have expected it to show more than 90C. Nothing has been done to the car over winter apart from fitting new rear brake cylinders and changing the brake fluid.

So, what do we reckon? Faulty gauge? Faulty earth on the gauge? Faulty voltage regulator (but the other electric gauges work fine), or faulty temp sensor? Obvious thing is to start with getting a new temp sensor as it's the original from the 1986 CVH, but I'd appreciate any and all ideas.

Hamwich, I think that's good news, looks like you've almost cracked it. Thermostats can when starting to fail behave in strange ways often failing to respond to the correct temperature spec but also present with symptoms like they are becoming tired and after a dozen or so cycles the range of operation reduces until it fails. Now technically they are designed to fail in the open position but can't be relied upon to do so. I think yours is in the tired stage and isn't opening enough so the water temp has a slow constant rise until it registers as too damn hot! Mind you it'd done well and had a long life!