If you've read my first Caterham blog post Road Trip... in a Caterham? then you'll be aware that when I bought this car, we attempted to drive it 800 miles back home sight-unseen. However we only made it about 200 miles before encountering an overheating issue that ended with the last 600 miles being completed with the Caterham on a trailer. Ever since then I've been struggling to pinpoint the exact problem, and I've finally resolved it. The solution turned out to be extremely silly, and the entire ordeal is an excellent example of me (and my wallet) falling victim to Framing Bias, which would wind up drastically over-complicating my troubleshooting.
Initial Symptoms
Everything started during our road trip home. It was 90 degrees out, the Caterham was loaded with myself, the girlfriend, and several small bags of cargo, and we were hustling it through the Appalachian mountains. Suddenly I noticed that the temperature gauge was reading quite a bit higher than it had been. Initially I chalked it up to "old car things," and the fact that we were loaded with cargo, pushing it up a mountain at elevation. Over the course of the next two hours, though, the gauge would creep higher and higher, eventually nearing the 250F (120C) maximum on the temperature gauge. It was at this point that we arrived at our hotel for the evening, and after some basic troubleshooting using the tools I had at my disposal, I wound up throwing in the towel and having it towed home. Womp womp.
Throwing Parts at the Problem
After some Googling and discussions with friends, it was determined that the most likely culprit given the specific symptoms we had observed (including things like a lack of fluid flow through a specific return line that I didn't fully understand at the time), was a failed water pump. The good news was that the seller had included a spare water pump with the car, and all I needed to do was swap it! This wasn't too difficult, as the front timing cover is easily accessed once you remove the alternator belt. I've never messed with timing components on an engine before, but it turns out as long as you align everything it isn't all that difficult. One water pump later, and the temperature gauge was... still jumping up quite high.
Next was a replacement thermostat. I tested the old one in a pot of boiling water and it did seem to be working properly, but I just so happen to have another 1993 Caterham owner living just down the road for me, and he had a thermostat with a lower temperature setting of 167F (75C) instead of 197F (92C). This did seem to make the car run cooler initially, but I still wound up in situations where the temperature would spike; particularly while sitting idle in traffic. I also drilled a hole in the flange of the thermostat at the recommendation of... pretty much the entire internet... in order to ensure that an air lock wouldn't be able to form there since apparently this engine can be a bit stubborn when it comes to bleeding air from the coolant system.
It was at this point I considered that perhaps I was getting false readings, as I noticed sometimes the temperature would jump 20-30 degrees very quickly. As I was losing patience, I decided to simply replace the sender and the gauge in the car. Neither solved the issue.
I decided to bypass the radiator fan switch, thinking maybe that was the problem, but the engine still ran hot even with the fan on constantly. I then replaced the expansion tank cap, thinking maybe there was a slight leak that was preventing the coolant loop from fully-pressurizing. No dice.
Quickly running out of options, I decided to bite the bullet and replace the entire radiator and fan assembly. The new radiator from Radtec is quite a chonky unit; a 60mm aluminum radiator with a much larger 11" fan. It also has a fan switch set to turn on at 185F (85C) and off at 167F (75C). It was a perfect drop-in replacement for the original unit, and after refilling and burping everything, it proceeded to... not solve the issue.
Critical Thinking Time
At this point, the ENTIRE cooling system, and everything related to it, was brand new. This simple fact - that it likely wasn't a "hardware issue" with any part of the cooling system - seemed to be all my brain needed to know in order for it to start looking at the problem without the Framing Bias it had been applying to everything thus far. Maybe "hot" is a symptom, and not the problem! Almost immediately I noticed something I hadn't noticed before: the temperature gauge was only reading hot while the radiator fan was running!
Those of you familiar with older vehicles - especially weird old vehicles with... let's just call them "over-simplified" wiring setups - are probably shouting at your computer screen right now. "IT'S A GROUNDING ISSUE!!!" I am not well-versed in older vehicles, and grounding issues are not something I've ever really had to worry about before, apart from a loose battery terminal, or any aftermarket accessories I've wired in.
After futzing around with running dedicated wires from the temperature gauge straight to the sending unit and to the battery, which didn't improve anything, I noticed that the oil pressure reading was also being affected by the radiator fan state. This suggested a the grounding issue wasn't with any one thing, but possibly everything. So I decided to trace all of the main ground wires on the vehicle, and this is how I finally identified the problem.
The Problem, and the Stupid Simple Solution
As it turned out there was only one ground wire connected to the negative terminal of the battery. It went from the battery directly to the engine block. From there, a second grounding cable attached to that same engine block bolt snaked up under the carburetors... and when I pulled on it to try and see where it went... it just came out in my hands. The other end wasn't connected to anything!
Even worse, though, was that I couldn't find any wire grounding the chassis of the car to the battery. This is important, because most things on the car ground through the chassis. Things like - oh, I dunno - the gauges in the car! While the engine is connected to the chassis through engine mounts, those mounts have rubber isolators in them to reduce NVH, and so there is no direct path from engine to chassis. There are arbitrary ways for stuff to ground, such as across the throttle cable linkage, but needless to say, a proper ground was needed.
The Caterham groups explained to me that I should have a grounding strap connecting one of the bolts on the block-side of the engine mount, to one of the bolts on the chassis-side of the engine mount, which would complete the circuit between the chassis, the engine, and the battery. So I removed that second, going-nowhere ground cable from the engine, and relocated it to the engine mount to act as the chassis ground.
Too Cool?
Result: the gauge no longer so much as twitches when the radiator fan kicks on. I now see a much less worrying oil pressure reading of 60-70 psi instead of 30-40. And the temperature readings? Well, now the car refuses to exceed 180F (82C), fluctuating between 160F and 180F as the thermostat and fan cycle on and off.
I should probably be happy it's running so cool, and that might actually be the ideal setup for the car if I was planning on tracking the crap out of it, but it's probably a bit too cool for my use case. I'll likely swap in a higher-temperature thermostat now so that the engine isn't perpetually running cool, but that's easy-enough to do, and I'm just glad the problem is finally sorted. It's been 8 months since I bought this thing, and what I thought was an overheating issue had prevented me from getting any proper enjoyment out of the car other than extremely short trips around town. Now I can finally drive it without one eye on the temperature gauge, and am no longer limited to driving it only when I know I wont encounter any traffic. And yes, it turns out all of those parts were entirely unnecessary when the solution was just relocating a single wire already installed on the vehicle, but I'll just call it a bit of preventative maintenance, as everything I replaced was original to the car and over 30 years old. In fact, circling back to the beginning of this story, I think the original setup on the car probably wasn't able to keep up with our drive through the mountains. The 20-30 degree bump in reading on the gauge with the fan running, combined with an engine probably already running a bit hot, and slowly running hotter and hotter, perfectly explains how we wound up with the gauge reading close to 250F on that initial trip... even if realistically it was probably closer to 215F at most.
Time for some grand adventures!
