Eunos Roadster losing coolant

I own a 1990 Eunos Roadster 1.6 V Special. It’s a terrific car.
Unfortunately recently I hit a deep pothole on the front nearside.
Initially everything seemed fine, so I continued for about 20 miles.
The next day I drove for around 30 miles. Gradually the car started to feel rough, and then the temperature shot up to about ¾. I discovered steam rising from the radiator and liquid bubbling from it.
I got the car to the local garage the next morning. They found that the car had no coolant at all, and the engine was misfiring. They left the car for 2 days and the misfire cleared.
They then looked at the radiator and found there was a hole in it, probably caused by hitting the pothole.
They replaced my damaged radiator with a new independent one made by NRF. They used all the car’s original hoses and its original Mazdaspeed radiator cap, and then refilled with coolant.
I have done around 100 miles in short runs since then, and the car seems to be running OK.
However the coolant expansion tank level keeps dropping below Low.
Initially the garage thought there might be an air pocket, so they’ve refilled the expansion tank twice. I haven’t taken the car out since its second top up. There appears to be no leaks from the pipework and the expansion tank has stayed at the same level (about half-full) for 2 days.
The garage estimate that the car could be using ½ litre of coolant per 100-150 miles. It seems possible that the cylinder head gasket may have blown.

    1. I’ve read on the net that a new radiator cap can cure coolant loss. Is it worth a try?
    1. If so, should I look for a 1.3 bar radiator cap, like the Mazdaspeed one I’ve got, or would a 0.9 or 1.1 be OK?
    1. Should I go on a longer run (say 100 miles) to get everything well up to temperature and see if the coolant is continuing to drop as it did before?
    1. If it is head gasket failure, the garage is talking about skimming the cylinder head and replacing the water pump, timing belt, etc, simultaneously as this is good practice. That makes sense to me, but is it strictly necessary?
    1. The garage is talking about £750 ++ for the work. They can’t be certain on the price until the cylinder head is off and they can see what the state of play is and how much work is required. I’d like to retain the car’s original engine – and it seems to be running well – but could a new engine be a better option?
      Any advice much appreciated.

The cost of a cambelt and water pump is minor compared to the other costs. The Belt has to come off anyhow. Not worth not replacing.

Getting a replacement engine can be fraught. These engines are now 25-30 years old, you might end up with an engine out of an auto of a late detuned UK car, both of which will be down on power compared to what you have now. You put in a scrapyard engine (or rather, you pay your local garage to do that), there’s no way they will guarantee anything about that enigne, unless they are sourcing it for you. Properly reconditioned engines (not the scam recon engines) will be over £1000 plus fitting.

A new radiator cap is worth a try. I would only use a new genuine Mazda cap, even though they are more expensive.

If there is no leak from the cooling system and coolant is disappearing, It would definitely sound like a head gasket issue.
If this is not obvious to the garage and coolant loss continues probably a good idea to get the garage to do a leakdown test to confirm.
cambelt/waterpump change say £250 so head gasket work £500+
May be worth getting another quote if that’s what is wrong but assuming engine otherwise good worth considering.
It would cost similar to source a good engine and fit with associated risk.
Worry about this after head gasket breach confirmed. Leakdown test pressurises cylinders and sees if that pressure drops over time.

Any signs of “mayo” on the oil filler cap or dipstick yet?

Not at the moment. Will check again shortly.

Update: I have now done a further 160 miles, including one long run of about 120 miles. Engine running OK, temperature OK. No sign of mayo on the oil filler cap or dipstick. However coolant level has slightly dropped in the expansion tank, but it is still above the Low mark.
Another garage - a reputable MX5 specialist - believes the head gasket has gone, and has recommended a reconditioned engine at just under £1000, including fitting.
Advice much appreciated.

For £1000, it won’t be a reconditioned engine, unless reconditioned means a jetwash of the block. Has the garage conducted a sniff test?

You may be right about the second hand engine not being reconditioned. I’ll double-check with the MX5 specialist garage.
The local garage - who fitted the new radiator a few weeks back - has not conducted a sniff test or a leak down test.

For a £1000 with a used engine, make sure they are going a few seals with that.

As a little low cost upgrade, its worth replacing the flywheel with a Mk2 1.6 flywheel. For some reason, on Eunos Roadsters and US Miata 1.6s, Mazda put on a heavy flywheel. For the UK, and for all Mk2 1.6 cars, they switched to a significantly lighter flywheel, which coincidently has the same part number as an upgraded flywheel from Mazdaspeed. It makes a small but noticeable difference, the engine is a bit revier, and the flywheels cost next to nothing.

Update: good news. The MX5 specialist garage has recently done a leak detection dye test on the engine’s coolant system. No leaks found. They believe that the coolant loss was due to an air lock after the new radiator had been fitted. Although the system had been bled by my local garage after fitting the new radiator, a small air lock remained. It gradually sorted itself with the two top ups of coolant and running the car for around 200 miles.

1 Like

I many years ago had a similar problem and when I stopped the car to check for leaks I could not find anything. It turned out that the top rubber hose had a pin prick hole in it that only let water escape when the engine was running. Check your coolant pipes with the engine running so that the pressure is high enough to let the coolant squirt out of any small leaks.

Thank you. Will check this over the next few days…

Somewhat related:

  1. Coolant leak from the top of the radiator; there was a very fine hairline crack on the engine side. It only opened up when the engine was hot. The giveaway was the foam pad at the end of the radiator (not present on early cars) becoming soggy with antifreeze.
  2. Coolant leak following a cambelt change. I (foolishly) didn’t bother with a waterpump change at 60k miles, as it was “fine”. It seemed the new belt decided the pump wasn’t fine. Never any visible waterleaks at rest, no puddles, just loss of coolant.

I switched to Evans waterless a couple of years back; there is the attraction, to the scientist, of a more efficient engine (no cavitation at surfaces, cooler at sustained runs), but benefit that its kind to elderly coolant systems (doesn’t pressurise like water). Difficult to bleed though, as its much gloopier than water-based coolant.

A similar problem was raised on another forum. Car lost fluid on a regular basis and everyone moved to head gasket etc etc, all the expensive repairs. Turned out that there was a hairline crack in one of the hoses which leaked under pressure. The water, being hot, evaporated almost instantaneously so no trace when cool or stopped. Looking at the problem and troubleshooting I got to wonder if a UV tracing dye as used in drain testing would work. Leaks would show up under UV light so should be easy to spot the cause . Anyone tried this?

Well, one upside of using vividly coloured antifreeze is that, if you get that kind of leak, the coolant may evaporate before you get to inspect for leaks but it leaves brightly coloured marks behind. I used orange/pink OAT coolant for a while in my previous car and found when wiping under the front of the sump to check for any oil leaks the paper towel came back shocking pink - new water pump time.

I don’t use that coolant any more as pink started showing up everywhere and I became convinced it was part of the problem as much as the diagnosis. I have never got my head around the range of coolant chemistries and what is and isn’t safe to use in a Mk.1.

I find the misdiagnosis really concerning.

Can you imagine if you’d shelled all that money on engines, head gaskets and everything else… Only for it to be a bloody air lock?

I’d have serious questions/doubts for the people who provided those diagnoses.

1 Like