Keeping the Hot Miata Engine Cool

Miata's twin-spark distributorless ignition system delivers twice more frequent sparks than other ignition systems. One of its disadvantages is to produce high combustion chamber temperature and thus temperature-spiking problem with the cooling system.

To try understand the problem further, let me point out the specification of Miata's cooling system:

Most of the time including 100F summer, Miata's cooling system is sufficient to keep coolant temperature below 93C (200F) where the cooling fan turns on/off to lower coolant temperature dramatically. Thanks "Fletcher Blades" from the miata mailing list for verifying it by sticking a temperature probe in the upper radiator hose and test drove in various conditions. However, people have been reporting temperature spiking under prolonged high-load condition. This is caused by the limited flow through the thermostat.

Miata's thermostat has two openings. The "sub" starts opening at 83.5C (182F) and can only lift 1.5mm (0.06") maximum. The "main" starts opening at 86.5C (188F) and can only lift 8mm (0.31") maximum at 100C(212F). It converts to a 1mm lift per 3F rise in temperature. Assume the equilibrium coolant temperature is at 195F. Then, a 7F increase in temperature can lift no greater than 3mm and will not provide enough coolant flow at high engine speed. It's not difficult to understand why the equilibrium temperature rises at prolonged high-load condition.

To remedy the problem, some people suggest a thermostat that starts to open 20F lower than the factory one. Other people argue that both 160F and 180F thermostats have already opened at the equilibrium 190-200F coolant temperature and therefore, would not make any difference. They would have been right if the thermostats only have on/off states like thermostats used in A/C. If an engine is not running in prolonged high-load condition, limited coolant flow through the thermostat at the equilibrium temperature is good enough to maintain the equilibrium. Therefore, 180F and 160F thermostats would NOT make a difference in the equilibrium temperature in that case. However, at the equilibrium temperature where both thermostats would have opened, a thermostat that opens 20F lower than the factory one will lift 6mm more than the 3mm lift calculated earlier. A 9mm lift obviously allows more coolant flow at high engine speed than a 3mm lift does.

Here is a proof to the factory thermostat flow problem. I had two track days in '96 during the 90F summer. During the 1st track day, the temperature gauge went up quite close to 'H' position on the track but returned back to normal on the way back home. On the 2nd track day, I ran the car without thermostat and the temperature gauge stayed at the normal operating temperature for the whole day.

Gathering information from the miata mailing list, I conclude that my cooling system is exactly normal like many other Miatas. One interesting thing is that people report their temperature gauge staying at 11:30 position while I've seen more than 10 Miatas all having their gauge staying at 12:30 position. According to the factory workshop manual, the drivetrain and cooling system for both US and Canada models are exactly the same. But I overlooked the discrepency in the "Body Electrical" section where US and Canada temperature gauges have different ranges. So the 11:30 position most people found corresponds to the 12:30 position I found in all the Miatas I know of. Again, "Fletcher Blades" further proved my hypothesis. His upper radiator hose temperature reading through a temperature probe has reached 197F yet his temp gauge has never climbed beyond 11:30 position. Since people having temp gauges staying at 11:30 position can achieve coolant temp of 197F like I do, I think they will also encounter temperature spiking problem as I do. If you haven't encountered that before, chances are you will if you need to push the engine harder.

So after all these observing, experimenting, thinking, posting, arguing, and so on, I decided to attack the temperature spiking problem by changing the heat range of the spark plugs, which is the source of the heat afterall. Mazda recommends using #6 spark plugs which is the only heat range I can get from dealers or auto part stores. I tried using the NGK platinum #7 plugs but the result is only marginal (See My Modification for more details).
 

Next: Site Map