Join Date: Mar 2008
Total Cats: 1
ebay radiator, parallel fans, thermostat, ducting
I had to do a good bit of separate searching when I changed my three year old OEM auto transmission radiator out and thought it might be useful to put my experience here for easy search results.
I installed a FM no-electronics kit last December. As I live in a region that can see consecutive 100 degree days, my plan was to watch my cooling as weather began to warm. I was having zero cooling issues prior to this spring.
Prior to the first eighty degree day, I did some pulls and logged my temps for reference. I would typically cruise between 197 and 208, and the fan would pull that down pretty fast over 205. Ambient temps were around sixty-five degrees. I was already running the FM 16psi radiator cap. For myself, my goal is to see no higher than 215 degrees temps this summer, and if that is reached, cooling back down in short order once the heat inducing load is removed.
To be honest, I have the mod-bug. I went ahead and purchased the ebay radiator before tracking any serious out of range temps. Once the radiator was installed and ducted, I was surprised by the impact on my coolant temps. They went up.
The temps were not outside the normal range, but the first eighty degree day, they were up about 3-4 degrees. Typically, under 55 mph was ok, over 55 the temps showed the increase. Idling was building up temps pretty fast as well. Adding heat to the system via boosting was also increasing my temps. I was not happy. I had spent a good part of a day incorporating the stock fans at stock positions, and working up the ducting. I had burped the radiator a couple of times. Fresh 75/25 water/coolant mix with Redline water wetter to boot.
My theory was that the engine bay was filling with air that was not being pulled out. I fly large scale RC airplanes using 80cc plus size two-stroke engines. Our rule of thumb is you must have 2.5 times the cowling outlet area relative to the cowling inlet area. Seemed reasonable. I'd compared my ducting to what was considered good and I thought mine was relatively equivalent to what I'd seen. Research had also shown louvers/additional ducting/aero mods that folks had used to combat this particular type problem. I believe this theory is later proven wrong.
I had replaced my OEM thermostat three years ago with another MAZDA factory unit, and as it had been working fine on the stock radiator just before the ebay went in, I generally trusted it. Still, my plan of attack was to address items one at a time, beginning with cheap/easy actions first and then if that didn't work, ease into the difficult stuff. The list was to try parallel fan operation, new thermostat, further seal ducting with spray foam sealant, modify splash tray to pull more air out and finally test ebay radiator flow.
Here is how that worked out.
Parallel fan operation was fairly easy for my 2000 using the stock computer.
I just put a jumper between the AC fan signal wire over to the wire for the cooling fan signal wire. That way when the main fan signal drops to ground, the AC fan relay also kicks in and runs in parallel. The same is true for when the AC signal drops to ground, the cooling system fan kicks on. Each relay still handles its own electrical load. The only down, is that the computer is not expecting the increase in current draw when the second fan kicks in, so you may get a momentary shudder at idle speeds. These are the wires for the 2000 Miata...
Fan motor red/green to (1R on wiring harness) jumper-ed to
condenser fan L/W (blue with white) to 1I on wiring harness.
This change was never intended to be a fix, rather an easily reversible patch to help drop temps faster as they climbed into the higher range. This change certainly shorted the time required to back my temps over 205 degrees back down.
The next item on the list was the thermostat. As there is a lot of dissension regarding MAZDA OEM, vs NAPA vs Stant vs SuperStant, I went into this one without a great deal of confidence. My MAZDA OEM unit had appeared to have been functioning properly prior to the radiator swap. After viewing Joe Perez's Stant vs SuperStant video (and an additional once or twice for the entertainment value, great stuff!) I settled on the Stant 180 degree unit from ADVANCE (#15868), as the NAPA was out of stock on their 180 degree unit. I wasn't expecting much, as the openings in the MAZDA unit I had were larger, and the comparison in apparent engineering behind the two seemed to put the MAZDA OEM unit ahead. I was wrong again...This change brought my temps down to be on par with other ebay radiator threads I'd researched. 75 mph it sits rock solid at 185 degrees. Warmup distance was increased by about 1/4 mile. At idle, eventually the fans are needed, but it takes very little time to bring the temps down. Runs with boost increase the temps in the system, but those are easily brought back down by either slowing down, or if the temps get over 205, the fans knock them right down.
I am not positive, but my idea is that the MAZDA OEM thermostat is engineered to flow a certain rate of coolant proportionate to the size of the OEM radiator. The new ebay radiator was larger, and the coolant was not getting out of the radiator and back into the engine quickly enough to allow it to do its job cooling the engine. I welcome other’s thoughts on this. Each of these changes were made singly, and the new thermostat had a huge difference in the performance of the ebay radiator. Since I have a 2000, I am just a bit concerned running below 195 degrees, but so far the car seems to be running fine.
Last edited by fwman1; 05-12-2013 at 08:04 PM.
Reason: bad link