Assuming the E-fan wiring isn't buggered up anywhere, with the switch currently plumbed into the BAC line, it's possible that you still have lots of air trapped in the cooling system, and coolant isn't circulating thru that line. If the switch sensor is sitting in an air pocket, it won't work. I used to have an aftermarket water temp gauge sensor plumbed into my BAC line, and it worked fine, but whenever I changed coolant, it took lots of air burping of the system to get that gauge to work right.

If you want to relocate the switch, what I would do is remove the OEM 207* temp switch for the aux. fan and use that location on the back of the WP housing for your new switch. The OEM wiring runs the aux. pusher fan anytime the A/C compressor is running anyway, so you won't loose that functionality. And since you now have a main e-fan that is supposed to kick on at a lower temp than 207*F, the switch becomes redundant.

 

I'm leaning towards I've messed up the wiring somewhere. Coolant is leaking out of both sides of the brass adapter, so I've got some flow through there. Air pockets are also still a likelihood though, as I don't think I've got all that sorted yet.

I wish I would have thought about ditching the 207* temp switch earlier. It would have been a much easier swap when I had everything apart. I even already had the Starion switch, which I believe is the same thread size and pitch.

 

Midday update. I took some time on a break this morning and at the beginning part of my lunch to go through my wiring on the relay for the e-fan. Everything looks good there. My +12V ignition source has 12V when key is ON. My positive from an unused spot on the fuse block has 12V at the backside of the fan connector. The backside of the fan connector ground terminal had continuity where it goes into the relay, and the ground to chassis also had continuity at the relay. As a final check, I jumpered the relay connector for pins 30 and 87 and the fan immediately powered on. I suspect the adapter at the BAC line, while getting *some* flow as evidenced by the leak, was not getting *good* flow exactly as you said. I am also taking your suggestion and relocating the temp switch to the factory location, I have already removed my BAC adapter and put it back together stock, but pulling the intercooler and alternator to get to the temp switch is an after work job.

I cleaned up the area by the Koyo's drain, so I can tell if I really have a leak or not. I think it may have been residual overflow when I created a geyser at shut off after the initial start-up and bleed attempt.

 

Sounds like you got it sorted out. Good luck, see you at DGRR!

BTW, you probably already know this, but you can use that plastic plug on the Koyo upper rad hose to assist in bleeding the system of air when you refill the coolant - just keep it open until coolant starts leaking from it, then plug & continue filling. Also, if you don't have one of those Lisle funnel kits, get one - makes it so much easier to get air out of these cooling systems.
The way it works is you attach the funnel to the rad, it makes a leak free seal, and as long as you keep the funnel about 1/3 ~ 1/2 full, as the engine runs and gets hot, air will burp its way out, and you just keep adding coolant to the funnel until it won't take anymore and stops burping bubbles. Then you use the handy plug piece so you can remove the funnel and return the excess coolant in the funnel to a container with no mess.

 

So, I'm about one or two steps away from wrapping this up so I can hopefully get some other things done before DGRR!





Out with the old and in with the new. The auxiliary fan switch removed in favor of a Starion switch. Access was probably even worse than I thought. I no longer have a 12 point 17mm wrench, only a 6 point, so I also had to use a pair of needle nose pliers to gently turn it a little more so I could get the wrench back on. The Starion switch is 21mm, and I had a 12 point, so other than going 1/8 of a turn at a time, no issues.

I started it up and after a few minutes of warming up, the fan came right on. So, check that one off. I do have one of those Lisle funnels, but the adapter does not seem to want to get onto the neck of the Koyo. Thankfully, I believe it is mostly bled. I also did a pressure test to verify I have no leaks and all is good there. It held pressure and there were no identifiable leaks. All that's left for this project is to warm it up one last time to adjust the TPS. When I had the upper intake off, I did some throttle body adjustments since it was out of spec according to the FSM. After that should be good to go other than keeping an eye on the coolant levels for a few more cycles to make sure I really do have it bled.

Fan in motion:



Also, I had a glass company come out and get the windshield installed last week. Some more notes on that to come later.

 

Ahhhhhhhhh! Ten days until I leave for DGRR, and I still have a few things to accomplish. Thankfully, they are becoming much smaller in scope and don't require nearly as much destruction as the cooling system project.

So, about that windshield...

Aftermarket companies do not have any glass when you ask for glass for a 1991 RX-7. I tried several. That's how I eventually ended up talking to Ray Crowe and discovering Mazda had a few in California of a certain FC coupe windshield part number. I *thought* this part number was not the correct one because it wouldn't have the antenna, which I was fine with. Turns out it did have the antenna, and the real issue was that it was for an S4. The difference in S4 and S5 coupe glass is the width of the frit on the sides, the S5 has a larger frit to cover up the mouse belt track. Well, it would have been helpful if I had figured that out *before* the glass guy got here, as I maybe could have taken a different approach and had him lay down some more primer for a larger frit. Instead, my friend came up with this idea, which I executed today:





This was the easiest, 100% reversible thing to do. I cut some strips of ABS about 0.75" wide and 28" long and used some mount tape to attach it to the A pillar. Some careful finagling into place and I was done. I probably could have done a little more up top to cut a curved piece to match the A pillar, but I didn't realize that was also going to be an issue until it was clipped in and I'm not taking the chance with FC plastics. I wager that nobody really notices this unless they know what to look for or are staring way too hard at my windshield.

Fixed the instrument cluster hood mounts thanks to this ingenious tab designed by forum member cone_crushr. I have a friend whose step-dad owns a 3D printer. I offered money for time and materials and they refused. I told them my car thanks them very much.



Waiting on a new (used) cruise control switch as I think I finally found the issue there. Then just a few more things to try and improve that are more about comfort at this point than anything else.

 

Can you get us the S4 windshield part number please?

 

S4 windshield shaded green with the diversity antenna is FB03-63-901E.

 

Starting to feel a bit pessimistic about engine health. The system had been holding pressure for long after the car had been shut off (talking ten hours and the upper hose was still tough to squeeze). I wanted to really make sure I had the cooling system bled, so I put the funnel on there and started it up. After the thermostat opened, there were pretty regular big bubbles coming through, but the level in the funnel never really dropped. I can hope for the best, but am assuming the worst. I had pressure tested it last week, and I probably didn't wait long enough - but pressure seemed to be bleeding off at the exact rate it was when the pump was pressurized by itself. Probably gave myself some false hope on that one.

In other news, I fixed the clock! And the cigarette lighter. Turns out the L/B wire somehow was in the wrong position on the front harness. I put it in its proper spot on connector X-20 and they both worked. This is not something I can ever recall messing with and I don't remember if these two things ever worked when I was daily driving the car from 2005-2009.



I'm going to go drink a beer while I tell myself maybe the engine really isn't on its way out.

Cheers!

 

I actually noticed sometimes this happens, but when it cools the coolant contracts back down. I assume the air is bleeding out like normal and the expansion of the hot coolant makes it appear as if the level is not dropping.

If yours continues bubbling after a couple heat cycles I'd assume the worst, but you have nothing to lose by trying a few more times to bleed it. As long as it isn't overheating, that is.

Best of luck! Hope it all works out.

 

Pete suggested to me that it could potentially be localized coolant boiling making it's way through the system and the level never dropped because the system is actually full. But he also made that comment while asking if it was already warm and up to operating temperature.

I'd definitely like to find out I'm overreacting, but I think back to the initial cooling system failure - a pinhole leak in the radiator. Possibly from too much pressure? The thing was likely original to the car though, so who knows.

 

Do you have an accurate temperature gauge to keep an eye on coolant temps? What kind of coolant temps are you seeing when the car is running normally with the pressure cap installed? Does any coolant puke into the overflow tank when the car is running and at operating temp?

For your current cooling system setup, depending on ambient temps and the set points you have to control your e-fan on/off cycles, you should be seeing temps in the ballpark of 185~195*F. And the only time you should see coolant puking into the overflow tank is after you shut down a hot engine - temp/pressure increases when the coolant stops flowing; the excess vents via the pressure cap thru the overflow hose to the tank. Then as the system cools, it pulls a vacuum, sucking the coolant recovered in the tank back into the engine/radiator.

 

Unfortunately, no accurate temperature gauge. Only the stock S5 gauge, which has never wavered off of halfway up once warm, so that only means I have never reached "oh, crap" territory.

I have not noticed any coolant going into the overflow while it's been idling in the garage, but that's also all it has done so far other than a quick trip around the block last week. I'll take it for a drive today or tomorrow and take careful notice of levels in the overflow tank at the beginning, a time or two during, and then after I get back.

 

I think I've talked myself off the ledge?

Here's pressure test #2, done to some more exacting standards:

Picture #1, 10:25am, slightly below 15 psi


Picture #2 10:28am, and also the biggest drop, right around 14 psi.


Picture #3, 10:55am, a full half hour after the test started, and a hair width below 14 psi:


And here's a cranking test with the focus on the coolant neck. This is a cold engine, I confess in my 22 years of driving a rotary, I've never had to perform this test, so still looking to see if this was supposed to be done cold or hot. FAQ seemed to allude to cold. And yes, I have the click, click, vroom issue.

 

I don't claim to have the most experience with coolant seal issues (in fact my previous engine had two broken coolant seals and zero bubbles or overheating symptoms), but I don't see anything I would worry about there.

The pressure test doesn't bug me even if it does technically lose ~1psi. I would suspect a certain amount of that is due to the tolerances in the tester and the seal onto the rad opening. And as for the second test I don't see any bubbles at all.

I would definitely get a real temp gauge though, just for peace of mind.

 

Took it for a couple of drives last night, for a total of over 20 miles. There is no bad news to report, only good news and things to possibly keep an eye on.

The good news - I had one of my kids take a video of startup, and there was no noticeable smoke. On either drive, with ambient temps in the high 80s, and a decent bit of stop and go on the neighborhood type streets, the temp gauge never wavered from the halfway mark. The add coolant light/alarm also never went off.

The thing to possibly keep an eye on is not what overflowed into the tank, but the fact it never got pulled back in overnight. Here are some pics of the overflow I took at various times:

Before initial start-up:


End of first drive after shut off:


End of second drive several minutes after shutoff:


And finally, this morning:


The upper radiator hose this morning was not stiff as if still pressurized, but could be squeezed.

I will say, the view out of the brand new windshield was nice!

 

Since nothing had been drawn into the system from the overflow, I wanted to check the level. Opened the radiator cap and was greeted with coolant puking out, so the system is maintaining pressure overnight.

I have a new radiator cap on order (even though this one was new, it had sat around for many years), but I think I'm back to not being entirely sure what the issue is here. If the cap doesn't fix it, I'm running out of potential issues.

 

Since coolant isn't being drawn back into the engine from the overflow tank, it means that either the pressure cap seal is defective, or there is a tiny leak somewhere along the overflow hose run or its connections into the tank - it doesn't take much of a leak here to break the vacuum needed to draw coolant back into the engine. Double check the hoses on and inside the overflow tank cap to ensure they are correctly plumbed, especially if you removed them for any reason. The hose that connects to the engine should connect to the nipple on the cap that connects to the hose under the cap that goes to the bottom of the tank and remains submerged. If you accidentally reversed the hoses on the cap, it won't draw coolant back to the engine.

 

I'll double check the hose run, but that should be inconsequential to the system staying pressurized, no?

My thought was it's not drawing anything in since it is staying pressurized and never creating any sort of vacuum in the first place.

 

Given the pictures posted, it looks like coolant is getting to the tank - which means the cap is at least partially working. But the reverse flow back into the engine on cool down isn't happening, so anywhere in that circuit that is leaking will break the vacuum needed to draw coolant back into the engine. The seals on the cap may be degraded, so I'd just get a new OEM radiator cap, and double check the hoses & connections to the tank.

 

Simpy put - maintaining pressure when cold is bad. A new cap is the right answer as a start. Alternative is to pressure test the cap to ensure blowoff at set pressure (~1 bar); it may be stuck.

 

Absolutely, definitely shouldn't have maintained that pressure overnight. I don't have a cap adapter for my pressure tester, so I went straight to attempting a new cap.

I'll get some new line for the overflow, I think what I put on may be a bit oversized. Discovered the stock line had been pinched when I was taking everything apart and what I replaced with is not as snug a fit. Come to think of it, that pinched line could well have accelerated the radiator's failure - probably never allowed anything to ever reach the overflow.

Today's my tenth wedding anniversary, so won't be until tomorrow before I have any new info. We shall see.

 

Since the traditional 10th anniversary gift is aluminum, tell your wife your new rad is her gift to you. Congrats!

 

Tried a new radiator cap and running a new line from the radiator to the overflow tank. Took it for a short drive yesterday, at which point the system was still pressurized eight hours later.

Did the combustion gas test twice this morning and (for anyone unfamiliar, the fluid is originally blue and turns yellow in the presence of hydrocarbons from gasoline)



The bubbles, the system staying pressurized, and now this can only lead me to one conclusion. Unfortunately, my engine is on its way out.

I'll wrap up this thread later, still a couple of things I fixed but didn't mention.

 

Yeah, the combustion gas test 'failure' is the final straw.

I never considered the criticality of the overflow tank hose integrity. A kink in the hose can overpressure the coolant system risking not only the radiator, but the entire engine! Thank for the lesson - unfortunately. May want to keep an eye out for white smoke and water in engine oil as an indicator of coolant seal leak severity. Head gasket sealers seem like a bad ide, although I suppose the logic goes "what have I got to lose?"

You might be OK driving it for a little while, as long as you keep it out of boost - no fun, but keeps it roadworthy until you're ready for a rebuild.