I have a 1991 N/A RX7 with the strangest problem I've ever heard of. I've googled endlessly, and searched these forums fairly extensively. I couldn't find anything that seemed relevant.

When the weather gets really cold, even for just one night, a strange problem appears. The engine will start fine, and run fine, until it starts heating up. It runs normally for about 2 minutes or so, but it's status gradually degrades as it warms up, until it reaches the point of being unable to run when it reaches normal operating temperature. If I press the accelerator, it stumbles and struggles to rev, or stumbles and is completely unable to rev, as if it's not getting enough fuel. When / If I get and keep it above 4k rpm, it seems to run fine, except if I floor the accelerator, it stumbles momentarily before revving just fine, but if it drops back below 4K rpm, it wants to stall. Strangely, if the RPMs drop to the very bottom of the gauge, and I have the accelerator floored, the car will try to save itself, and if successful, will rev just fine, but when I release the accelerator, it still wants to stall. Needless to say, the car is undrivable in this condition. Usually if I wait for a few days of warmer weather, it will run just fine.

I checked the MAF (temperature portion was checked by electrical resistance vs current temperature), TPS (see below), Air Intake Thermo Sensor (checked cold, electrical resistance vs. current temperature, and breathing on it made the resistance decrease), Water Thermo Sensor (electrical resistance vs. current temperature, as I didn't want to remove it without having sealant to put on the threads), alternator (Auto Zone checked it), spark, (grounded plug in wire on the intake manifold, both leading and trailing gave spark, but leading gave a significantly smaller spark, but I figure this is normal, as the leading coil is much smaller than the trailing), Electronic Spark Advance (electrical resistance as in the FSM), O2 sensor (as per FSM), and I checked all the vacuum hoses to the best of my ability.

The Water Thermo Sensor is very hard to get to in my specific car, I think the shop that rebuilt the engine put a set of wires in the slightly wrong spot, and I can't get to the sensor without removing the alternator, which pins the wires in place and is very hard to remove due to my strut bar, which doesn't want to be removed.

When I checked the resistance of the TPS, it was a little out of spec. The narrow range (the one that can be adjusted with the screw) can be barely adjusted back into spec. The full range (the one that can't be adjusted with the screw) was about 500 ohm out of spec, but if I made a makeshift spacer out of duct tape on the end of the plunger, I could get it back into spec. If I plug in test lights that you would use to adjust the TPS, I can't get any light by adjusting it, but if I push in the narrow range plunger further without the spacer on the full range, both (or one if I push it just right) would light up, but pushing it in with the spacer on the full range would not give any light. I also checked for gaps by having someone very, very slowly (digital multimeter) move the throttle to full open.

Because I am currently living in an apartment away from home, I can't do more extensive things like removing the intake manifold to get to the fuel injectors (even though I don't think they're the culprit due to the engine running when the engine is cold or the weather warms up) to check them or to have them cleaned off of the car. I can do this if I can get the car running good enough to get it home, but this would mean the problem would have disappeared by waiting.
I would check the compression, but don't have a gauge here, and as I said, it revs fine if I keep it above 4k rpm and it runs fine when the weather is warmer.

I also don't think it could be a fuel system problem or bad gas since it runs fine if the engine is cold. I think it's more of a control system problem, but if there is sufficient reason (other than "you should just check it"), I won't rule out the possibility of the fuel system malfunctioning.

It hasn't always done this, as last year it ran fine on the coldest days of winter, except for when it sat for nearly a month after I got hit by a flying deer (bounced off SUV) and had to drive a different car and my dad didn't start it a few times a week, and it ran fine once I got it running after that. Now, there is no additional car that is available, and the rx7 always has been and always will be my DD (and I wouldn't have it any other way).

 

You say it was sitting for a while. Did you put fresh gas in it?

BTW, you should be checking the sensors when the car is at full operating temp. Checking it when cold can give very different readings.

Does it rev to 4k stutter and keep going or does it have trouble at any rpm under 4k?

 

No, it hasn't been sitting for a while, I guess I was a little unclear. It sat for a month, but that was a whole year ago, and I got it running like normal back then. The last time I had the car running before it's messed up now was I think Sunday, but this problem has been showing up since the coldest days of this winter.

It has trouble at any rpm under 4K.

When I checked the O2 sensor, I couldn't check the sensitivity, because I don't have the SST, I only checked the voltage as described on F1-83 in the FSM. I did this with the engine at operating temperature, above 4K rpm so it wouldn't stall.

Also, this happens when the car is in any gear or neutral.

 

First off, I decided to replace the O2 sensor since it was at least 1.5 years old and not expensive if you don't get the one with the $100 plastic clippy thing.

It got cold again, but before it got too cold to where it wouldn't run at all, I took it for a drive. It kept having power loss, and was very tricky to take off with even less torques than it normally has. The check engine light came on for a sec, and the car ran better. When I got back, I checked the code, and it gave me code 17.

Code 17 - Feedback System - Oxygen sensor output remains 0.55V 10 sec. after F/B system operation beginning - Feedback system canceled (For EGI)

 

besides a visual check of the vacuum connections, did you spray carb cleaner to locate vacuum leaks?

Did you try to set the idle speed using the FSM procedure?

 

Sounds to me that you have a problem with one of your primary injectors. The reason it clears up after 4k is because the secondaries are kicking in. Backprobe the primary injector wires at the ecu (only way to check without pulling intake) to see if your getting a signal out of the ecu. If you are than i would recommend trying good known injectors. There's nothing else that kicks on or off around 4k.