When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
I just tested my coils. All 3 coil packs are showing in-spec resistance. T1 .6 ohms, t2 .6 ohms, Leading pack .6 ohms and L1 to L2 13.6 kOhms
Gonna splice the fuel pressure testing kit to my feed line and see what I see there.
I hooked up the fuel pressure gauge to the feed line and ran the car up to operating temperature on jack stands.
Learned a few things about what's happening. The car is still exhibiting the same behavior which is that it runs mostly fine for 5 minutes, slowly degrades in idle quality for 5 minutes, and then rapidly chokes sputters and stalls itself out.
the good news is that im not having any fuel delivery issues. The car maintained 48 psi fuel pressure at the feed line for the entire duration of the test and held 48 psi even after the car stalled out.
the bad news is that my coolant temp sensor is still being temperamental. The first 1-2 minutes the car is on, the sensor behaves and reports almost exactly the same coolant temps that i see on my aftermarket gauge which has its own separate sensor. Suddenly, it jumps to 175c. Surprisingly, the temp jumping to 175 doesnt seem to affect how the car is running at all.
The car runs mostly OK for the next 8 minutes. I start to hear an occasional miss in the idle as time progresses, but nothing of note in the logs.
THEN, the temp suddenly goes from 175c to -38c which puts the ecu into panic mode. From -38c the coolant temp sensor rapidly counts up as if the sensor just "reset" and is climbing back to its actual reading.
As soon as this happens, the iacv jumps to 80%, i can hear air rushing past the throttle body, and the car stalls out.
The tps doesnt freak out, it reports correctly. (I gave it a little throttle right as the coolant sensor was freaking out.)
Also of note was that the battery voltage dropped from a constant 14v the entire time down to 13.3v which seems to imply that there may be a small short to ground when the sensor suddenly dips from 175c to -38c. (But this could just have been a side effect of the rpms rapidly dropping as the iacv stalled the engine out)
my new coolant sensor will be arriving in the mail tomorrow but not a lot of hope that itll fix the issues since it seems more likely that theres a temperature sensitive break in the wiring somewhere which would cause this weirdness.
Now my question is, what could be causing the iacv to suddenly full open?? Is it simply because the ecu suddenly thinks the car is freezing cold? Or is there some signal pollution from some kind of other short?
The next thing I'm going to try, is to zero out the coolant temp correction in the PFC map and see if that stops the ISCV from going nuts.
If so, then at least I know this issue is isolated to the coolant temp sensor circuit and doesn't involve the ISCV getting random signals from somewhere other than the ECU
Surely you mean 175 degrees feranheit. Boiling is 100 degrees "c".. Probably you don't want to see more than 110 degrees "c".
its not that i want to see it, its what the pfc logs are saying the sensor is reporting. Obviously its not accurate, since i have another water temp gauge reporting the actual temp which is 190f or so
Some folk have simply recommended unplugging the ISC if it is suspicious of acting up.
What is the status of your idle screw
For my flooding issue, in the end, we took off the ISC off altogether and the tech used his fingers to moderate the air intake (so we did not open the throttle at all).. They were his fingers not mine, and I am not recommending it, but it was something that let us finally get the car started.
Some folk have simply recommended unplugging the ISC if it is suspicious of acting up.
What is the status of your idle screw
For my flooding issue, in the end, we took off the ISC off altogether and the tech used his fingers to moderate the air intake (so we did not open the throttle at all).. They were his fingers not mine, and I am not recommending it, but it was something that let us finally get the car started.
Deleting the iscv definitely seems like a "band aid" solution where i'd much rather fix the underlying issues. The valve itself does exactly what the signal voltage is asking it to so i gotta figure out why its doing that.
Most likely culprit so far still seems to be the harness but im just looking for some more conclusive evidence.
its not that i want to see it, its what the pfc logs are saying the sensor is reporting. Obviously its not accurate, since i have another water temp gauge reporting the actual temp which is 190f or so
Maybe wait until you get the new sensor? Or can you switch the sensor that is feeding your analog gauge to feed the power FC input?
Or put another analog gauge onto the sensor that is now feeding the Power FC?
Any chance that somehow the wiring for the thermoswitch for the fans and wiring for the water temperature to the ecu, got crossed up? You said you rewired those?
Looking at the Version 1 JDM wiring diagram the wiring for both the temp sensor and the thermoswitch run through the "EM" harness to connect to the "D" harness through connector x-19. The wires from the temperature sensor then connect to temp sensor through the X-16 connector. The thermo switch wires connect to the "f" harness through connector x-06.
I would caution that the JDM and USDM wiring may be different. Also Mazda actually came out with many iterations of the "d" harness in the first year.
Seems there is another water temperature sensor that connects directly to the ecu at pin 3E. It seems all three sensor's wiring commonly run through the emissions harness at some point.
I just tested the sensor out of the car with the car off and cold. Sensor works fine up until about 60-70c and then it develops an open circuit condition which is making the ecu read 175c. I guess when the circuit intermittently closes is when i see the temp jump down to -38c and start counting back up.
new sensor should come in the mail today and i will update the thread if that conclusively fixes my stalling issues
I just tested the sensor out of the car with the car off and cold. Sensor works fine up until about 60-70c and then it develops an open circuit condition which is making the ecu read 175c. I guess when the circuit intermittently closes is when i see the temp jump down to -38c and start counting back up.
new sensor should come in the mail today and i will update the thread if that conclusively fixes my stalling issues
i was about to say that 175 and -38 are like opens and shorts in the circuit
Car idled fine for half an hour without sensor wonkiness. I think the issue's been fixed. Will take it off jack stands and test drive during the weekend.
crazy that just a bad coolant temp sensor can do this to the car. These things are such drama queens.
Well, the stock ECU would throw "a code" if the temp sensor signal was faulty. It is code "09". So you would likely get a check engine light on USDM and, possibly, "limp mode" on JDM.
Does the Power FC ever emulate the stock ECU "limp mode"?
Come to think of it, I've seen similar coolant temperature sensor failure from a 1989 Honda CRX. It's good you had video showing the sensor screen, that might have been hard to find with other methods. We got lucky on that CRX, the car had a standalone ECU and the laptop software showed a plot/trace line for coolant temperature so it was pretty clear to see when the calculated value jumped around erratically. The sensor isn't very expensive, so it might make sense to replace them proactively on these cars as they get older.
Car idled fine for half an hour without sensor wonkiness. I think the issue's been fixed. Will take it off jack stands and test drive during the weekend.
crazy that just a bad coolant temp sensor can do this to the car. These things are such drama queens.
Bad temp sensors (not just coolant but air temp as well) will cause these issues in many cars, not just RX7s. It causes the computers to go into a type of limp mode in an attempt to prevent damage. Same thing with O2 sensors and such for OBDII vehicles, not so much with OBDI or non OBD. Any computer that uses sensors as feedback usually has these types of "safety's" built in one way or another.
Bad temp sensors (not just coolant but air temp as well) will cause these issues in many cars, not just RX7s. It causes the computers to go into a type of limp mode in an attempt to prevent damage. Same thing with O2 sensors and such for OBDII vehicles, not so much with OBDI or non OBD. Any computer that uses sensors as feedback usually has these types of "safety's" built in one way or another.
Limp mode is expected but total loss of control over the car + stalling out and dying is both very concerning and a bit melodramatic. More modern cars would probably have safeguards in place to prevent a throttle bypass like our iscv from suddenly slamming full open and killing the car. But i guess more modern cars wouldnt even need such a bypass as theyd be throttle by wire.
(I'm just throwing a fit because a stupid little sensor made me relive the trauma of all the times ive experienced rotary compression loss)
Well, the stock ECU would throw "a code" if the temp sensor signal was faulty. It is code "09". So you would likely get a check engine light on USDM and, possibly, "limp mode" on JDM.
Does the Power FC ever emulate the stock ECU "limp mode"?
FC Tweak would have caught it when he went to go actually tune the map.
FC Tweak would have caught it when he went to go actually tune the map.
It would have if i didnt enable 2-step to see what it does. Apparently this makes the map read-only and prevents you from starting an autotune or manually editing on that map. I was also confused why it refused to do anything with my logs/map until i read that part in the manual and asked xavier about it.
Essentially, once a map has been 'drag race mode' enabled it becomes immutable and any logs generated while running that map cant be used to edit that, or any other map.
08-28-24, 04:51 PM
