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.
My ECU is apparently stuck in limp mode on my nearly stock series 7 '96 FD. Is there a code reader that is readily available that can read what the fault may be from the diagnostic plug on these cars? I just recently learned that there is no check engine light to count the flashes like my older Mazdas so that's a bit disappointing.
My ECU is apparently stuck in limp mode on my nearly stock series 7 '96 FD. Is there a code reader that is readily available that can read what the fault may be from the diagnostic plug on these cars? I just recently learned that there is no check engine light to count the flashes like my older Mazdas so that's a bit disappointing.
Any recommendations would be helpful. Thank you!
What behavior(s) is the car experiencing? I could probably diagnose it from here as I have the procedure for the Eunos Cosmo already translated from Japanese and it uses the same connector/pinout. Should be a piece of cake to adapt it to a Series 7/8 FD. On the Cosmo, it is essentially "Short the TEN Terminal to ground, then watch the dashboard indicator flashing pattern". Manuals spells out how to turn this into a code number, which then describes what parts or systems are acting buggy.
What behavior(s) is the car experiencing? I could probably diagnose it from here as I have the procedure for the Eunos Cosmo already translated from Japanese and it uses the same connector/pinout. Should be a piece of cake to adapt it to a Series 7/8 FD. On the Cosmo, it is essentially "Short the TEN Terminal to ground, then watch the dashboard indicator flashing pattern". Manuals spells out how to turn this into a code number, which then describes what parts or systems are acting buggy.
The car starts and idles just fine, you can drive slowly but as soon as there is any boost it cuts spark or fuel and won't rev. From what I've read searching around on here that is the ECU protecting the car because it has sensed some sort of fault. Also entirely possible I'm off on assuming this just trying to diagnose by searching old posts before I do any more.
I've jumpered the diagnostic plug pins like described in the technical posts I've read, problem is apparently the JDM cars do not have a check engine light to count the flashes. I've done this procedure on my Mazda 929 to read fault codes. With a JDM car I assume I will need an actual reader?
The car starts and idles just fine, you can drive slowly but as soon as there is any boost it cuts spark or fuel and won't rev. From what I've read searching around on here that is the ECU protecting the car because it has sensed some sort of fault. Also entirely possible I'm off on assuming this just trying to diagnose by searching old posts before I do any more.
I've jumpered the diagnostic plug pins like described in the technical posts I've read, problem is apparently the JDM cars do not have a check engine light to count the flashes. I've done this procedure on my Mazda 929 to read fault codes. With a JDM car I assume I will need an actual reader?
I'm sorry, I was mistaken before. Covid has me off my A-game a bit right now. To read Engine Fault Codes, you have to short the FEN Terminal to Ground, not the TEN Terminal. The following information is from memory, as I cannot locate the information I previously translated for the Cosmo, nor its pulse pattern example images right now.
At the moment, I cannot recall which Indicator Light flashes for Engine Fault Codes, but I do know the HOLD Indicator flashes for Transmission Codes (Short FAT Terminal to Ground) and the ABS behaves in the same manner (Short FBS Terminal to Ground). All use the same "Morse Code" style Flashing Pattern format. Displayed codes are typically done as follows:
When you turn the Ignition to ON, the first 3 seconds are to Initialize Self-Diagnostic mode (Light is ON), then 2 seconds OFF. Now it begins with the first code by flashing the tens digit (1.2 seconds per flash, 0.4 seconds between then), wait 1.6 seconds, then flash the Ones digit (0.4 seconds per flash, 0.4 between them), wait 4 seconds, then it goes onto the next code in the same manner (tens, 1.6 second gap, then ones).
For example, if it is code "23", you'll have
3 seconds ON, 2 Seconds OFF, then 1.2 ON, 0.4 OFF, 1.2 ON, 1.6 OFF, then 0.4 ON, 0.4 OFF, 0.4 ON, 0.4 OFF and 0.4 ON.
For ABS on the Cosmo, Code 23 is a Solenoid Valve issue for Left Front Output
For the ECU, Code 23 on Cosmo is Fuel Temperature Sensor Signal.
All FDs would be similar if not identical in behavior. Not exactly sure offhand what indicator would be used for Engine Fault Codes, perhaps the Shift Indicator? Take a look at your cluster for something distinctly engine-related which would be on both automatic and manual transmission cars.
EDIT:
So I just looked through my Japanese manuals, going backwards from 2002 and found that they all use the same procedures for Engine Fault Codes. Either the DT-S1000 (Part# B099-49-0A0) or the System Selector (B019-49-9A0) plus DISC Monitor (H018-49-9A0). The Indicator Lamp is NOT in the dashboard, it is on the DISC Monitor. Codes are set up the same as everything else though, format is identical.
The DISC Monitor/System Selector method is the "analog" method of getting fault codes. I'm going to venture a guess that the System Selector's 7-position switch just applies Ground to each of the respective terminals in the Diagnosis Connector, so that the DISC Monitor can flash the appropriate system's Fault Codes. If this hypotheses is correct, you MIGHT be able to use a 12v Test Light in place of all of it, connected to Ground and the FEN Terminal.
Last edited by Akagis_white_comet; 12-21-21 at 08:01 PM.
Reason: More FD-specific info
The DISC Monitor/System Selector method is the "analog" method of getting fault codes. I'm going to venture a guess that the System Selector's 7-position switch just applies Ground to each of the respective terminals in the Diagnosis Connector, so that the DISC Monitor can flash the appropriate system's Fault Codes. If this hypotheses is correct, you MIGHT be able to use a 12v Test Light in place of all of it, connected to Ground and the FEN Terminal.
Thank you so much for the help! Just to review that I'm doing this correctly and if there is any difference on the JDM cars i've attached a photo of how I'm shorting the diagnostic plug.
My big question is, in the absence of a check engine light to count the flashes for the code, is there one of these pins I can connect a test light to that would allow me to read it? Just wanted to ask before I do something stupid.
Thank you so much for the help! Just to review that I'm doing this correctly and if there is any difference on the JDM cars i've attached a photo of how I'm shorting the diagnostic plug.
My big question is, in the absence of a check engine light to count the flashes for the code, is there one of these pins I can connect a test light to that would allow me to read it? Just wanted to ask before I do something stupid.
If it were a Cosmo, TEN Terminal to Ground would put the ECU in Test Mode. That locks ignition timing and a few other things. TEN= Test Engine. FEN=Faults on Engine, and so on.
Thankfully, the Diagnostic Connector is the same on every Pre-OBDII Mazda. Here is a pinout and definition of it on the Eunos Cosmo. Same info applies to ever FD.
I'm not entirely sure if the FEN Terminal puts out 12v pulses or Ground to make the DISC Monitor's indicator flash, but in lieu of a schematic for the DISC Monitor/System Selector setup, or an original to make a schematic from, using a Test Light connected between FEN and Ground seems like the safest method at the moment. No chance of shorting out the ECU that way, especially if using a LED bulb.
You are a genius, thank you! Since this might help someone else here is what I did. Connect TEN to the far ground in the diagram above. Then use a 194 LED with the pins bent out and they line up perfectly with FEN and GND. Check polarity with a battery if you want or flip around if it doesn't work. From there you should have the LED flash to read the diagnostic codes.
From what I can tell I have a few errors but I'm still not entirely clear when it moves to another error code. I have the long and short flashes but sometimes they are close together. The sequence starts with one very short flash, two second pause, one long followed by three short, one long then three short all about a second or so apart. Then one long into four short flashes but these are all super close together, maybe a half second apart. The initial brief flash before the pause does not return in the sequence as it starts over with no breaks. Any ideas or would it help if I had video?
You are a genius, thank you! Since this might help someone else here is what I did. Connect TEN to the far ground in the diagram above. Then use a 194 LED with the pins bent out and they line up perfectly with FEN and GND. Check polarity with a battery if you want or flip around if it doesn't work. From there you should have the LED flash to read the diagnostic codes.
From what I can tell I have a few errors but I'm still not entirely clear when it moves to another error code. I have the long and short flashes but sometimes they are close together. The sequence starts with one very short flash, two second pause, one long followed by three short, one long then three short all about a second or so apart. Then one long into four short flashes but these are all super close together, maybe a half second apart. The initial brief flash before the pause does not return in the sequence as it starts over with no breaks. Any ideas or would it help if I had video?
I'm glad it worked out for you. Anyway, my apologies as I am currently packing up to move right now so I won't be able to give you a breakdown of the codes very easily until I'm settled in and set back up at the destination. Give me a week or so and I'll be able to answer your question in proper detail.
This trick definitely deserves being added to the 3rd gen FAQ.
No rush at all, when you have time I'd appreciate if you or anyone else wants to take a crack at trying to decipher the code for me. I have uploaded it to YouTube. I just let the video run for about two minutes then shut the key off.
Sorry I didn't get back to you sooner, been busy with other stuff. I'm seeing 03, 13 and 14, but I may be mistaken here. Still waking up right now.
Code Descriptions:
03: Crank Angle Sensor G Signal
13: Pressure Sensor Signal (MAP sensor)
14: Atmospheric Pressure Sensor Signal (Inside of ECU)
Test the Map Sensor in the same manner as a North American one as follows on the signal wire:
At -66.7 KPA (-500 mmHg, -19.68 inHg), it should read 1.25-1.55v
At 0 mmHG (0 PSI), 2.38-2.78v
At 98.7 kpa (740 mmHg, 14.31 psi): 4.35-4.65v
Crank Sensor (G) is the LOWER one. Should be 1.5mm +/- 0.5mm from the Crank Trigger Wheel. Resistance spec is 950-1250 ohms at 20 degrees Celsius (68 Fahrenheit). The mounting nuts are torqued to 7.8-10.7 Nm (5.75 - 7.89 ft.lb)
Thank you! This helps a lot. I'm guessing the crank sensor is likely the one serious enough to put this thing into limp mode.
Since I am new at these engines, could you help me physically locate the MAP sensor on the car? I am guessing the ECU sensor there is nothing I can do about at this point. Is there actually a pressure sensor inside the ECU or is this just a signal pickup from a sensor somewhere else?
I'm also curious why there is no pause between the codes as I expected reading guides online. Well at least if this helps me narrow down my issue, thanks so much for looking into it.
Thank you! This helps a lot. I'm guessing the crank sensor is likely the one serious enough to put this thing into limp mode.
Since I am new at these engines, could you help me physically locate the MAP sensor on the car? I am guessing the ECU sensor there is nothing I can do about at this point. Is there actually a pressure sensor inside the ECU or is this just a signal pickup from a sensor somewhere else?
I'm also curious why there is no pause between the codes as I expected reading guides online. Well at least if this helps me narrow down my issue, thanks so much for looking into it.
MAP Sensor is bolted onto the firewall, directly behind the Throttle Body and Intake Elbow. See Post #4, first picture here. Left side of picture, roughly centered with the connector on the right-hand side: https://www.rx7club.com/3rd-generati...-help-1140037/
The ECU's internal Pressure Sensor measures Atmospheric Pressure to compensate for altitude changes. Cosmo has the same setup too. The FC had a similar setup, but used an external sensor tucked away in the kick panel. As for the Crank Angle Sensor(s) flaking out, one acting oddly is more than enough to make the car "nope" out like you're seeing. However, they are Magnetic Reluctors which are generally quite reliable. A MAP Sensor is a Voltage/Potential Divider, which functions using a flexible membrane with electrical traces on it. It changes impedance based on Pressure, either positive or negative to output varying voltages that the ECU can understand in order to calculate fuel needed. After 20+ years, a Map Sensor conking out is more likely than a flaky Crank Angle Sensor. Test per the specs I posted above, procedure and pinout are the same as a North American FD.
If the ECU sensor is bad, I would assume I need to get a new ECU at that point or upgrade to a Power FC.
I'm an amateur at electrical testing so actually diagnosing this by measuring voltage at the specified pressures is beyond me. Is there a guide to how I can physically do this? I was going to swap it out for a new one until I saw the price. May be headed there anyway.
If the ECU sensor is bad, I would assume I need to get a new ECU at that point or upgrade to a Power FC.
I'm an amateur at electrical testing so actually diagnosing this by measuring voltage at the specified pressures is beyond me. Is there a guide to how I can physically do this? I was going to swap it out for a new one until I saw the price. May be headed there anyway.
The good news is that this is a purely old-school setup, so no specialized tools are required. All you need is a common multimeter. With the Map Sensor connector (Connector B1-40) still plugged in, put the positive probe on the connector's Green/Yellow Wire's terminal (AKA backprobing) and the Negative Prove on any convenient chassis ground location. Battery Negative works fine for this. Turn the key ON and read the voltage. Then change the pressure that the sensor is seeing and read the voltage again. As above, it will read within a certain voltage range at each of the 3 pressure readings I posted before.
A Mityvac can make this easier to do, as you can precisely pump the pressure up/down to get accurate, repeatable results. If memory serves, the vacuum line is 4mm ID (Inside Diameter).
The good news is that this is a purely old-school setup, so no specialized tools are required. All you need is a common multimeter. With the Map Sensor connector (Connector B1-40) still plugged in, put the positive probe on the connector's Green/Yellow Wire's terminal (AKA backprobing) and the Negative Prove on any convenient chassis ground location. Battery Negative works fine for this. Turn the key ON and read the voltage. Then change the pressure that the sensor is seeing and read the voltage again. As above, it will read within a certain voltage range at each of the 3 pressure readings I posted before.
A Mityvac can make this easier to do, as you can precisely pump the pressure up/down to get accurate, repeatable results. If memory serves, the vacuum line is 4mm ID (Inside Diameter).
Thank you, that makes perfect sense. I can see I'm going to have to invest in a vacuum pump tool to properly diagnose this car. Really appreciate the detailed help on this!
Thank you, that makes perfect sense. I can see I'm going to have to invest in a vacuum pump tool to properly diagnose this car. Really appreciate the detailed help on this!
Autozone has a Vacuum Pump in their Loan-a-tool program if needed. Otherwise, you can get one at Harbor Freight for around $25. The cheapest way I know of to create positive pressure is with an Air Pump intended for bicycles, preferably with a gauge built into it. From there, all you have to do is adapt it from a Schraeder Valve fitting.
There are other ones with 1/4" NPT threads used for Air Compressors in the store's tool section, but it can be hit or miss finding them in-store. I modified a BMW radiator cap with one of these fittings, combining it with a sub-$10 "Slime" brand 12v tire air compressor from Walmart to make a dirt cheap Coolant Pressure Test Kit. The compressor's air delivery volume is pretty low per minute, so there's little chance of suddenly over-pressurizing the system. Sure beats paying over $100 for a commercial setup.
To the OP: I'm not sure if you found your issue, but after discussing with Akagis_white_comet about Mazda diagnostics earlier, I could find that for S6 models, Miatas and Cosmos, the led should be connected between FEN and B+ (Rather than FEN and GND).So in case you are still troubleshooting that, you may want to try again with the led on B+ and see if the codes make more sense, as it could have been inverted (Your "light on" may have been "pause")Nico
I'm reasonably certain FDs use the same operating logic as the Cosmo in their Diagnostic procedures. In this manner, Terminal FEN/MEN/TEN pulses out 12v+ in a specific pattern to make the LED blink in a Morse Code manner whose message contents is detailed in the Workshop Manual. Pretty unlikely that the pulses could be inverted.
Other vehicles such as the Miata could work in the same manner but with flipped polarity. There is a 12v terminal in the connector. On the Cosmo, it's from the 10A Turn Fuse and the Workshop Manual says "don't short this terminal to ground, you WILL blow the fuse".
The funny thing about using the original tools (System Selector, Disc monitor, etc) is that the light is most likely an incandescent bulb so it could work either way with just a SPDT polarity reversing switch on one side of the bulb, while the other side gets connected to FEN, MEN, TEN, etc through the Selector Switch. Probably one for Rotary and the other for Reciprocating? Might be possible to replicate this with some electronics sorcery but that's another project for another day.
My issue is still a mystery to everyone I've had check the car out. Still reliably throwing the following codes:
13: Pressure Sensor Signal (MAP sensor)
14: Atmospheric Pressure Sensor Signal (Inside of ECU)
Funny thing is I can almost predict when it will happen. If it is really humid like after a rain the car is often in a low-boost condition, maybe 6 psi or so. After a while of driving it will generally act normal and both turbos boost great. When the low boost condition happens the ECU throws those codes and goes into "limp" mode requiring a reset before I can drive again.
MAP sensor tests good and was swapped with a known good one just to be sure. Car displays the same behavior with either installed. I wouldn't think low boost would put the car in limp mode. The recurring ECU sensor code has me worried I'll need a new ECU.
Something to keep in mind is that if the Map Sensor is Known Good, the next step is testing harness integrity. Wouldn't be the first time a FD had harness issues due to heat. Also, heat has a funny way of increasing resistance too. The most common scenario is increased resistance to Ground, which skews sensor voltage higher, making the ECU think the intake pressure is higher than it should be. The same can be applied to the Atmospheric sensor inside it. If the ECU ground (under the UIM) is iffy, it'll push the Atmospheric sensor closer to Zero mmHg instead of the actual atmospheric pressure reading it uses to calculate with.
Inversely, increased resistance to vRef (5v) will have the opposite effect, pushing the sensor readings toward Vacuum (closer to 0v than normal). One way to test both ground and vRef is to backprobe the two TPS Signal pins at the ECU and see if it returns the correct voltages through the entire throttle range. Throttle fully closed has a range of voltage values, Fully open does too, and it should be a nice smooth increase in voltage from closed to open.
Grab the multimeter and start testing for continuity/resistance. Specifically, where it should be almost 0 ohms but is iffy, and where it shouldn't have continuity but does.
12-13-21, 04:14 PM
![[L'e$kro]'s Avatar](data:image/svg+xml;utf8,<svg xmlns='http://www.w3.org/2000/svg' viewBox='0 0 50 50'><circle class='avatar-circle-default' fill='%2366b5a3' cx='25px' cy='25px' r='20px'></circle><text class='avatar-text-default' x='49%' y='53%' fill='white' text-anchor='middle' alignment-baseline='middle'>[</text></svg>)
