FD RX7 in "limp" mode
 

Hey everyone.
I was out of town and I let my fd rx7 sit for two weeks.
When I took it out for a spin I noticed that in 2nd gear it would not go higher than 2.5k rpms. After a little while on the throttle it will climb to 3000 and then start to sputter out at 3.5k rpms.
So far I have tested:
- the throttle cable is tight
- when fully on the throttle the valves open
- my tps sensor is reading slightly out of spec on fully throttle the 2nd pin is 5.04. So .04 out spec. My readings are when the car is warm. When the car is cold I'm not able to get the readings in spec
- when in neutral I can rev up to whatever rpm I want without an issue.

What should be my next step in debugging this issue?
Start checking for vacuum leak in the vanos system or get the tps in spec?

If its US spec, check codes

welcome to the board. :)

my money is on a malfunction with your MOP, but as stated above, your first step should be retrieving the codes.

is this some kind of Freudian slip on your part? there will be none of that talk around here. :)

This behavior is typically "Artificial Rev Limit under load". Before digging in too deep and creating a bunch of additional variables, I'd suggest starting with the basics. Are ALL of your grounds Known Good? Same with the main power cable too.

I'll spare you the in-depth jargon, but an iffy ground connection can skew sensor voltages such as the TPS HIGHER than they normally would read. If it is the Main Ground, it will also impair anything that receives its ground connection through the engine shortblock such as the ECU. Injectors and Ignition signals are quite susceptible to this because they are ground signals from the ECU. So the car behaves like it's running a marathon with its shoelaces tied together. On a FC, this comes out as the "3800rpm hesitation".

More details about the ins and outs of grounds can be found here: https://www.aaroncake.net/RX-7/grounding.htm

Going to read the codes, I have a JDM model but I have seen videos where you just need a single led so will attempt to that.
I didn't start off with that since I was misinformed and told that I couldn't read codes on these cars.

There is quite a bit of misinformation running rampant about anything rotary. Even moreso if the vehicle in question was only offered in Japan.
The process you are referring to still needs the issues ironed out. It was based on a hypothesis I came up with about the inner workings on the "correct" tools. However, the rhythm/behavior of the LED blinking seems to not mesh up with what is in the Japanese Manuals. Every time I saw the videos of it, it was difficult to pin down what it was saying. What appeared to be a failed atmospheric pressure sensor inside the ECU turned out to be something else as the same LED behavior happened after swapping ECUs with another one.

In other words, it's incomplete work and I wouldn't feel confident with the results in regard to a diagnosis. Kinda hard to debug from another continent...:lol:

Plus, it is wise to note that the FD has the same Middle Terminal on its Main Ground Cable that the FC and Cosmo do for the Body connection. This is where corrosion begins and is why things go goofy as Aaron Cake described, especially after 30ish years. This is where I would begin any diagnosis efforts. Is your car a Series 6 or 7/8?

So I pulled the codes and here is what I got
25, 26, 31, 32, 40, 43, 50, 54, 71, 73

Using Banzai's chart(can I used this?) it leads me to believe there is an open circuit.
Since the last two codes only have open circuit as a condition.
It' safe to assume my groundings are not correct, right?

Just so we're all on the same page, how exactly did you get these codes? Last time I dabbled with it remotely, I couldn't keep track of each code for some reason. Perhaps you've discovered something that I have not? Anyway, I checked the codes for Japanese Series 6 and here's the results:
25: Solenoid Valve PRC (Pressure Regulator)
26: Stepping Motor (OMP)
31: Solenoid Valve (Secondary Air Relief)
32: Port Air Control Solenoid Valve
40: Purge Solenoid Valve
43: Wastegate Solenoid Valve
50: Double Throttle Solenoid Valve
54: Air Pump Relay
71: Injector- Front Secondary
73: Injector- Rear Secondary

Every single one of these items relies on a Ground Signal from the ECU. And that relies on...the ECU Ground and Main Ground Cable. Follow Aaron Cake's guide to the letter and it should straighten things out. The most annoying part will be to un-tape the harness to extract the starter/ground cables from the bundle.

To get the codes I open the diagnostics port and ran a wire from GND to TEN. Then using a resistor soldered to a led I ran the resistor side to +B and then the led side to MEN,

I recorded the codes on my phone and just counted the short and long flashes to get the codes.

Appreciate the help so far!

Smart man! Think you could put the video on Youtube? It would be a massive help to all who have a Japan-spec FD, especially since the codes being conveyed are known. Makes it easier to identify each pattern.

I could do that, just took the Upper intake manifold off for the first time to get to the ecu ground located underneath. Looks clean but will attempt to clean it out. While I'm under there, is there any way to test solenoids and lines for leaks/make sure they are working.

Went through the grounds and tested them, they all seem to be working as intended.
Going to check the injectors next.
https://cimg7.ibsrv.net/gimg/www.rx7...1f94a69f3e.jpg
Main Battery Ground Terminal
https://cimg1.ibsrv.net/gimg/www.rx7...4f5177ed66.jpg
Main Battery Ground
https://cimg3.ibsrv.net/gimg/www.rx7...0405cb45c7.jpg
ECU Ground

Does my ground for the ecu look bad?

You are pulling so many codes for solenoids.

Almost seems that you may have had a "emissions delete".

Where all those solenoids have been removed and your vacuum system drastically simplified.

Problem is, the stock ecu hates that (for good reason).

Start looking into an "Apexi Power FC" to replace your stock ECU,

If you have an emissions delete, you likely have "parralleled" turbos, and your second turbo gate is likely been welded open.

It is a long road to build it back to original stock configuration and you may never be able to escape limp mode, even if you do.

There is afterall, about 75 things that can cause limp mode.

If there is a successful sequential turbo, emission delete recovery, out there, I have never heard of it.

If your solenoids have been deleted, but the PO was kind enough to leave the emissions harness intact (and hanging in a ball under your UIM) you can try plugging in the solenoids (which will cost you about $1000) even if you do not restore the vacuum lines. The ECU will be happier about that.

But remember your stock ECU is 25 plus years old. How many 25 year old circuit boards would you trust?

You might want to open up the ecu (carefully) and check the condition of the circuit boards (don't bend any pins).


(Note: A new OEM TPS may run as much as $400; but there is a continuation replacement part number. Best to look for that. The TPS can read to be in spec, but may go out of spec when the engine is hot. Best to replace so you know it is not a variable.)

What year is you car. What version? There were some major changes when you get into version 4 , 5 and 6 cars.

As versions come along, the more there is differences from the USDM factory service manual.

My car is a 1996 type rb bathurst, so a version 4.
And I'm starting to think it's the ecu being dodgey, I have a power fc just need to flash it to a stock twin setup. Will try to get that plugged in asap.
Things I have tested so far:
- my air pump relay and air pump
- checked the vacuum lines to make sure they weren't broken
- checked the omp
- cleaned grounds

I guess the ecu is the next step of the debugging.

Congratulations on the Type RB Bathurst!

Note that for Version 4 and above the location of the coils were rearranged.

If you use the USDM FSM as a guide to hook up the leads, then you may have them in the wrong order.

This might not make a noticeable difference at low RPM, but if you go higher you could blow a seal or two.

(Ask me how I know.)

Also your car should have the "rat box" instead of the "rats nest". . For Version 4, and above, seven of the key solenoids were incorporated into a single assembly located where the rats nest was.

Again it is strange that your code reading shows most of these solenoids not working.

If just one solenoid is blown, you have to take the entire rat box apart to change out that one solenoid.

Often people will just change out the whole box (about $600).

[After that you can fix the broke rat box at your leisure and resell it.)

[There is a guy on Buyee selling used rat boxes for about $50 each - good luck with that.]

This is from the 1995-12 FSM change manual. The old coil location is above and the new location is below.

(I take no responsibility if your coils are still hooked up wrong.)


Maybe Akagi can helpfully translate this page.


https://cimg8.ibsrv.net/gimg/www.rx7...6e85469a93.jpg

Sorry this pic is too small. I will try to get a bigger version. It is from My Albums in the community pics section.

You will want to google "WM4029" which is the change manual for Version 4. Maybe Akagi has this on his translation project list.

It is an important one.

Also you should get part catalogue AJFA06-00 (95-12) FD3S-400001 , and any updates of such AJFA06-00.

Again ,although in Japanese, it has exploded diagrams of all the parts and is very useful (and kind of cool).

They run about $100~$200 on Buyee.

https://cimg5.ibsrv.net/gimg/www.rx7...22ea780ba7.jpg
Memorize this.
https://cimg1.ibsrv.net/gimg/www.rx7...ab803ee883.jpg
What could possibly go wrong?

https://cimg3.ibsrv.net/gimg/www.rx7...6a34f778f5.jpg
https://cimg6.ibsrv.net/gimg/www.rx7...8aa3905b7f.jpg
https://cimg1.ibsrv.net/gimg/www.rx7...6e6b938cc4.jpg
Dismantalling not recommended, Haha.
https://cimg4.ibsrv.net/gimg/www.rx7...e635cd549b.jpg

https://cimg0.ibsrv.net/gimg/www.rx7...ad19ca74eb.jpg
Upper Version 1~3. Lower version 4~6.

Just got your message, the FD manual collection is on my to-do list once I get the JDM Library up and running. As for what's in the above picture, here's what it says in English from left to right:
Old (AKA version 1-3):
Trailing #1, Leading, Trailing #2
VVVVVVV
New (Version 4-6):
Trailing #1, Trailing #2, Leading

Give me a holler if you're still stuck.

[Caution: Old Man Story!]

So my car came with a DIY emission delete and turbo parralled, and stock ecu. Naturally it was going into limp mode all the time.

(The ecu will reset when you turn the car off, but once it runs through another self inspection cycle, back comes the limp mode.)

With a help of a young friend that had apprenticed under a local Rotary sensei we set out to reconstitute the sequential set up, which seemed simple enough.

We had a Japan OEM code reader (try to find one) and blew nine codes. Mostly missing solenoids but also the OMP. Luckily the PO had left the emission harness intact, so we attach a bunch of old solenoids and the car seemed free of limp mode. (Not!)

After that we replaced many of the old sensors including MAP, O2 Sensor, twin solenoid (the green pair) and TPS. This further helped the limp mode situation.

When we removed the front air feed pipe to the front turbo we saw the fines were chewed up.

So we replaced the turbo with a spare, but discovered the flapper on the back turbo had been welded in place (dremelled the welds off).

(New set of turbo related gaskets: $1800.)

In the meantime I had new plugs put in by a local garage. (Change the plugs and O2 sensor every 5000 km.)

Still we were getting limp mode. And running the revs up to 6000 we got a loud bang. After hitting the bang 4 times, I lost compression.

So engine rebuild.

Rebuild shop found the following contributing factors.

Leads hooked up wrong.
Oil update blocked by oil pan dented up.
O-ring in-line in-tank fuel uptakw bent.

The first work I had on the car was replacing the fuel pump. So the o-ring was likely bent the whole time. Likewise the oil uptake was restricted from the beginning. So the bent o-ring might have caused a simulated limp mode.

But it was the miss-placed leads that were the final straw.

The OMP may code on and off, if it is in the process of dying. So it may fool you.

We also found the filter for the Map sensor was facing the wrong way.

(Version 4 got a new MAP sensor by the way.)

The ECU really hates a faulty OMP, so will limp mode very quickly if it gets a bad reading.

You are so very much ahead by running those codes first.

Imagine being a shop trying to fix an RHD without the codes!

So many stories of "whack-a-mole" attempts to search out the problems.

Recently some people have found a sticky IAC as causing idling/running issues.

New, about $450, used, about $100. Easy to swap out. (Gasket is NLA.)

Burnt out coil(s) can simulate limp mode.

New air filter always helps! ($15)

The power FC, cycles like crazy on cold start up, but will find its groove eventually.

I have 40,000 km on my rebuild and the car runs beautifully.

Pays to have a knowledgeable shop behind you.