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've been working on getting my FD back to complete functionality, which I've thankfully made some strides in recently.
However, I fired it up for the first time in a while, and realised pretty quickly it really doesn't like the midrange RPMs. It hesitates quite heavily at 3000-4000 RPM, doing a couple of small backfires, but then goes from ~4500 to redline without issues.
It idles perfectly fine, and doesn't seem to run rich or lean when it's at operating temps.
For the things I've at least thought might make a difference, I've listed, but I'm just confused by how specific the hesitation is.
I've changed the air-, fuel-, and oil filter, changed the oil, filled it with 100 octane, and the other basic maintenance stuff.
Compression tests were 117-120 on both chambers, and the engine and electronics are bone stock, so no changes to the ECU or anything like that.
As far as I can tell the MAP sensor works fine, but I unfortunately don't have much knowledge about those, compared to MAF sensors.
The only other thing I can think of would be a vacuum hose leak, as I'm not sure if any of them have been changed previously, but does that only affect a specific range of RPMs?
If it helps at all, it's the 99' 261 BHP RB spec.
Thanks in advance!
P.S. Apologies if this has been seen before, but I couldn't find anything on a stock engine setup, at least not at first glance
I've been working on getting my FD back to complete functionality, which I've thankfully made some strides in recently.
However, I fired it up for the first time in a while, and realised pretty quickly it really doesn't like the midrange RPMs. It hesitates quite heavily at 3000-4000 RPM, doing a couple of small backfires, but then goes from ~4500 to redline without issues.
It idles perfectly fine, and doesn't seem to run rich or lean when it's at operating temps.
For the things I've at least thought might make a difference, I've listed, but I'm just confused by how specific the hesitation is.
I've changed the air-, fuel-, and oil filter, changed the oil, filled it with 100 octane, and the other basic maintenance stuff.
Compression tests were 117-120 on both chambers, and the engine and electronics are bone stock, so no changes to the ECU or anything like that.
As far as I can tell the MAP sensor works fine, but I unfortunately don't have much knowledge about those, compared to MAF sensors.
The only other thing I can think of would be a vacuum hose leak, as I'm not sure if any of them have been changed previously, but does that only affect a specific range of RPMs?
If it helps at all, it's the 99' 261 BHP RB spec.
Thanks in advance!
P.S. Apologies if this has been seen before, but I couldn't find anything on a stock engine setup, at least not at first glance
Sounds like a turbo control issue (solenoid/check valve/vacuum lines/etc), and right in the middle of the sequential changeover. This is a N3G2 ECU, correct? Been a bit since I looked at the 99-spec manuals, but I do know that 96+ ones have the "black box" which has everything turbo-related in a tidy little package. As a result, the vacuum diagram is completely different from 92-95. I'll take a look this weekend and see if I can dig up the 96+ diagram to get you in the right direction.
That sounds a lot like it matches the problem. I haven't had it over 1st gear, as I haven't had space for it, but if memory serves, the hesitation was worse when it was actually moving.
It should definitely be the N3G2 ECU, but I'll try to check next time I see the car, as it's a little ways away.
Last edited by Sumerei; 07-20-21 at 04:19 PM.
Reason: CPU -> ECU
That sounds a lot like it matches the problem. I haven't had it over 1st gear, as I haven't had space for it, but if memory serves, the hesitation was worse when it was actually moving.
It should definitely be the N3G2 ECU, but I'll try to check next time I see the car, as it's a little ways away.
Moving=Engine under load. That would generally make any symptoms more pronounced as you discovered. My memory is a tad fuzzy on the exact RPM points for when each part is triggered in the turbo control system when it changes over from feeding only the front turbo to feeding both turbos, but 3000-4500rpm covers the whole changeover on both FD and 13B/20B Cosmo. This might sound like a dumb question, but do you have a boost gauge? Stock FDs have a 10-8-10psi boost pattern. 10psi up to the changeover, 8psi during changeover, then 10psi afterwards. It would help to see if it is behaving correctly, or if boost is acting funny in the RPM range you're having issues in.
You're absolutely right, I realised pretty quickly after posting it that I was just being silly about the engine load.
It does have the stock boost gauge, so I'll check the pattern of it as soon as I can, most likely on the weekend.
You're absolutely right, I realised pretty quickly after posting it that I was just being silly about the engine load.
It does have the stock boost gauge, so I'll check the pattern of it as soon as I can, most likely on the weekend.
Like other FDs, boost should ALWAYS be verified with an external gauge. Although I've not verified the specifications for the stock boost gauge yet, and this is up for debate/verification, I would wager it is calibrated similar to the Coolant Temperature Gauge with 3 positions: Low, Normal and "Buy a New Engine". Even a $20 mechanical boost gauge from Ebay is better than nothing here. Might not be the prettiest, but a tube-fed mechanical gauge taped to the dashboard for testing is more peace of mind than hoping the stock gauge is behaving correctly.
Very belated hi,
It's managed to take an incredibly long time to diagnose at least part of the issue (mainly due to time constraints), so I apologise for the huge delay.
I'm not actually sure if anyone will receive this reply, but regardless, the wastegate & turbo precontrol solenoid (N3A1-20-285) had snapped on all 4 plastic buds, along with half the vacuum lines going to it.
So we replaced the lines, along with most of the other upper rubber vacuum, with a silicone set and a new solenoid. That fixed a good deal of it, but far from all of it.
Both turbos are completely fine, the vacuum has no leaks whatsoever now, and the engine itself is about as healthy as it can be.
However, the hesitation is still there, albeit at a lesser degree, at somewhere in the ~2-4k RPM range under load. I also somehow only later realised the "redline" before it would backfire was at a measly ~5.5k RPM (Odd thing not to notice, I am well aware).
Following that I replaced the '99 MAP sensor, as it was reading very low voltages compared to factory, at least for the earlier sensor years.
But in my infinite wisdom I bought a new '93-'95 sensor with the part code N3A1-18-211, and not N3F1-18-211, as the original had been.
We figured there wouldn't be a difference in the way it registers boost (asked at Atkins Rotary about it as well), but what it did instead was change the "redline" to ~7k RPM, make it run rougher cold, and the same at idle and under load.
All in all my question was whether there's a significant difference between the two MAP sensors? We noticed the older spec N3A1 sensor was larger, with a smaller vacuum opening, but I'm having a fairly high bit of difficulty figuring out whether the voltages are different.
Any help with it would be greatly appreciated, but if this thread has run it's course I'll post about the sensors specifically instead.
Watch out for the coils being hooked up wrong. They changed the position of the coils for the S8 (and probably s7). But if a mechanic is using a USDM FSM, he might hook up the coils to the wrong plugs.
If you can find someone with a JDM code tester it would be best to run "the codes",. There are 75 or so codes.
Run a smoke test to see if you have any more leaks from missing gaskets, cracked solenoids, leaking intake pipes, etc.
Best stick with the S8 MAP sensor. Make sure the little arrow on the MAP sensor filter is facing the right direction.
Change the o2 sensor.
TPS could be going out of range, especially when it gets hot. (Note little TPS screws like to strip, or drop out of site, in the awkward space. I use a small ratchet with a phillips head bit.)
Check vacuum line to fuel pressure regulator. The regulator needs vacuum on to work properly.
Very carefully open up th eecu and check that the circuit boards are prisitine.
Check for any bent pins on the ecu. Don't bend any pins on the ecu when you reconnect it.
Check and clean all your grounds.
On the twin solenoid that you replaced lines to, make sure the vacuum lines are indeed connected in the right order.
I've looked around for it a fair amount without much luck, but is there anywhere I can see how the coils are meant to be positioned for an S8? It's only been serviced in Japan so far, so I don't think the replacement I did for that and the solenoid messed anything up, but you never know...
Other than that, I tested the vacuum after replacement, and it was finally holding completely tight, so that should at least be fine.
Regardless though, thanks! I'll double check those points when I next see the car (within a week, and not 4 months), look for a JDM code tester, and give an update on what I've found.
This is from the the 1995 change manual for the JDM FSM. It is the changeover to 16 bit system from 8 bit system. The diagram shows the old S6 coil positioning on the top and the new set up below. The pair of "leading" coils moves to the right side, rather than being between the trailing coils in the old set -up. In my case I think my local shop, when giving new plugs, likely looked up the hook up in the USDM FSM. It cost me an engine.
Thanks again for your help, and the car's at least sent me in what I pray is the right direction.
Ended up using the LED bypass to read the codes, and only one came up every single time, even after resetting the ECU - 45, for the charge control solenoid.
If my memory serves, I changed the vacuum line without breaking the buds on that solenoid as well, so if the car's correct in its code, the coil inside it is probably fried.
I read through and saw that all the turbo control and emission solenoids should be identical through every gen (barring the "special" 3, and at least from an S6 and earlier), but is that true for the S8 as well?
The S6 pre-control and wastegate solenoids were identical, but the question is what solenoid I should get to replace the CCS? 1481-13-240a looks like the right one, but I must admit I'm treading deep water on the subject.
The coils were correctly placed, the sensor filter was in the right direction, the fuel pressure regulator was good, and I'll be replacing the O2 sensor soon.
I haven't had the ECU out yet, as I'd rather leave that for a last resort, especially considering how much of a pain it would be if I ended up breaking it.
Also found out the previous owner intentionally removed the CEL in the dash for unsuspecting remote auction buyers like myself, so that was a bit rude..
There may be space for a CEL, but there never was a CEL. Limp mode is the method. It is designed that you take the car to the dealer to get codes read.
Have you tried Nengun or RHDJapan for the replacement solenoid. Solenoids seem to run about $80 Canada.
Google search the part number and see what comes up.
I have a few scrap "rat boxes".
if you are unsuccesfull finding one, I might start pulling one a part for parts.
Note: I have read on here that there is a history of new "solenoids" failing right out of the box. Depends how long they have been rattling around in a parts bin somewhere.
[The wise guys, who did a DIY parrallel turbo and emission delete on my S8, luckliy left the harness intact in a neat bundle under the UIM. When we first ran codes, we got nine hits. We plugged in a bunch of S6 solenoids into the harness without otherwise reconstituting the deleted items. (i.e. no vacuum lines hooked up ) This seemed to fool the ecu into not going into limp mode. So it seems the ecu is checking for electrical function and not diagnosing whether the vacuum lines are hooked up or not. (It may otherwise sense something is wrong with the way the car is running, I don't know.]
Also be aware that the ecu assumes a certain amount of bypass air is entering the exhaust stream before the O2 sensor, at certain stages of operation. A common delete mod is to remove the air control valve. If this is done, the stock ecu will be misinterpreting O2 sensor readings.
(It is good to have a wideband afr gauge!)
A code reading will not identify any air/vacuum system leaks. This is why I recommend a smoke test. Our local shop charges $50. (or at least pre-covid they did.)
We assumed we had everything buttoned up, but the test found four leaks. (Note: If there is a single gasket left in your driveway once you are done, find out where it belongs.)
Did I mention checking the little c-clips that hold your actuator arms onto the turbo valves arms? (Have spares before checking.)
Got yet another wonderfully slow update, and a correction of the main issue.
We did the smoke test again, and the whole thing is still completely sealed, so if there's anything wrong with the solenoid, it's only electrical. Got a new one, but haven't fitted it yet, so hopefully it works as intended.
Finally replaced the O2 sensor, and got a proper S8 MAP sensor fitted. It runs a fair bit better cold, but like all the other things replaced, hasn't fixed the main issue. The air control valve thankfully hasn't been removed, so that shouldn't cause any problems.
What we recently noticed was that pumping the accelerator pedal relatively quickly, at least from a standstill, would produce pops and crackles every single time, regardless of heat and RPM.
It's also inconsistent in how badly and when it backfires at constant throttle, which we're guessing is a slow buildup of fuel causing one huge exhaust backfire.
How I neglected to mention all this previously I'm not entirely sure, but the albeit very uninformed guess I have, is that the actual main problem is a fuel/air mixture mismatch when on the throttle.
There's still no smoke whatsoever from the exhaust, so it's at the very least not burning undue amounts of oil or coolant.
We checked the TPS, and it was within spec across the board, so that's at least one thing crossed off.
Regardless, my question now was what can cause a fuel/air mixture problem for it? We're gonna check and clean the fuel injectors for starters, but are woefully unaware of what else could cause a rotary to run this way.
Still haven't opened the CPU or checked the actuator arms, as it would seem odd for them to be the cause of the "new" issue, though I obviously can't say for sure.
Got yet another wonderfully slow update, and a correction of the main issue.
We did the smoke test again, and the whole thing is still completely sealed, so if there's anything wrong with the solenoid, it's only electrical. Got a new one, but haven't fitted it yet, so hopefully it works as intended.
Finally replaced the O2 sensor, and got a proper S8 MAP sensor fitted. It runs a fair bit better cold, but like all the other things replaced, hasn't fixed the main issue. The air control valve thankfully hasn't been removed, so that shouldn't cause any problems.
What we recently noticed was that pumping the accelerator pedal relatively quickly, at least from a standstill, would produce pops and crackles every single time, regardless of heat and RPM.
It's also inconsistent in how badly and when it backfires at constant throttle, which we're guessing is a slow buildup of fuel causing one huge exhaust backfire.
How I neglected to mention all this previously I'm not entirely sure, but the albeit very uninformed guess I have, is that the actual main problem is a fuel/air mixture mismatch when on the throttle.
There's still no smoke whatsoever from the exhaust, so it's at the very least not burning undue amounts of oil or coolant.
We checked the TPS, and it was within spec across the board, so that's at least one thing crossed off.
Regardless, my question now was what can cause a fuel/air mixture problem for it? We're gonna check and clean the fuel injectors for starters, but are woefully unaware of what else could cause a rotary to run this way.
Still haven't opened the CPU or checked the actuator arms, as it would seem odd for them to be the cause of the "new" issue, though I obviously can't say for sure.
My apologies for the lack of detail in this post, I'm just waking up right now and have an idea. Pretty much everything in a FD is Ground-referenced, meaning the ECU is looking for "X ohms to Ground under Condition Y". If there is additional resistance introduced, it will skew the voltage readings HIGHER than what the ECU is supposed to see. Therefore, the ECU will think the conditions are skewed too, such as throttle position being more open than it actually is or the MAP Sensor indicating higher pressure than actual. Results in it putting more fuel in than normal, and possibly backfiring like you're seeing now.
The MAP Sensor's ground is where I'd check first, along with the Negative Battery Cable. I'll take a look at the Wiring Diagrams later today for anything else that would produce similar results.
So I did some digging and found that both TPS and MAP sensors are grounded THROUGH the ECU. ECU then is grounded at Point #11 and #12 under the hood, though I'm not sure which is relevant to this matter. If in doubt, address both of them.. Both #11 and #12 are under the UIM, similar to a FC, and subject to the same failure conditions Aaron Cake described in his article about Grounds. Here are their locations.
Since both are bolted under the UIM, their ground path relies on the Battery Negative Cable, which has a known failure point due to the Center Terminal bolted to the Strut tower. FC and Cosmo are the same way. Follow Aaron Cake's repair guide and, assuming no other issues going on, it should straighten right up.
If it does not have a proper vacuum feed it could cause fuel pressure increase..
If vacuum is Ok, although robust, the unit could still be defective, again causing higher fuel pressures.
You have got an AFR gauge, right?
(One clever DIY'er simply clamped an afr sensor in the end of his tailpipe. This might work, if you are not yet prepared to mount a second bung in your downpipe. If you do mount the afr sensor in the downpipe, mount in a way that condensation will not pool in the bung.)
So I yet again haven't gotten much done outside of the engine grounds (one of which was loose, but didn't seem to change anything), but I have a question/rant about the whole thing, before I continue to whack-a-mole the problem..
It runs badly when it's cold, on the throttle, at high rpm, and on the turbo switchover, though that got a bit better with the vacuum fix-up.
What I just don't understand is that the operating temp idle is perfectly fine? The second it's at about 800 RPM, it purrs without a single hiccup for the duration.
I can't wrap my head around how that works, when I would imagine things like a carbon build up, or a vacuum failure related to the AFR, would cause it to constantly pop and hesitate, not just under specific circumstances.
How would that even be possible? It feels like it's entirely something AFR related, but that doesn't explain why the cold idle runs that badly, and then suddenly works just fine warm, until it's on the throttle.
I do also desperately need an AFR gauge, but that might be after I give up and bring it to someone who might know how to repair the damn thing.
I apologise for the rant, but it's absolutely infuriating to me that I'm doing something wrong in all of this, explaining it terribly, and that I'm probably not seeing something obvious.
You're absolutely right, still have the charge control solenoid to replace, but as I found out while cleaning grounds, the solenoid housing is a whole lot less modular than I had first hoped.
From what I understand it should help with the turbo switchover, but not much in the fuel ratio department.
It's got the original N3G2 ECU, which I still need to get open for a clean/checkup, albeit very reluctantly with my luck, but it's worth a try if I can find another functional one.
07-20-21, 08:16 AM
