To give you guys an update, I've been at this latest problem for well over a month and I just cannot seem to figure it out. All important back story before I post video.

Over the past few months, my tuner has struggled to get my idle under control. Sometimes it would be perfect, idling right at 1000 rpm. Turn the car on the next day, I'd have to keep the car alive for the first couple minutes from a cold start because it wanted to idle somewhere in the 600/700 range and it would die (engine is street ported). After a few minutes, idle would stick to 1000 rpm....until we start driving. During stops (stop signs, stop lights, etc), it would dip to roughly 800 and jump to 1000..back and forth, back and forth. So I'm reading through the forum and come across numerous threads about adjusting the throttle body. Hell yea, I'm gna fix this. Then...it left me stranded on the side of the road. I was driving along per usual...maybe 35 mph, nothing crazy. Radio goes out first, dash lights next, then the engine, and it will not start. Friend gives me a jump and we head back home. So I'm thinking alternator.

Couple of weeks go by and I get my refurbished alternator back from IR Performance. Everything looks good, sounds good, and runs good. Send tuner the logs from a drive and everything looks stable from his perspective. The very next day I get ready to go for a drive, the car won't idle correctly again. As it warms up, the idle stays around 900-1000 rpm. So I hop in and go for a drive.

-At first, maybe 2 miles down the road, the idle starts to pulsate. Its significantly worse than before...1000-3000-1000-3000...so on and so forth.

- I get on the highway another mile or 2 down the road to come around town and head back home. Take it thru the rev range. Its driving excellent, pulling harder than a mule under WOT...but I get a red light. I downshift to 3rd and I thought I accidentally rev matched. Problem is, I typically don't rev match in traffic. Shift the car from 3rd to neutral....idle is 3000 rpm. Put it in 2nd, car begins to engine break like it normally would. Back to neutral...3000 rpm. Great.

-Heres the icing on the cake. I get off the highway and I'm getting closer to my house. Speed limit is 35. Coming to a stop again with car in neutral...3000 rpm. However this time, I put it in 2nd AND THE CAR IS ACCELERATING TO 3000 rpm!! I nearly smash into the back of a minivan because it caught me off guard. So now no matter what gear I'm in, car wants to be at or above 3000 rpm...even in neutral.

I get the car back to my garage (thank God I didn't hit anything or drive thru the wall) and the car still wants to idle at 3000 rpm. Cut it off and start looking for anything on the throttle body that looks out of the ordinary. Cable isn't sticking, throttle cable mechanism is in working order, gas pedal is working as normal. I'm baffled. Plug up my computer and this is what I see. Keep in mind, car is not running...key is just in the on position

https://www.youtube.com/watch?v=rX_v...ature=youtu.be

I've added grounds, I've cleaned grounds, got a new TPS and I cannot figure this out to save my life. The worst (and probably most important) part about it is even while looking at the screen of the video above, I would unplug a connection and nothing would change. What could possible cause that?
Your help will be MUCH appreciated. If I'm missing any important details, please let me know. Thanks.

 

Disregard my post, early morning and I missed that you already checked what I was thinking.

 

It was hard to see the specifics in the video, but I gather you're running an aftermarket ECU with a DBW throttle, correct? What kind on both counts, and what are you using for the accelerator position sensor (APS) and how is that mounted?

I'm running a Link G4+ ECU with a GM DBW throttle body, and RX8 accel pedal, and with that setup if anything goes out of whack electrically (e.g. the 2x TPS sensors don't track close enough, your DBW PID tuning is jacked up, etc.) the Link's safety fail safe logic will put the DBW into a limp home mode - it won't rev beyond ~1500RPM and it will buck like crazy when you try to give it gas. I'd assume whatever ECU you're running has similar safety fail safes, so my best guess is there's a mechanical issue with the APS - perhaps it came loose and is hanging up mechanically instead of returning to an idle/zero throttle position?

 

He said it's a mechanical throttle. What ECU is this? The only way it can idle at 3000rpm on a mechanical throttle is the throttle valve sticking open (dashpot or fast idle cam for example), idle air control valve duty too high, or throttle cable/cruise control issue. That's assuming no vacuum leaks. You should be able to take logs and look at TPS voltage and idle air control duty cycle to see what's going on. Spark timing can affect it too of course.

 

Ok, I missed that detail - from what I could see in the video, I could tell it was some sort of aftermarket ECU, and I assumed he was running a DBW setup because the video shows what looked like DBW settings.

 

I looked at the video. It's a Haltech Elite 1500 it looks like, and that's the Haltech software. It's entirely likely the tuning of the idle air valve is off. basically without the Haltech map file and datalogs there's not much we can tell OP besides very general advice.

 

Read this -

https://www.rx7club.com/3rd-generati...ncing-1147118/

If you have a high idle, especially 3000 RPM, there's air getting into the engine that shouldn't be. Period. Either the throttle is stuck open or there's a hole of some kind letting in air between the throttle plates and engine. It's not a ground, it's not an electrical thing, it's air. Idle is set by the air you give the engine.

Start with that and track that down. That could also have been something that was going out for a while (like a block off plate coming loose) and that was causing the idle weirdness.

Dale

 

I had a minor issue with the secondary throttle plates sticking causing a hesitation during light-throttle acceleration. It was caused partially by carbon buildup in the throttle bores. So it occurs to me that since your issue is intermittent, you may be having a similar, but more serious version of the same problem. Intake manifold temperature changes and thermal expansion, etc., could either relieve or exacerbate the issue.

So I would check to make sure it isn't sticking slightly open - it wouldn't take much to cause a 3K-RPM idle, and might not be immediately apparent on inspection. The 1st thing I'd look at while it's doing it is whether the secondary throttle stop is in firm contact with the lever that is attached to the secondary throttle plates.

 

One solid test on the TB is to remove it from the car and shine a light behind it. If you see light around the throttle plates, they aren't totally closed.

The fast idle cam will hold the throttle plates open, you do have to manually kick that down to close them. Also I have seen cases where someone removed that cam and the linkage was still there and jammed up/got stuck and kept the TB open.

Dale

 

With key on, engine off, record a datalog. Look at TPS raw voltage if you can. while taking data, go under the hood and move the throttle plates as well as the fast idle cam (it can be rotated away with your finger). if you see that the TPS voltage is decreasing when you physically try to close the throttle plates more, then it's possible you are sticking under normal circumstances. I've seen the physical dashpot on the throttlebody stick and not close smoothly.

 

Thanks for everyone who responded. Today I made sure to update my Haltech ESP to the latest version, update the Haltech 1500 firmware to the latest version, and in doing so I wanted to disconnect everything from the UIM and TB to show you all what issue I'm having. Note: car is not running, key is in the on position. Video posted below, then I'll start individually responding to your replys. Thanks again.

 

Thanks for the response Dale. I actually saw and saved that post for reference when I was initially thinking it was an idle issue. However, my mind was changed while I was trying to recalibrate the TPS. Is it possible a mechanical issue could cause the electrical system to go haywire?

I definitely understand what you're getting at, because thats the route I was chasing when I thought it was just an idle issue. I'm still going to change out the dashpot (is it possible to just remove it?) whenever I can get everything stabilized. In the video you can see the TPS and IACD voltage is all over the place...even with the TPS being disconnected. All of this is happening with the throttle cable disconnected.

Sorry about that Pete. After your comment, I went back and watched the video so I can see how that can be confusing. I'm not running DBW and I'm running the Haltech Elite 1500.

Will the data log tell me anything I can't already see regarding voltage to the TPS? If you watch my second video, you'll see that my TPS is disconnected and I'm still getting crazy voltage variations from the ECU regarding the TPS. I am completely stomped.

 

I've had similar issues with setting up my 1500. If you disconnect the IAC valve while your idle is up there does it drop? If so there's your culprit. Take it out, give it a clean and see if that helps. I'm suspect of mine sticking open or closed. The other thing to play around with is the frequency setting on the haltech for the IAC.

As for the voltages moving around, could be a noisy earth signal to the ECU?

 

I wouldn't worry about those voltage fluctuations, it looks like the reading in your second video is between 0.135V to 0.138V which is probably less than 1% throttle position. I wouldn't expect to see that reading with the sensor disconnected, but I'm not familiar with Haltech input circuits.

Assuming this is not a simple vacuum leak, my next steps would be the following:

1. Learn how to configure your ECU's datalogging, so you can record the throttle sensor raw voltage, idle duty cycle (final result including any idle feedback settings), engine speed, battery volts, MAP, CLT, and IAT. In case there is a harness problem, watching for jumps or spikes in the IAT/CLT signals can help troubleshoot and identify some problems (CLT/MAP/TPS/IAT are likely to share the same sensor ground wire, but only MAP/TPS are likely to share the same sensor 5V supply wire).

2. Double-check the throttle sensor by connecting the throttle sensor and taking another video while slowly sweeping from full closed to full open with the engine off. Get familiar with viewing the ECU datalogs so you can watch to make sure the trace is 'smooth' without weird jerks or steps or jumps. If you press the throttle three times while datalogging, you should see three smooth 'bumps' in the throttle sensor reading.

3. Once you have datalogging working, connect all the sensors and then record a datalog while you tap/shake/bump the harness wires near each sensor and near the ECU. Watch the datalogs to see if any of the signal readings change.

4. If you don't find anything when shaking the harness, drive the car while datalogging and try to capture a log of the problem happening. Of have a passenger record a video of the laptop screen while the problem happens and hope the camera is pointed at the relevant readings.

5. Might not be related, but the ECU's indicated BattVolt value was 11.7V near the end of the video. Check with a multimeter to see how closely this matches your actual battery voltage. I would charge the car's battery if it's actually lower than 12.2V when sitting, if you don't already have a trickle charger it's worth getting.

 

If the TB is cracked open slightly and you calibrate the TPS for that position you'll have weird things happen. When the TPS reports the throttle body is closed it HAS to be totally closed.

Dale

 

I'd recommend as many have, that you should check your throttle plates and check for any vacuum leaks. It is always easier to check for the mechanical issues first then move on to the electrical.

You should next recalibrate your TPS sensor. There is a step-by-step calibration in the Mazda FSM.

When I first took delivery of my RX-7 it was having idle and backfire issues. Immediately recognized it as being a TPS issue and once I had the PowerFC installed, I recalibrated the TPS. All the problems went away.

 

What you are seeing is normal electrical noise. A tiny voltage fluctuation like that is not indicating a problem. Now there may still be something going on, but from the video, I don't see any obvious electrical issue.

You're just on the learning curve for how ECU tuning works. You basically got thrown into the deep end. This isn't a modern car with a stage 1 reflash/"chip" tune.

 

I'm the tuner here.

This has been an odd one. A good number of months ago I remotely tuned the car and it was cranking/cruising/boosting all cleanly and consistently across the entire range. Fast forward a month and it starts having this odd issue where the idle wanted to stumble, but only during warmup. I revisited all the settings and tuned it back to stable several times through modification of the ISC settings, however with time it would always seem to revert back to the stumbling after some time and THEN on the same file it started having the issue where the idle would 'hang' high.

To me, it isn't an issue with the ISC as that valve simply doesn't flow enough to push it to a 3000 rpm idle even at full flow given the normal no-isc active idle is right at 1000 rpms. I've thought that there has to be something in the mechanical throttle linkage that is sticking OR a massive vacuum leak somewhere.

Skeese

 

need datalogs of the sticking idle, that's the next step.

 

I re-watched your video today, and it almost looks like your car still has the factory thermowax / fast idle cam. If you've still got that installed, it's basically a variable throttle stop depending on the coolant temperature of that thermowax valve. Learning how to use the ECU datalogs would help catch this, the calculated throttle position should be about 4-10% higher when the thermowax is cold. If you accidentally calibrated the ECU's throttle sensor when the thermowax valve was cold, the measured throttle position might do weird things when the engine is warm and you're pressing the throttle very gently.

 

Update:

I thank all of you for your input. I took the TB off once again to reset everything to make sure there was nothing mechanically wrong with the TB itself and to confirm that my TPS is aligned correctly. We're about to take another stab at it so we'll be back with you momentarily.

 

Check any block off plates. I've been having an odd issue with idle for a while. Turned out I had a leak at the ACV. Took a long time for the leak to become big enough to locate.

 

So we had an interesting break today that ultimately led me back to square one...but pointed me in the right direction.

So I called Skeese on Facetime earlier and we narrowed it down to an electrical issue.

First I tried starting the car to get some logging data...wouldn't start. Last time I tried it, it started but it was really rough, so somehow I've gone backwards. Called Skeese and we tried to figure out what was happening when he suddenly had a eureka moment. So to rewind a few months, this past summer I pinned the WBO2 gauge into a male connector to plug into the female WBO2 connector off the wire harness. After looking over my wiring allocation, Skeese discovered the voltages didn't match up. So with the key still in the on position, I unplugged the connector and boom....everything stabilized. So we came up with a game plan, decided to de-pin the ground, relocate it to a chassis ground, plug the connector in and see if that fixes it.

Thought we'd found the problem? Wrong again. After I hung up the phone and de-pinned the ground, relocating it to a chassis ground I tried again. Turn the key on and everything started going haywire again. TPS is fluctuating between 0-100, just like in the videos. Voltages are going crazy across the board for all of my sensors...BUT...something I remembered from our conversation, Skeese mentioned that all of the sensors that were going haywire were sharing a common ground...the ECU ground. The only issue I'm having now is now that the ECU ground and WBO2 are not sharing a common ground, shouldn't one or the other be more stable? I know this is hard to fathom without seeing the issue in person for you guys, but can some of you point in the direction of your aftermarket grounding setups? Since this problem has reared its ugly head, I've added approximately 4 chassis grounds in hopes they would resolve my situation and this has not been the case. Should I remove them all and start from scratch? What do you guys think?

 

So a few months back we completely disconnected the IAC valve because we couldn't get it quite where we wanted it with stabilizing the idle (now that this problem has revealed itself, it makes sense). As for the noisy earth signal to ECU, that is seeming like the likeliest solution at this point. Where did you locate your ECU ground? I though the old ECU mounts would be the idle location; now I'm starting to question that decision.

So I understand that a little variation in voltage is normal, but as you indirectly pointed out in your 5th point whatever is going on is draining my battery. Usually when I'm done messing with it, I'll disconnect the better. And yes I need to invest in a trickle charger like yesterday. Regarding your other points;

1. I've done that already under the guidance of Skeese. I've gained a significant amount of knowledge regarding the configurations of the ECU. Now I haven't gotten into the tuning aspect of it all just yet because I feel as though the rotary is too delicate for a novice like myself to toy around with, but one day when the car is running fine, I will.

2 & 3. So I'm not sure if I've already mentioned this, but my TPS is brand new. The problem is every time I calibrate the TPS, the voltage values on 0 and 100% change. One time they might both be .02, and this last time they were both 0. However I like your idea of shaking the harness around near various sensors and the ECU. I'll have a friend come over this weekend and give me a hand with that.

4.If only I could get her to start.

I've saved a couple of your write-ups from over the years regarding idle settings and whatnot. The issue is every time I calibrate the ECU, I get a different voltage. I feel as though the TPS voltages itself fluctuating the way they are is a direct indication that it is an electrical issue rather than mechanical, like I was hoping. However I did follow your advice and remove the throttle body to make sure there was no daylight coming from behind the throttle plates. I got nothing, so I reset the TPS again and zero'd out all of my previous idle settings just to verify nothing was mechanically going wrong.

Done and done. While I'm not ruling out any vacuum leaks, I feel as though we have to tackle these electrical issues first because now the car won't start and every time I calibrate the TPS, its getting progressively worse. Just like in the video (I'm not sure if I recorded that part, or mentioned that the car was not running in either video) the TPS values are jumping from 0% throttle to 100% throttle and everything in between while I'm not touching the throttle at all.

I totally understand what you're saying, and you have a point with your 2nd comment. Back to the first, if you read my sentence in the post directly above this one, we got the voltages to stabilize for about a solid minute before they all went crazy again. The car refused to idle before because of the voltages to the TPS, and it won't start now. There is no calibrating the TPS at this time, its just not cooperating.

Thanks for the advice. Reading this made me do a bit of research regarding elimination it and now I'm on the hunt for a block-off plate. For the time being I just took the adjust scress spring out and screwed the screw all the way down. Seems to work for a few people so hopefully it'll work for me for now.

Noted. I'll be sure to do that when I get another chance.

 

I'm of the opinion that there's more than just one issue here however pins 14, 15, and 16 should only be interconnected with themselves and directly to the battery.