Sudden rough running [resolved]
 

So a few times I have been sitting at idle, everything is fine, then all of a sudden the engine will start idling very roughly, then die. Revving the engine does not help, it still runs rough then dies. Once this happens, I cannot get it to idle well after restarting the engine unless I completely shut down power (turn key completely off) THEN restart.

Some things that may have triggered it:
1. LC-1 oveheating and throwing a E8 code (sensor too hot). This happened once, but another time when I had a rough idle incident the LC-1 was fine, no error code.
2. Saving a map file to the lap top. I was saving off a map after doing some driving while sitting at an idle. When I hit save, the car did the rough idle right after I hit the save button. When I shut it off and restarted, it was fine again.
3. I thought the rough running might have been due to bad AFRs at idle speeds below 1000 rpm but I think I have that licked.

Anyone have any ideas?

I'm battling the same problem right now. One time up to the light, idles at 14.0. Another time it is lower, like 10.X. Another time it will die. You can't catch it with the throttle often either.

I've been working on idle and trying to balance a good lean idle with a richer idle to prevent the tip in problem. I don't know if they are related, but I know I have been trying to idle richer as it does help the tip in a bit. If I could get these two problems fixed, I'd be golden.

Yeah I have been enriching the mixture at idle too to keep it from hunting, and to be more stable. Right now I am at about 12:1 with no load (no fans or ac) and about 14:1 with load (fans and/or ac). At first I thought that was it but now I am not so sure.

On a different subject, have you played with accel enrichments yet?

Crap, I didn't even think about electric load. Sometimes the fan is on, sometimes not.:facepalm: I need to investigate that. I haven't adjusted anything in the closed loop idle control for exta effort for electrical load. I did play with the recovery RPM and I could make it catch the dies but it was a super unstable idle.

I tried to work with the asynchronous settings which, according to the manual, is supposed to be like a power jet on a carb to get it off idle. I've tried everything from 50, to the 100 it was set to, up to 255ms which is the max it will allow and it doesn't seem to change anything at all.

I need to try that. My idle drops 300-400 rpm when the ac comes on. Is it pretty clear in the manual how to adjust it, or did you do something special?

Yeah the fans I have (Spal) add a decent load on the alternator. Idle drops noticeably when they come on.

I haven't done anything with it, but in the IDLE tab, you can adjust extra RPM for electric load, along with extra effort fuel for AC, low battery.

I need to get in there and poke around more myself.

I've had something similar to me happen about 8 months ago. It was right after I had reflashed. I loaded a new map up and when I took it upstairs the the car just fell on its face. Thought I'd popped the motor. I reflashed again, and 8 months later and some 9000 miles, I've never had another hiccup. It seems like the solution to anything weird is to reflash. Once you get a good flash in there are no problems. I have feeling it is only an issue because we are using these USB to serial inputs. If it was straight serial the quality of the flash would probably be better, as something more then likely gets lost in translation along the data conversion that causes the ECU to be unstable.

I thought about trying a reflash so I'll do that next. I think there is a new firmware release anyway. I have a straight serial port on the laptop I am using so that should help.

The sudden rough running happened twice on the drive home. No commonality that I could pinpoint. Once was at idle, the other while cruising at 65 in 6th gear.

That blows that theory out of the water. I always thought it was the USB to serial interface... Maybe its the program?? Who knows, but what I do know is that if you reflash it it goes away!

OK I reflashed and verified the ECU with the same firmware version you sent out Travis. So far so good. I'll report back in if I still have the same problem. I was datalogging when it went wacky both times today so if I have the problem again I'll look at those logs closely.

Speaking of wacky, have you had any time, generally fairly early on when the car is cooler, where it will just stutter/shudder for a fraction of a sceond? Just a hiccup. It can be accelerating, at cruise. It's random. I can feel it, but it really doesn't show up on the datalog as a lean spot, injector issue, MAP pressure, anything...just an anh... and it goes along like nothing had happened.

I'm going to try a reflash as well.

I had this happen for a while. I think I figured it out though.
Changed:
Control Band from 300 -> 100
Neutral timeout from 3000 -> 500

Now it idles solidly at 900rpm. I still have some weird warm up issue, sometimes. Overall though it idles much better now. I haven't tried increasing the control band since I changed the Neutral timeout.

At some point I also increased the minimum value to 60 w/o out doing this it would try to idle low regardless of the idle target.

Yeah I have had that too. I'll watch for that as well now that I have reflashed.

I need to try that too. My idle RPM is all over the place from 800-1400 depending on what is going on. Thanks.

Do these setting have an effect when you're in slow or rapid converge fuel mode or just when in closed loop?

I know it works for slow and closed and I'm almost certain it works for rapid as well.

Stupid question but, am I supposed to be daily driving in closed loop or open loop? I've been in open loop for the last three days. The way I read it, open loop is for when tuning is "done".

You want to run closed loop for daily driving. Open loop ignores the O2 signal.

I think he meant while still tuning the map, not once the map is tuned. I ran open loop while tuning the map with the megasquirt.

OK it is still definitely doing the same thing. The flash did not fix the problem. It took me about two hours to get home because it kept doing this popping/hesitation/death thing same as described above. I changed about everything I could think of in the ECU settings, even stuff that did not make sense but tried anyway and it still did it. Coil dwell, air temp corrections, coolant corrections, turned off MVSS, etc. etc.

The ONLY common element I could find is that it only did it when the coolant temp was above about 210 deg F. If I stopped, let the car cool off, it would be OK for a bit, then do it some more once it warmed up again. Every time, the AFRs would go very lean and jump all over the place, like 17-19:1, wild swings between. When the problem went away the AFRs were steady. I checked the logs and it looks like the inj pulse is staying about the same when this happens. So I do not think it is the ECU freaking out and dropping the injector open time.

I suspect vapor lock, especially since this is a returnless system and there is no continuous circulation. Do the 99-00 systems vapor lock?

On the one hand I never had this problem until now. And, though I have seen this before, it was on carbureted engines with low fuel line pressures. I would think with the higher rail pressure a fuel injected system would be less likely to vapor lock.

On the other, the car is making more power so it is generating more heat. Plus I added a fan shroud and the fans are flowing better, thus they could be blowing more heat onto the fuel line. And, it is a returnless system. The only flow for any of the fuel under the hood is what is going through the injectors. At idle or even cruise that is not much of a flow rate.

I searched through old threads on vapor lock and did not see anything on this. I also looked at m.n in the NB forum and saw nothing there.

Oh yeah the same thing happened with the tank nearly empty and the tank completely full.

I have never heard of vapor lock on the NBs, not once. Nor have I ever experienced anything like what you're describing. Hopefully these issues don't show face tomorrow when I try to get the car running well. If I come up with anything I'll let you know though I'm still oblivious to everything adaptronic currently. Guess I should start reading.

Thanks Chad. Yeah though the symptoms point to vapor lock, I am not so sure that is what is going on. I do have a Walbro 255HP pump sitting on the shelf in the garage. I think I will install it tonight.

closed loop just makes corrections to the pulsewidths based on O2 feedback (based on AFR target). it doesn't permanently change the map.

open loop injects what's in the map regardless of AFR target.

I plan to run closed loop as a permanent solution once i'm tuned.

The base map has the fans coming on at 85c (185F) and shutting off at 60c. With the fans wired in parallel, I don't think mine's ever run that hot but summer is coming.

Yes it is already getting into the 90s here. Plus I was running the A/C.

I could have sworn that I changed that. It didn't make sense to have the fans come on before the thermostat was open. I had 93/75 but I just checked and mine is back to 60/85. I think when I got an update from Travis it defaulted back. I'm going to change it back.

I raised mine too for the same reason.

That is one of the things I tried changing this evening, lowering the fan set points to where they were basically on all the time. I still had the same problem.

how fast are your temps going up?

Good point - I'll have to change the set points.

Back to the original thread, I did some reading on the Adaptronic board and there is a detailed procedure for setting idle that I haven't done. First thing they say to do is zero the max value and adjust the idle by the bypass screw. I've not done anything like that. I've been trying to do it all by ECU.

Here's the thread:

How to tune idle

I'm way over my head when people start talking about open loop idle and closed loop idle when I don't really even know the true difference between the two.

Letme email Rob and see what he thinks, there has to be somethign weird going on. Maybe a triggering sensor isn't reading correctly, a long shot could be the adaptronic itself is not functioning properly, but I've never heard of a failure...

It's gotta be tuning if a couple of us are having the same problem. I was on their board today and there are a lot of threads about idle tuning problems. I think that we just haven't had the "right" procedure. I'm planning on trying the one that I linked this weekend, but I'm still blind leading my blind-assed self.

if I ever get mine running, we'll be able to rule out if it's your idle tuning. I plan to let the mazda ecu run idle.

Keep in mind this is just not happening at idle, but also at cruise, so I do not think now it is just an idle tuning issue. I drove all the way into work today without changing ANYTHING (I did not install the Walbro) and it ran great. Idle, cruise, WOT, all OK. No hiccups anywhere. I was not running the A/C but I drove the same route.

So far the common element seems to just be the coolant temp when it gets over 205F (96C) or so; this morning it was cool outside, and I did not run the A/C, so the coolant temps were lower. So either there is something in the code that flips out when the temp gets over a certain point, or there is a setting that triggers the rough running at that coolant temperature point. Or it is non-ECU related, like I have vapor lock, or the coils are overheating. But, if it was the coils overheating (and misfiring?), it seems like the AFRs would go down instead of up since there would be more fuel in the exhaust.

Thanks Travis.

Speaking of sensor wierdness, I noticed yesterday while sitting in a parking lot trying to figure out what was going on that the MVSS was doing something unexpected. When sitting still, with the car in neutral, if I revved the engine, the MVSS would climb and fall. That is not what I would expect to see, since I thought it is tied to the speedometer sensor input.

O2 sensors sniffs just that...oxygen.

Along with the unburnt fuel you also have all the oxygen that didn't combust with the fuel...gauge would read hella lean.

^^ Ahhh OK good point maybe you are right. So that could be it. Would it not be strange though for all of the coils, or at least two or three of them, to all start misfiring at the exact same time? I wonder if the wasted spark batch firing (in pairs) could do it. If one misfires, do both in the pair misfire?

UPDATE:

Today going to lunch I noticed that the car was running better off idle. Still spiking, but only to the 15's and not stumbling instead of pegging at 17.9 and stumbling. It went this way for a couple of miles. Didn't die at stoplights. On the way home I was fully hot and it died at a light and the lean stumble was back. I didn't have a laptop hooked up so not sure of WT, but it was likely warmer than when I went to lunch. So, maybe more to your temp thing. Could be temp enrichment, coils (for me) or plug wires, but it runs out great after I get past the stumble. I don't know how old my plug wires are, which should be it's own reason to replace them. My coil is 3 years/20K miles old.

Hmmm... interesting. That still sounds similar to what I am seeing. So you tend to see the problem also when the car is warmer?

Definitely yes. The dieing is only when hot. Stumble off idle is less before temps hit 100 the first time. Never put it together until today.

Travis have you heard anything back from Adaptronic?

Stein which firmware are you running?

Dunno, but should be whatever the latest and greatest is when I last reflashed in March to fix the first MAP issue. I can check tomorrow.

I had a similar intermittent problem which I thought was PC pro related at the time. Turned out to be the egr valve acting up. I blocked it off and no problems since. If you still have yours unplug it next time it acts up.


http://i16.photobucket.com/albums/b3...h_MOV06125.jpg

My EGR is blocked off

Mine is also.

OK I just installed the new Walbro 255HP fuel pump. It was a pretty easy installation.

At any rate I idled the car in the driveway, running the a/c, top down, revving the engine, to get the coolant temperature up. It got up to about 210-215 depending on which gauge I believe.

I did get it to do the rough running thing again, though it was at a higher temperature than before, maybe 5-10F higher. But, it cleared up faster, and with some tweaking of the 2500-3000 rpm fuel table cells I got it to go away completely. However I am far from certain that the problem is fixed. I need to take it for a drive on a hot afternoon for more testing.

While I had the fuel pump assembly out of the tank I tested the fuel pressure regulator using a compressed air source. It was completely closed up to about 60-65 psi, then opened up, just like it should. So that is not the problem.

I think it will be Monday before I hear back from them, I'll post up as soon as I do.

I thought I had mine licked today. Drove for about 30 minuts tuning cruise cells. All of a sudden it died and from then on it was a bitch to keep running. I know it is not strictly temp, as I was watching my very first warmup of the day. It went to 110C before the thermostat opened, then dropped to the normal 100C. It convinced me that I need to change the thermostat as it must be getting sticky. Not tied to the problem, but it shows that it isn't strictly temp related as I drove die free for 30 minutes after the temp spike.

I also did some large scale tuning and hit the 2500-3000 ranges hard as well. I did get it to stop dieing with a lot of effort on tuning idle cells but I'm sure it is just masking the problem.

I didn't get to do the full tuning cycle that I linked earlier as I didn't know where the idle air tuning screw was.:noob: I have an autocross tomorrow so hopefully can get back on it tomorrow afternoon/evening.

Is your problem only around idle? Have you tried cleaning the IAC valve and the air screw port?

Are you guys using the correct wari software version with the correct flash? They come in pairs, and if you don't use same with same it can cause instability. To my knowledge you guys are the only 2 with the problem thats this severe. It has to be something that has just gone overlooked, but I swept the software settings looking for anything and couldn't find a single mis-step.

OK I did not know about the pairing. So based on this Adaptronic page we should be using V1.0N firmware, Flashit V1.0, and WARI V1.11?

The thing I do not get though is if they are paired, how do we try out the beta releases like this one of WARI if there is not a corresponding beta of the firmware?

Right now I am using the beta version of WARI linked to above, and using the firmware supplied with the base maps Travis sent out about a month ago.

If you goto Help on the wari panel and then goto about. That window that pops up will tell you what software you have and what flash to use with it. The newest version on that page you referenced uses V1.12 wari and V1.0O flash.

I like the new Wari by the way. Pretty nice setup...

OK then I'll try checking WARI and flashing the corresponding firmware. I'll do that this morning and report back if I have problems. It may take a day or two to be sure I do not have the same problem again since the weather is going to be cooler here for the next couple of days.

Yes the new WARI is better. The active cell locator feature on the tuning table is a big help.

UPDATE: OK I loaded firmware V1.0O (using Flashit V1.0) and am running WARI V1.12. I verified the firmware version required under 'about' just like you stated above.

That new Walbro is loud during the initial key on 3 second prime. When I turned the key on this morning I thought "WTF is that?!" :giggle:

those guys need to start naming their releases with version numbers.

i've got like 3 wari.exe files and fuck me if i know what version they are.

Agreed, lol. I think I named the one I sent out in the packet. Drives me nuts.

I can't wait to get mine in... those big walbro's remind me of the space shuttle hydrogen pumps. 3...2...1... ignition :giggle:

OK problem still not fixed. When the temperatures climbed to about 205-210 on the way home (a/c running, cruise speed, light to moderate throttle) it happened again. So matching the firmware to WARI and the new pump did not cure it. This time though I noticed that the AFRs did not appear to be going lean like they did before, which is interesting.

On a positive note I have the EBC working now.

I did notice that VVT was enabled on trigger 2. That should not be checked. Rob is on the case now, so we'll see what he comes up with.

In for EBC install thread with pics!

Rob needs logs of this happening. Anyone by chance logging when any of this occured?

I'm sure I do but all logs were when I was tuning and changing fuel cells so might be bad data.