There definitely was no oil on the cyl #1 wire itself. It was straight clean rubber. I did check that pretty well last time, but I only checked the one cyl. and didn't really get all intimate with the tube and my finger, lol. I will check again, though, and all 4 cyl's just to be sure.
Finally getting some logging answers: Searching the interwebz is hard work. If anyone wants to suggest some important parameters to log I'm all ears. I'm guessing there will be several that are obviously important and some that maybe aren't so obvious to me. Preliminary reading says I may have trouble logging IAM and engine load, since I'd need the memory addresses for these parameters, and AFAIK the 00-01 USDM ECU's have not been decoded at all to determine this. Hmmm .....
Got some "normal" operating logs in both EcuExplorer and RomRaider. So far I like the way the parameters are listed in RomRaider better. Now I have lots to learn to have any idea what any of this means. I'll probably continue to log every to and from work trip and any especially long trips until I log through some bogs.
Yeah, RR logger is much better. Do some looking at logs on romraider or the openecu websites, and you'll quickly learn what parameters to log.
So far I'm logging: A/F Correction #1 A/F Learning #1 (always 0?) A/F Sensor #1 Atmospheric Pressure Engine Speed Ignition Total Timing Injector Duty Cycle Intake Air Temperature Knock Correction Advance Learned Ignition Timing Manifold Relative Pressure (Corrected) Throttle Opening Angle Engine Load (Relative) A/F Correction #3 (always 0?) Idle Speed Control Valve Step (pesky P1507 - figured worth watching this) I think I'm supposed to be logging Ignition Advance Multiplier and maybe the directly read Engine Load - but I haven't had time yet to try all the different addresses looking for mine (suppose I could google that). I'm seeing Learned Ignition Timing values up to +12.00 - there's some positive value in there almost any time I'm accelerating. Also, Knock Correction Advance down to about -4.50, negative values show usually when I'm on it more (>75% throttle), otherwise its 0. A/F is usually around 14.7, but goes as low as 10 for bits and actually goes as high as like 21.94 quite a bit (thinking that's leaner - if it really is air:fuel ratio - air being the larger number). Of course, I haven't had time to research what any of that means or if its normal or not. And I haven't logged a bogging event yet. I'm getting tired of reading the internet, lol. I'm also a bit confused about some of the other tools - like Learning View. They read like they only support ECU's with ROM definitions (i.e., not my car) - because they're reading values out of RAM or something (fuel trims, for example, which I was confused not to see in the parameter list, but then again I really know nothing about all this still). I just need to plug the cable in and try the app, though. Haven't had time yet. Did give me an excuse to pull a couple WOT runs to redline in 4th. Hehe. Not that I need an excuse, though.
Ok, total timing and AFR will be the main 2. Watch those while you drive. You will get to learn what the car like running for timing as you cruise. You are best to NOT have the bog for a bit so you can get a good read on the timing. When the bog happens, you want to be able to compare what is "normal" and what is currently happening. It could just be going real lean, you might be a timing thing, but you want to make sure you know what the car "normally" does. When I was tuning my car, I could drive around and tell you ballpark numbers (+-2) of my timing while driving. But when something wasn't right, I knew right away because I was so used to seeing the same numbers all the time. Russ
Gotchya. Was actually thinking the same thing - get some baseline data and learn what "normal" is. I'll need to find something more convenient to monitor while driving, though. There must be some giant gauge type app I can run. More googling.
I see now. Had noticed the MAF & Injector duty graphs, but missed the Dashboard view entirely. lol Thanks.
Got logs Caught a log this morning. Driving to work. Bogging pretty good at WOT in 3rd, 4th, & 5th. Here's a scap of the log. I know its out of context, so if anyone wants to see baseline data I can post some. However, it seems glaringly odd that the injector duty cycle drops to basically nothing while I'm WOT. Also, there's a some amount of learned timing there, but I'd have to dig back through baseline's to see if it was also there in similar conditions when it wasn't bogging. I also have a longer log where I saw this same bogging this morning - like one where I did it consecutively in 3rd, 4th, and 5th gears. It does the same thing, though. As I approach WOT, Injector duty cycle drops to literally 0, if I hair off the throttle the car surges forward and the duty cycle jumps back up to 30%-40%-50% area. Notice in the second short log how I go WOT quickly and at like 35% throttle opening the duty cycle's like starts climbing - hits 9%, then bam it hits WOT and duty cycle drops back down to nothing.
Hrm, yeah that certainly looks wrong. Even your timing curve is messed up. The timing curve should slowly ramp up as RPM increases, too. At those points were IDC jumps to zero or about zero, the timing also jumps down a few degrees. It's like the ECU thinks you clutched in or let off the throttle in those areas.
The logs seem to show appropriate values for the Throttle Opening % - isn't that what the ECU is actually reading form the TPS? If it was faulty I'd expect that % number to be dancing around. But that's just a guess. Also, I'm wondering if the duty cycle is a measured value or a determined value - that is, is the ECU driving the injectors to < 1% or is something else causing that and the ECU is measuring that the duty cycle is < 1%? I'm pretty sure this is showing that the ECU is telling the injectors to operate at < 1%, so then my question is why? But I'm not sure how to answer that.
I'm not sure exactly where the ECU measures throttle (ie, if the number comes from the TPS or if there is some other sensor), but your hypothesis about the throttle numbers jumping around as well certainly sounds plausible to me. IDC is a calculated value calculated from engine load, injector latency, and RPM.
step 1. Put a wbo2 in the tail pipe. See what is actually coming out of the exhaust. It will also verify your stock o2 to see if it is right. Step 2. Run a 60-70% throttle run. See if it will pull clean to red line. trying to verify if it is a load/rpm problem, or a throttle position problem. Then maybe try a 80-90% run is the lower one pulls clean. When your injectors turn off, your afr goes to 15:1 (super lean). It is like you are losing the load on the motor (which you are). So that is why your timing is spiking up. Are you still running the automatic ECU, or do you have a 5 speed ecu in?? It is acting like the ecu wants the tranny to change gears. It is a bit soon, but base off of rpm ramp speed and such, if might try to shift that soon. It adds gas back in, and the timing goes down and the afr goes rich again. I would also say that you are running a bit lean when it is pulling ok. What octaine gas are you running?? Russ
Great points by Russ. Makes sense that that the auto ECU may be expecting a shift and is freaking out.
I can answer #2 right now. 60-70% will pull all the way to redline. 80%-90% I'd have to test more carefully to be sure I'm at the right throttle position, but my point is that it definitely seems related on throttle position and not load/rpm. I haven't been able to notice any difference between gears (different load, same throttle position) as to what point in the throttle the bog occurs, and I can always back the pedal off to a certain point and it instantly goes away and pulls again. The ECU is a 5MT ECU. But its still on the auto harness and the TCM is still hooked up, so to the best of my knowledge, for all intents and purposes it is behaving like a 4EAT ECU. I know that's a big variable in this whole situation, but here's why I continue to lean towards it not being the problem: If it was, it would be more consistent. This occurs some days and not others seemingly at random. If it was ECU behavior, it should do it under the same circumstances all the time. Also, there's not a lot of data going back and forth between the TCU and the ECU. The TCM gets its own signals for rpm, tps, etc. and decides for itself when to shift. The ECU is minimally involved in such an operation. Here's all I can find in the FSM for what signals are passed between the TCM & ECU (only signals I didn't list are those going to both TCM & ECU - like TPS, and the AT Diagnosis Line [for throwing AT codes]): View attachment 15273 View attachment 15272 (b20 on the TCM is the AT load signal, but c11 on the ECU is labeled as Mass air flow signal for AT) I certainly don't know everything (or anywhere close to) Subaru ECU's, but I just have a hard time fathoming how a couple "torque control" signals could cause this - unless it was very, very predictable, which its not (except for the bog after 40 min at 70+mph, which is similar but different arg. I hate that I even have to qualify it like that, sheesh). Also, I've read of more than a couple people still running the 4EAT ECU's who never mention any issues like this (except the high idle that learns away) - including Dynapar on here. The slightly lean could be my I-Speed SRS-10 reflash. I'm also now catless, though. And I run 89 octane - usually Marathon or BP.
I don't know that the ECU can expect a shift. The only way would be if those torque control signals indicate it. Cause, as far as I know, the shift points learn a bit - making prediction of a shift impossible. I forget the factors it looks at - I believe rpm & tps, not sure about load - and there's a graph out there somewhere that kind of shows the base intersecting lines (might have seen that in the FSM, even), but I think they adjust slightly over time - could be wrong though.
Well, let me qualify that statement, it is possible the auto ECU would have different timing and fuel values in the cels you were hitting while WOT, so the actual mapped values were causing your bog, where with an auto the car would've shifted and hit those cels in different driving conditions. I don't know, honestly just guessing here...:dunno:
Also, threads of note on TCM signals (will continue to add as I find): http://forums.nasioc.com/forums/showthread.php?p=12463689 - guy asking about the torque cut signals, appears they maybe drop timing during a shift http://forums.nasioc.com/forums/showthread.php?p=11628665#post11628665 - legacy77 (has a big AT/MT swap write-up on his own site) on the AT/MT identification pin and the same ECU's being able to operate in either mode. http://www.extreme-machines.nl/down...SSION/12. Electrohydraulic Control System.pdf - some pdf explaining the electrohydraulic control system in a Subaru AT. Talks about the torque signals specifically. http://www.endwrench.com/pdf/feb2004pdf/4EAT.pdf - A Phase 1 4EAT diagnostics and service guide - different from FSM's
Yeah, I know the TCM and ECU talk, not really sure how much (haven't had a chance to read those threads). I wouldn't be surprised if the TCM is sending signals that tell the ECU to switch modes/maps during a shift. Is there a gear pattern? IE: does it bog every time going WOT in a certain gear? Or is it completely random?
Also, I can verify this is true. All of the WRX ECU ROM's that are older then ~2003 have one single ROM for both AT and MT (with different timing, boost, and fueling tables for AT and MT).
Best info I've got on those signals is this: Torque control signals 1 & 2: Reduces engine torque during range selection and gear change. (but how, exactly - reduce timing?) Torque control cut signal: Sent from engine control module (ECM) to TCM to temporarily inhibit the torque control when starting off with low coolant temperature. (probably the culprit of that high-idle that learns away) Not really. It never seems to bog in 1st or 2nd. And it didn't used to bog in 3rd. But, again, it only does it some days. It could happen in the morning, but not in the afternoon. That's one big thing that says its not ECU/TCM behavior - too inconsistent. Its got me thinking about snipping those wires, though, just to see......
Yeah, but you may (well, are most certainly) hitting different load cells in higher gear (or also, when it is colder out), and maybe the two (ECU and TCM) are talking together at just that certain point and the result is causing the bog. I guess I wouldn't rule this out a the root cause.
But I think I hit the same load cells, revs, temps, etc. combination(s) far more often than when it bogs. I mean, I know that every single weekday day when I hit the new 35w bridge on my way to work I get two more lanes and I drop to 3rd and floor it and get around all the 40mph-ers that have been holding me back for the last couple miles. Also, the temps must be same or close to on several days - but yet I may only get the bog once every few weeks. Just playing devil's advocate a little here, though - I see what your saying and can't rule it out entirely yet.
Oh snap: -- http://www.endwrench.com/pdf/feb2004pdf/4EAT.pdf I guess I'm too damn stubborn, assuming that's accurate. And I didn't search well enough before today. My bad. That explanation fits the logged data/symptoms/conditions pretty well, I'd say.
Yup. Thats exactly what I was getting at. Fuel cut and timing decrease during the shift. It seems like the TCM is telling the ECU it is shifting when with your 5MT, you've still got rev's to go..and this may only happen at a certain point in a gear, where the ratios are different enough between the 4EAT and 5MT to trigger it.
That article did a very good job of explaining how the TCM works and what those signals are for (unlike the FSM's). Of course, I'm still not sure what the "normal" signal is and what the "fuel cut" signal is - so I can provide the normal one always, but I'm guessing simply disconnecting those wires will do the trick since that's how it would be if its a 5MT harness. Then I just have that 40 minutes at 70mph+ bog - but that's so consistent I swear its some timed thing the ECU or TCM's doing.
Yeah, I would say the ECU is at "normal" always, and then gets tripped when it receives the signal from the TCU. Remove the signal output from the TCU (or input in to the ECU), and you're good.
to many posts..... Anyways, this is what I learned from my 4eat (in a 02 wrx). So it may be a bit difference. The tcu as load maps in it. It uses these load maps to determine when and how hard to shift. The tcu uses the load values from the ecu (in a round about way), but it is using mostly tps, maf, and rpm to determine what you (the driver) is doing and how to best shift the transmission. When the tcu map tells the ecu that it needs to shift, it cuts fuel (so idc's go down to nothing), adjusts spark to the anticipated load sites, tells the tcu to performe the shift, then ramps fuel back into the mix. If you cut timing, you load the cylinders with fuel, so it is easier to just cut fuel, then you also do not have to worry about back fire. Your logs look pretty much like the logs I would get from my car. I guess I would keep logging it and save the logs where it goes lean and you don't make any power. Then see if there is something consistant between them (rpm point, air temp, load site, something). I know in my car, they talked to each other a lot, and that all had to do with the the load on the motor. Remember, it will also do the same thing for downshifts. So if the load get so high, the tcu will tell the ecu to do the same things, then the tcu will just downshift the tranny instead of upshift. So maybe load the motor (4th at 40) and floor it, then try 2nd at 40. Just load the motor differently. Seems like the 40-45 area is the sweet spot on the 4eat, where it wants to grab 1st when you floor is, and will drop to 2nd for sure for anything else it needs to do. Russ
Advertisement removed by Readymix, and replaced by a delightful and relevant song, by an artist who also is unable to see these forums.
