You guys with the 5spds.. especially running $58.

Can you idle across a parking lot in first or second gear without it bucking/jerking?

I cant..

Even light light throttle does it. I have to accel, or decel. No light cruise for me.


-- Joe

 

i think mine bucks a little when not under load in first or second. when in 5th at say 45 mph i would not say it bucks but surges a bit. if i go super slow in 5th it might buck but never have tested that. what i can tell you is before i tuned eprom it bucked so severe i would lay rubber in 1st(kinda cool now i think of it) on stock ecu and prom. tuning took care of that. most likely very lean with no AE. my spark tables are very conservative and will look at that when i tune my changes(7/04) and buy gtech.

 

Maybe you could try experimenting with turning certain features off, like DMAP accel, fixed spark advance, open loop, decel enlean, etc... And see if you can find any links.

 

Joe,

Are you seeing sudden IAC or spark timing jumps while the car is "bucking"? I know I've seen situations where the ECM seems to be fighting itself while coming out of RPM control and going into throttle follower. This will generally tend to happen as the TPS is hovering around the RPM control threshold. While TPS is opening, more air is getting to the engine causing the spark to retard. (Because the ECM is trying to maintain idle speed.) Then, as the TPS passes the RPM control threshold, the spark "jumps" to the non-idle value. At the same time, the IAC is potentially commanded to a different spot.

I can understand your frustration. To make things more complicated, sometimes something small sets off something else. Then, another interaction comes into play. So, as RednGold mentioned, it would be a good idea run a test with DE, AE and DFCO "disabled". During your "bucking" events, things like MAP might be changing fast and large enough to kick something like DE or AE into action causing even worse behavior!

If you have any datalogs of this happening, I'd be interested to see them. Hopefully, the answer is in there somewhere!

 

I think you're on to something there. Try disconnecting the IAC once warmed and see if it's related. You may end up having to decrease the attack rate of the IAC, throttle follower, and stall saver, and godknowswhatelse.

 

Some stuff to look at/consider.


ISMLTR1: FCB 190 ;128 ;65 ;IAC - Closed Loop Gainword, Small Error Retract, Drive 97

ISMLTR2: FCB 100 ;68 ;34 ;IAC - Closed Loop Gainword, Large Error Retract, Drive 18

ISMLTE1: FCB 150 ;100 ;49 ;IAC - Closed Loop Gainword, Small Error Extend, Drive 33

ISMLTE2: FCB 1 ;IAC - Closed Loop Gainword, Large Error Extend, Drive 1

ISMPNR1: FCB 130 ;65 ;IAC - Closed Loop Gainword, Small Error Retract, P/N 97

ISMPNR2: FCB 68 ;34 ;IAC - Closed Loop Gainword, Large Error Retract, P/N 34

ISMPNE1: FCB 66 ;33 ;IAC - Closed Loop Gainword, Small Error Extend, P/N 33

ISMPNE2: FCB 2 ;IAC - Closed Loop Gainword, Large Error Extend, P/N 2

ISTCBNL: FCB 9 ;3 ;IAC - TPS Delta For Entering Throttle Cracker 1.17%

ISTCBNH: FCB 10 ;5 ;IAC - TPS Delta For Entering Throttle Cracker 1.95%

ISALPC: FCB 8 ;IAC - Initial P.S. Stall IAC Pulses 25 Steps

ACDLD: FCB 20 ;IAC - Initial Delta Step For A/C Base Drive 33 Steps

IACACDL: FCB 0 ;15 ;IAC - Initial NV RAM Fail Delta For A/C On 40 Steps

ISPKDL: FCB 5 ;125 ;IAC - Park Delta From KISSWNA 125 Steps


SPTMP: FCB 50 ;255 ; Idle Stabilizer - Min Cooltemp For ISS Logic To Function 151 DegC

 

Hi,

Ok. WIth IAC disabled it is about 50% as severe. Infact, driveability is much better. Except.. Idle warm is normal. then. all of a sudden. it kicks up to like 12-1400 rpm for no reason.
(or is this telling me something?)

DFCO, DE, AE, all disabled makes no difference in any combination or all at once..

Kicking cruise timing down helped cruise a LOT. We're talking from a non-driveable state, to bareable.

could TB shaft bushings cause this type of problem?
Airfoil?

log of bucking: http://members.cisdi.com/~anesthes/bucking/

in .csv, and .uni (datamaster) format. Shows a idle crawl example.


I have a lot of theorys. one of them is driveline play. Loading and unloading the motor, causing MAP fuel corrections, resulting in more bucking.. Anyone buy that ?

-- Joe

 

btw.. Those logs seem to be 10d short of my tables in TC, so i'm asuming the aldl doesn't output the initial sa. (or i'm 10d short everywhere. haha)

-- Joe

 

All I can "see" is a little difference in what I would assume is closed throttle timing vs open throttle, and some work needed on the VE table, especially at idle. It may be oscillating about the block boundary, or TPS requirement for spark advance.

 

Joe,

A few observations from the file.

1) The Datamaster file appears to indicate the the EST is in "bypass" mode during this trace. Is there any reason to believe this? Perhaps, if the ECM really isn't controlling the timing, this could be a big problem? (Or, am I just not reading things correctly...?)

2) The initial IAC position when idling around 900 RPM appears to be zero when MPH = 0. It goes to ~6 steps when the vehicle is moving. Do you have the TB adjusted specifically to get the IAC counts to zero? Or, is this a sign of a potential vacuum leak? (Again, I could be reading this wrong in Datamaster...)

3) During the first tip-in starting at sample 30, the TPS value oscillates between "open" and "closed" throttle (i.e. non zero value (18%) then 0% or something very close then 23%....). You can see IAC, Spark and MAP moving with the TPS change. IAC jumps up 12 steps. Spark Timing increases by 10deg. RPM soon follows. Are you certain your TPS sensor is good and adjusted properly? This seems to indicate a noisy signal as we transition from 0% to something larger. Or, perhaps it's loosely mounted to the TB? You mentioned things got better with the IAC disconnected. This would make some sense based on the file. Removing IAC influence should help. But, we're still fighting spark and fuel.

4) There are 4 places in the file where the BPW shoots very high. The first occurance of this happens to be synched with the big TPS delta (18% in ~150ms) pointed out above. There's also a corresponding MAP increase at this point, too. Datamaster is reporting a BPW of 28.56ms or 48%DC at this point! Was AE turned off here? The three other times there's a big BPW I don't see any reason for.... So, I can't blame AE on those....

5) I don't understand why spark advance stays static when TPS equals zero. Do you have the spark stabilizer calibrated out? I would have expected this value to move around a little when TPS is equal to zero..... It just seems to stay at 13.4deg whenever TPS=0%.

6) I'd agree with RednGold that some work is required in the VE arena. But, I don't think it's causing the oscillation problem.

I think there's merit to your idea of driveline oscillations affecting the behavior of the ECM. But, I think this is being "kicked off" by the effects of the *HUGE* TPS changes when the vehicle first starts moving. Once we remove that first kick, we'll see how good/bad things really are. Maybe one of the first subroutines we'll add to Grumpy's $60 code will be a driveline damping algorithm for manuals!

The first thing I'd try/check is the TPS sensor. If I could reach into the datalog and change one thing to see how everything else reacted, I'd like to see a "smoother" TPS signal. You mentioned something about worn TB bushings. If these are worn bad enough, maybe they could be contributing to the TPS oscillations, too???

So, two more questions/favors to ask. First, if this data log didn't have everything "turned off", can you run another log with everything disabled? Second, it might be helpful to look at your calibration to see if there's something else going on. If you can make that available, we can investigate further. If you want to wait on all this until after looking at the TPS, that would be fine. Ideally, with a smoother operating TPS sensor, the problem will go away!!

Let me know what you think. Thanks!

 

I noticed that, I think datamaster is high or something. All my logs show that flag the second the motor starts running. However, we both know, my car wouldn't make it 10 feet w/the ecm in bypass mode. I don't understand that one ?

I adjusted the IAC using the method described here.
Jumper aldl, turn key on. Unplug IAC, turn key off. Start car, adjust tb idle set screw until desired idle, shut car off replug IAC.

I checked the mounting of it today. Its on there tight. I can't account for its acuracy. Theres a chance it could be giving varied resistance values at the same throttle position. It is from 1989.. Think I should swap it out ?

No. In this sample, nothing was turned off. I'm a little confused about that too. (the bpw)

Yep. Which isn't exactly what my timing table shows either.

rpm 30 40 50 60 (kpa)
600 25.0 25.0 25.0 23.9
800 26.0 26.0 26.0 26.0
1000 29.9 29.9 29.9 28.8

That was a theory I had. I had read that worn bushings can cause vac leaks. I checked today, and it appears fine. I sprayed the WHOLE intake down.. no vac leaks.

Sure I'll scan some things tonight. I have a spare (old) TPS sensor, but perhaps a new one is in order?

I switched to $60 this eve, to rule out any inconsistancies. I'm happy to report $60 behaves exactly like $58.

I'm gonna toss my current $58 bin in that folder for your review if you'd like.

Thanks !

 

Ok. Heres some feedback..

Tried with DE, DFCO, and AE disabled. Car bucked a little on 0 tps crawl in 1st gear. part throttle cruise was fine though.

Changed out TPS, and reset IAC.. This seemed to actually make the problem worse. (with the normal bin, not the removed bin)

If you look in that directory, there will be 2 new logs called lot- and then a desc.


I then went on the highway to give some boost a try, and the car almost died!

For fun, if you wanna look at the spreadsheet wtf.xls I reproduced this every time.. Basicly, 2nd gear acceleration. Normally do it to 6k then shift and its fine (including last week). This time, at around 7psi it like. died. tire chirped, then came back to life. almost like ignition lost.

I went home, threw a $58 bin in (with same constants tables). Same issue.. Pulled over, disconnected the fuse to the water injection setup.. Tried again.. Motor spun to 6k under boost no prob. 12ish psi. Weird.



My water/alky injection (blue windshield washer fluid for now) kicks on via a napa switch at 5psi.. Think the methyl alcohol in the washer fluid is richening the car too much and stalling it ? Think I should give up?

-- Joe

 

Joe,

Still looking at the data logs. But, I do want to ask some preliminary questions :

First, would the following descriptions of the files apply...

1) lot-normal =
2) lot-novede =
3) woohoo = wtf.csv starting at sample 7822?

The bin you've included would apply to (2) above?

I notice that MALF 35 was active in the first two runs. So, I just wanted to make sure the IAC was really connected. I'm just learning what the $58 code does when everything is "working". I'm not sure of what, if anything, happens when MALF 35 is active.


I'm still kind of concerned about the TPS. It just doesn't seem like it should be as jumpy as it's showing in the log files. You mentioned you swapped the TPS out without seeing much change. So, here's another favor to ask.... Can you log some sweeps of the TPS (key on, engine off)? Try a few slow tip-in's/tip-out's all the way to 100% and back. If it's not a big pain to swap out the TPS, it would be nice to see how the two sensors compare in this test. It also might be interesting to try the same thing without the accel pedal. Jump under the hood and turn the throttle bracket by hand a few times. Maybe the problem is due to slack in the throttle cable/holder/assembly???


I've taken a quick look at the wtf.csv file. From what I can see from a *very* quick glance, it appears the ECM was commanding appropriate values during the event you have highlighted. Can you adjust the flow of your WI system? Perhaps it just needs to back off a little (rather than being completely disabled) to avoid this situation? Something like a smaller nozzle...?

I wouldn't suggest giving up on this. In fact, I'd look at this as an opportunity to make some changes to the software that might help your car (and, maybe, someone else's). For example, assuming you're not using the ECM's wastegate functionality, maybe we could find a way to turn on your WI system with that..... Keep the faith. I'm sure we'll be able to "fix" the problems!

 

The "Normal" is the $60 equal of the $58 bin I put in that directory. Nothing disabled. Iac and TPS not tinkered with at that point.

Removed, DE, AE, and DFCO. (zero'd out all tables). Still no IAC or TPS tinkering. woohoo.uni, is after I reset iac and new tps.

#1 above actually.. Though my logging was with $60, they're as far as I'm convinced, identical. Grumpy did a good job.

I can post the actual $60 bins when I get home tonight, if you like. Just lemme know.

I adjusted the IAC the night before. I didn't clear the code. I'm not sure.. I guess I prolly should have cleared it eh?


For sure. I'll do that tonight and post results. You prefer the datamaster log or csv?

Btw. I set the tps to .53v at closed throttle, and it maxes around 4.37 I think at WOT. This sound ok?

See thats the weird thing. I'm running the same nozzle that a lot of guys on here are, and same pump. Only difference, my pump is at 60psi, everyone else seems to be running 100psi.. Yet.. it appears mine is too much flow? Or perhaps the blue windshield washer fluid isn't the right stuff to use?

I can change nozzles for sure. I'm using: http://home.att.net/~alkycontrol/ima...onents/noz.jpg

Thats something I've asked about a few times. I've tinkered with some of the WGDC stuff, but without an ecm bench its impossible to test.

I'm gonna get an ecmbench soon, its just been a tough month.

Thanks !!

-- Joe

 

The intermittent high PWs of 28msec is because of the way the code works. Just ignore them.

Make this bin change and see if the bucking goes away:

At $10BC from $4F to $01

That is for the stock $58 code.

RBob.

 

Rbob,

Excellent idea!! :hail: I now see why the spark advance was always going to a single value with TPS "closed". (Well, there's still the MAP dependency.) This will be a good thing to implement to see if things get better. Ultimately, we'd probably want to implement some sort of "blending" function to ramp spark when TPS crosses the "closed" threshold. Right? You mention the abnormal BPW values are not an issue. Have you seen similar with the jumpy TPS values, too? I would guess the problem with BPW is the fact that the information is sent in 2 bytes over ALDL..!! I'm putting a lot of faith in the fact the reported value is *really* what the ECM is seeing and using. If this isn't true, we can start looking elsewhere....

Joe, just to keep the tests the same, I think it would be a good idea to clear the MALFs and try this again. Sometimes the thing you would think is the most trivial and unrelated to a problem is really having a larger influence. (Kind of like TPS going to zero forces a change in the spark lookup routine! ) I don't expect everything to magically improve with the MALF reset. But, it should keep us from chasing around in circles if we keep things consistent.

I still would like to see the TPS tests, too. Even if RBob's modification helps, there's going to be potential trouble with other functions that use TPS or dTPS. Your max and min voltage values look good. We've just got to be able to get the sensor to smoothly transition through the range relative to your foot! I'm now getting used to Datamaster and can always export the values if needed. So, you can stick with the DM .uni file. I do appreciate the offer.

Do you know if I can use a $58 .ecu file with the $60 bin? I don't know if anything "new" calibration wise has been added... If I can, I'd say go ahead and put the bin up.

I don't know much about the chemical differences between "water" and "washer fluid" to be very helpful here. I'm just basing by suggestions on what you describe the engine is doing. (Guy goes to the doctor and says "it hurts when I do this". The doctor says "stop doing that". That's my approach to this one. ) Do you see the same type of behavior with a water-only mix?

To be honest, I am interested in possibly setting up WI on my turbo, too. And, given Grumpy now has source code available, we should be able to write some code that will "take over" the wastegate PWM rather than having to figure out a way to calibrate the existing algorithms to our will. I'm thinking we can leave the existing wastegate logic intact and add a bit to the options calibration that will use either that or our "new" water injection algorithm to operate the PWM. This type of thing is *exactly* where having an ECM bench is essential....

But, I digress.... Let's get this part throttle bucking car under control first. Then, we'll handle the WOT stuff. Here's a good procedure to follow :

1) Clear all the MALFs before testing....
2) Get the TPS logs as mentioned earlier
3) Use the $58 code with all the functions "disabled" *but* without RBob's change (yet). Log the results
4) Now, change the code as per RBob's instructions. Do the same maneuver in (3) with no changes to AE, DE, DFCO, IAC, etc. Log the results. And, let us know if it *feels* better.

For the time being (until this TPS stuff is worked out), I think we'll do well to keep the TPS-dependent fueling functions out of the way. After TPS is corrected, we'll bring the other functions in one at a time.

 

Hi!


Made the change. I think it made it worse infact. Switched bins around a few times, seems worse with that change.

Did you expect me to disable dfco, de, ae, and so on with this change as well or were you wanting me to try it with my normal bin?


I didnt get a chance to do the logs today with the TPS. I'll try to get that done tomorrow. Had a minor crossmember pulling threads out of frame problem today that I need to tap to a larger bolt size tomorrow.. heh. So that put a damper on my eve.

-- Joe

 

Interesting. That change was to eliminate the sudden change in SA as the throttle moved. In the bucking log file the SA was changing from 13 to 25/26 degrees and back (+ 10 initial).

Also, graph the TPS & MAP together and notice how the MAP follows the TPS. This tells me that the TPS is in fact moving with the throttle body blades. Else the MAP wouldn't move.

The first half of the buck log shows a self-feeding oscillation. Like either the driver is being lurched to and fro or a problem with the throttle cable.

The second half of the log shows the RPM rising and falling in a rather smooth manner. This is accompanied by a low BLM of 108 and the O2 reading showing drop off and then RPM surges.

That change on a stock/regular bin is fine.

RBob.

 

Joe, you'll probably want to try it both ways to see if there's a difference. With as many things moving around as there are, it would be hard to tell what's causing what..... Ideally, we'll see an improvement with the functions disabled and otherwise out of the way. At a minimum, I'd expect no difference in driveability with RBob's change in either case....?


RBob, I noticed that, too. The question (to me) is whether or not the *magnitude* of the TPS values is correct. If it is correct, we're going to have to spend some time making sure the functions we're turning off are calibrated so they don't go active under these conditions....?


Too bad we don't have a pedal position sensor available to us! Hopefully, after Joe does the TPS logs, we'll have a better idea of which is causing the problem.

 

I'm going to try and log that stuff for you guys tonight. I had an ignition problem that got soo bad so quick it made the car almost not driveable.

I yanked the dist out, and it seems the pickup is faulty. I'm gonna bring it to autozone and get a complete new dist (lifetime warranty). I also bought a new coil, since the one I have is from 1989. I'll check the plug gap on all the plugs, reset it to around .025-.030, check all wires, then drive around for a bit. Once I know the ignition is cured (assuming bucking remains) i'll show you guys the TPS logs you've been patiently waiting for.

BTW .. my $8D calibration has the same exact bucking issue.
I prolly should have mentioned that before. I can make it much worse by adding a lot of timing, or leaning out the ve. But I can never seem to make it go away.


-- Joe

 

I can't seem to shake this skip. I wonder if maybe the plugs are too cold.

AC-delco R43Ts. 9:1 c/r..

Problem is, with a hotter plug I think cyl temps will get too high under boost (12psi)

-- Joe

 

Do you think it could be slop in the driveline. I think even a stock (manual) car will do the same thing....

 

I most definately have too much play in the driveline. Could be tranny, spring hub, etc. But thats not causing the missing. Though the missing might be causing the bucking which is worse because of the driveline play.

-- Joe

 

If you look in that directory, a file v23.uni exists.

I logged some not running TPS time at the beginning of that log. Thats with the new TPS sensor. I dunno where the heck I put hte old one. I'll find it.

It appears my plugs are fouling. I'm not sure why. I drove to autozone and picked up a few box's of plugs (1 set of R44s, and a set of R45s). I'm gonna try a few different heat ranges to see if they stop fouling at part throttle.

The drive time you see is with a reduction in VE accross the board(and a little timing too). I bought what used to be mostly 114blm to high 120s, low 130s. I was hoping to get rid of the misfiring. I think the plugs are just too fouled. I'll try a hotter plug, and drive around and see if it fixes it.

This whole problem may be related to fouling plugs, eh?

-- Joe

 

Remember that the BLMs are low.That is pulling fuel. Suggest a VE table rework.

RBob.

 

Yep. Thats why I reduced ve accross the board.

I pulled the plugs tonight. (2 week old plugs). They all had.. I gues its carbon. On them. like big chunks on the top. At first I thought it was detontion, but it flakes off too easy.

My gues is, my cyls have tons of carbon, and when I hit the water injection it steamed it all off on the plugs.

ANyway. I went 1 heat range warmer (R44ts) and gapped it to .030. Seems like the missing is all gone. Now the bucking is still there, but it was GREATLY reduced.

I wonder if the problem has been plug fouling all along.

Looking at that log, I still see some low blm points, and a lot of high blm points. Perhaps some VE work is in order for sure.

Of course, this bin worked great in the spring when it was 60f out every day..

-- Joe

 

If you look at that log, there seems to be a radom 28.56 bpw here and there at idle.. Weird.

-- Joe

 

Joe,

I haven't looked at all 11,000+ points. But, a quick glance at the engine off TPS data looks pretty good. Also, the first drive cycle at sample 1021 shows a much prettier TPS transition. That is, until sample 1091. At that point, it appears the throttle goes closed again. Soon after, it goes up again.... I can't tell from the data. But, is the bucking causing your foot to move the pedal back-and-forth? (I.e. what RBob mentioned earlier??) I would have expected this particular maneuver to be pretty smooth compared to the earlier logs.

The problem with the fouling plugs/missing cylinders certainly isn't helping matters. Again, there are probably various causes all conspiring to make you part throttle operation less than optimal.

Maybe it's time to start thinking about how we can use IAC to help us out of this situation (after any mechanical problems are addressed). I'm not sure if the existing IAC algorithms could do this. But, maybe we could open the IAC whenever the throttle "suddenly" closes. (Kind of like what appears to have happened around sample 1135.) If the behavior at that spot was really due to TPS, maybe we can find a way to increase the magnitude of what it was trying to do.....? Ideally, this would finish off the rest of the bucking problem.

The base pulse width anomalies are due (I think) to the way the information is sent over the ALDL. Because BPW is sent in two bytes, there's a problem whenever the two bytes haven't been updated in synch. For example, assume the current BPW is $0100 (first byte $01, second byte $00). The BPW is reduced one count to $00FF. But, the LSB is updated before the MSB is cleared. We'd get $01FF just before being set to $00FF. If the ALDL routine grabs the value at this time, we'll get an abnormally high BPW. I haven't confirmed this is the problem. But, I wouldn't be surprised if it was this or something similar.....

On a side note, I've been thinking more about a software change to use the Wastegate PWM signal for your water injection setup. At a minimum, I'd assume we'd want a MAP dependency when MAP > 100kPa. Would there be any benefit to TPS or RPM dependencies, too? If the table is small enough, I think we can find enough room for a 100-200kPa in 20kPa increment table. If we need a 3D table, some shuffling is going to be in order.

Let me know what you think. Thanks!!

 

Heh, Look at the rpm vs mph. Thats cuz I shifted.

In the situations of serious bucking, Its usually 0% tps, or around 5% tops. When it happens, I throttle off immediately and throw in the clutch to stop the bucking. Giving it more fuel sometimes helps, sometimes breaks the tires loose.

It didn't really buck much during that log cuz I didnt make it buck. If you look at the log, there is a shutdown and restart half way through. After the restart, I left the lot and drove down the road. I put it in 4th, at 0 tps. Normally after a second it would start bucking.. I really didnt notice anything until the rpms dropped to about 800rpm at around 30mph. Before, it would start bucking like crazy around 2k.

Unfortunately, its pouring out. Tonight, when it stops raining, I'll provide you with a 0tps log of bucking in 1st gear accross a lot.



Interesting. Ok cuz I looked through after my post, and I didn't see any obvious high 02 after those entries. So I was kinda confused.. At first I thought it was maybe battery pw correction when the fans kicked on, but I verrified no voltage drop at those points either. (and no spikes)


I think a few constants might pull it off. There should be a min RPM for water injection, so if you get xx psi boost at like 1500 you don't stall the moter. Then there should be xx psi to start injecting. Most people recommend 3-5psi to start.

I'm not sure about the TPS. On a turbo car for example, its possible to make plenty of boost at even part throttle.

Back to the bucking issue. Is what I'm trying to "correct" simply uncorrectable? I mean. Is it simply not feasable to expect my car to idle down the street by itself in any gear with 0 tps for any amount of distance without any type of bucking/hesitation?

I know a carb will do it, but thats apples and oranges. By by Vortec truck will do it too. I can't ever recall being able to do it on my firebird - not when I was running maf, $8d, or $58 (or now $60).


Thanks for all the help and support.


-- Joe

 

That is just the way the code works. Not really getting those PW's at the injector, ignore them.

RBob.

 

Can you include the IAC with the log?

The bucking should be fixable, there is something going on if it is that bad.

RBob.

 

Rbob,

What log do you mean? the v23.uni file has the IAC data in it. Or would you prefer I create a new log?

Thanks!

-- Joe

 

Doh!! That's what I get when I put the blinders on.... Feeling pretty dumb now....

We're probably going to want to break this into two parts. Getting the bucking eliminated @ 0% TPS. Then, eliminating the "low throttle" causes. You mention you didn't "make it buck" in the v23 log. Did you avoid this by selecting a higher gear? Or, by keeping TPS high. (Admittedly, I haven't looked at that data, yet.) Do things change based on how quickly you apply/release the pedal at non-zero TPS?

It would probably be good to go part way to the beginning in that you can try :
1) Use the software with RBob's spark patch enabled so we don't get the timing jump at 0% TPS like we'd seen.
2) "Disable" the IAC
3) I'm ASSuming the spark plug/missing cylinder issue is fixed....

It would be interesting to see if you get better results this time. (Try both 0% TPS and ~5% TPS.) If things are better, we should probably start looking at the IAC logic. If the bucking continues, we'll need to broaden our scope beyond just the IAC.

We can definitely "fix" this problem. We still have a few ***** to turn in the 749 before we give up. I realize it's hard to keep an optimistic attitude when you're being violently shaken (not stirred) by your car!!

 

I was looking at the buck.csv log. It's just that I prefer the CSV format for various reasons. Is the v23.uni log available in CSV format? I'll take a look at it.

RBob.

 

Heh.

If I drive downa hill, up a hill, at any TPS its fine.

If i'm on flat level surface, and the car is not decel or accel, and I'm either at 1800-2000 rpm (light tps) or 0 tps, it will buck. Not hard, enough to make the drivetrain jerk back and forth.

If i'm on the highway, going 70mph, in 5th gear, 0 tps, on flat surface, same deal. If there is load on the motor either way, be it accel or decel, its fine. But its when theres. no load it just.. bucks.

If i'm in say a parking lot, and I'm driving, and I just 0 tps and try to idle crawl accross the lot (remember parking lots are speed limit 5mph) its just impossible. It bucks like CRAZY!. At a shopping center, I have to throttle, clutch in, throttle, clutch in, across the lot. No crawling for me. This makes no sense cuz there should be a little load on the motor, and IAC should make it crawl smoothly accross the lot/shopping center.


As long as the 44s don't detonate. I'm gonna check the plugs today, then get a boost run in, and check 'em. If they look ok i'll continue.

Problem is, I can't disable the iac for longer than tinkering in the parking lot. I get unstable and floating idle.. Maybe that was the plugs. I'll try it again today..


Heh. true. Its not my daily driver, but I like to cruise once and a while. So I'd like it fixed. I'd like to be able to goto the mall for groceries without everyone thinking I'm driving like a ahole.

-- Joe

 

Now that works over time for play. Going out with teh laptop, will be back later with results.

The 44s don't seem to be fouling, or detonating so I think they might be the right plug.


-- Joe

 

Ok!

Tried Rob's hack with and without IAC.. Same result. Bucking at 0 tps across the lot.

if you goto: http://members.cisdi.com/~anesthes/bucking/rbob/

I put some files.

I'm going to create a .csv for each file for those who prefer that format.

Anyway. Interesting thing happened (after all my testing).

I noticed my header bolts were all pretty loose around the flange. especially the #7 primary. So I tightened them all up (like 5 turns each).

I drove around and found my blms were reporting 10 or so richer (lower) than before.. Look at v24, and v24-2. The -2 is after the header tightening. My guess is that air was entering the headers and making the o2 think the mixture was much leaner, which could have attributed to my plug fouling!

Anyway. The bucking didnt quit after tightening the headers..

-- Joe

 

in my 86 Trans Am i can let the clutch out softly with no gas and cruise. The regular idle is 700 while cruising its 550 or 600 no gas at all and its smooth. Stock right at this moment but going to let it breath with intake manifold and the works on exhaust. 3 inches baby

 

Beginning to sound like a basic ignition system problem. ie running rich to drop the firing voltage requirement down. Then it winds up too rich with any load. Flaky carbon deposits on an EFI engine is a serious sign of something being really off kilter. Not to mention 2 different codes have the same serious problems.

 

Joe, I looked at the rbob-iac.csv file. Before I comment I'd like to be sure of what I am looking at. Please correct as required.

Engine is started and allowed to idle down (rec 1 - 475)

Car put into gear, little TPS to get it moving (475 - 558)

Car allowed to idle across lot (558 - 745)

Surge gets so bad, throttle applied (745 - 764)

Continue to idle across lot? or de-clutched and rolled to stop (764+)


Also, can you re-create the rbob-iac.csv file and include the INTegrator term?

RBob.

 

I pulled the clutch in and rolled to a stop. It was soo bad I thought I was gonna break a tooth off the ring gear.

Yep I'll re-create that now.

Thanks!

-- Joe

 

Thats what I was thinking for a few days. I spent the weekend replacing: distributor, plugs, wires, coil, rotor, cap. Theres really no more to replace. I removed the MSD box months ago, thinking that was it.

I checked the plugs, they don't appear to be fouling now that the headers arn't leaking. (stupid problem I know, shoulda thought of that).

I'm gonna prolly pull all plugs again today and check for deposits. I know #1 and #2 are fine. Last time, #6 and #7 (when I changed 'em all) had a mound of crap on 'em.


-- Joe

 

I have been looking at the rbob-iac.csv log. After the inital startup the engine speed oscillates. This occurs shortly after the ECM goes into closed loop. The pic shows from record 280 through 480 of the original log file.

As the O2 sensor value drops so does the engine RPM. This is turn causes both the SA to fall and the MAP to increase. As the O2 sensor value increases so does the engine RPM. Now the opposite occurs: SA increase while the MAP drops.

There may or may not be idle correction SA involved with this.

I have a couple of recommendations: reduce the proportional gain term(s), make the 600 RPM row of the SA table the same as the 800 RPM row, reduce the idle correction SA table values.

From the log it looks like most of the SA change is from the main table. Here is the bottom couple of rows from the my-58.bin:
With the engine speed and MAP moving from 700 to 1000 rpm and 50 - 65 KPa, that would explain the SA moving from 21 to 27 deg.

RBob.

 

Hi!

Thanks for the recommendations. I made the 600rpm row the same as the 800Rpm row in main spark advance. Where would I find "idle correction sa table" and "proportional gain term(s)" ?

Either not in TC, or I have no idea what your talking about. I can load up programmer, if you know the labels I should look for.

Thanks a bunch!

-- Joe

 

In looking for the proportional gain terms I ran across the mean O2 for closed throttle. The lower limit should be increased. I have listed the terms below:

Once these changes have been made work needs to be done on the VE table. The logs are showing low and mixed BLM values.

RBob.

 

Thanks! looking at these now in programmer.

I've never chartered into this area. What would you recommend I set them to? This is a bit more advanced tuning than I'm used to.


Theres a few other things i'm curious about:

$383 Vac. load above which blk learn disabled 10kpa
$384 Map. load below which blk learn disabled 50kpa

This gives me the impression block learn is disabled at idle.

Also:

$67f Desired idle vs battery voltage (table)

This doesn't match up to my idle table. Its much lower. Could this
be causing part of the problem?


-- Joe

 

I would set the prop gains low, such as:

$35D 1 Proportional Gain, Low Vacuum

$385 2 Proportional Gain, Open throttle
$386 1 Proportional Gain, Closed throttle

As for these two:

$383 Vac. load above which blk learn disabled 10kpa
$384 Map. load below which blk learn disabled 50kpa

The one is VAC the other is MAP, so basically the BLM learn is only enabled between 50 and 90 KPA MAP. The $58 mask only has tywo BLM cells: idle and cruise.

As for the desired idle RPM stuff, I don't know. I would be tempted to set all three tables the same. Then see if it idles where it is supposed to. I haven't looked at the code enough to see when the idle vs. Battery is used.

With a stick car the P/N wire should be open with the scan tool always reporting 'in drive'.

RBob.

 

I'll set those tables to be my idle rpm then.

THe P/N wire on my car is shorted together. No clutch switch. Car used to be auto. So your saying I should put them on a clutch switch, so when the clutch is out - they are open?

Thanks!

-- Joe

 

As far as I know the P/N line on a stick car should always be open. In this manner the ECM is setup to know that the vehicle is always in drive.

Just checked both the Sunbird and Sy/Ty wiring, the P/N switch is D16, Org/Blk, and is to be open when in drive. Stick cars are always in drive.

This will enable a better idle control.

RBob.

 

These limits are only applied to the "open throttle" (no depenency on vehicle speed) BLM cell. When the throttle is closed, there is only an RPM check for that BLM enable. You should be able to see quite a few areas in the log file where the MAP is out of the above range (at closed throttle) and the BLM is still updating.

The logic will use the higher of the coolant/battery idle speed tables.


I can see the influences RBob talks about in his posts. But, I don't understand why there's such a severe reaction in some cases. From the log, it doesn't look like there's a drastic change in average BPW from the non-zero O2 readings to the "0" O2 readings back to the non-zero O2 readings.

That being said, I think it's a good idea to try backing off on the gains to see what happens. I wonder if adjusting the transport delay tables might help, too? I didn't think to ask this before. But, is the O2 sensor located very far downstream from the turbo?

Also, I can see some IAC behavior that needs to be addressed. There's a 3 step "shuffle" that pops in about every second in the area before the wild oscillations. There are smaller oscillations of RPM that occur much faster than this. So, it's not all of the problem. However, it is probably the cause of the difference you felt when you disabled the IAC in previous tests. Still tracking that one down..... It's too bad the idle spark stabilizer gets turned off when the vehicle is moving. I wonder if it couldn't help some of the higher frequency oscillations we're seeing.....

Very interesting idea. Seems almost too simple to work.... From an idle standpoint, it would make sense. You'd be able to take advantage of the different idle gains between the "loaded" and "unloaded" engine. After we get your car/engine to behave with it in one mode, it would be interesting to see if this change would help.

I would suggest spending some time tuning the VE tables in this area. If you find the values move around too much, the delay calibrations might need to be changed. Let us know how the lower gain tests work out.