91L98Z28

background to this code change:
A while back I posted about my being annoyed with how the AUJP prom unlocks the TCC during decel. For example, going 80mph, TCC locked. Let off the gas, TCC stays locked but then unlocks at 75mph because of the 75mph threshold for "always lock". I could lower this threshold, but in actually i want to disable it altogether (set it to 255), because i don't want the TCC locking at WOT going down the dragstrip. So it's very annoying to me that I can use compression braking above 75mph but not below (and that if I am compression breaking through 75mph, it suddenly unlocks!). Because commute speeds around here average 70-80mph, this is a constant annoyance.

Anyways, I found a way to hack/patch the code to solve this. here's what I did for anyone who is curious (disclaimer below!).

(1) I set the TCC lock threshold to 255mph. This way, the TCC will not lock at 75mph at WOT on the drag strip, tearing up my single-disc converter. I don't really want it staying locked if I punch it at 80mph going down the highway anyways....

(2) I made a three byte code change described below.

Looking at the ANHT hac around LE103, you can trace the code and see that preventing the JMP LE19A command will prevent decel TCC unlock while retaining all other TCC functionality (took me an hour of staring at this one page of code to come to this conclusion, ugh!).

Problem: the AUJP code is not in the same identical places as the ANHT code. As luck would have it though, it's not that far off. The same JMP command is at 0xE10F in AUJP (0x610F if you look at the raw bin in a hex editor). The address that is JMP'd to is a little different than ANHT, again because the code is not in the same identical place as in ANHT. They are offset by 8 bytes.

The bytes at 0xE10F (0x610F if you look at it raw in a hexeditor) in AUJUP are as follows:

7E E1 A2

7E is the opcode for JMP, and E1A2 is the location to jump to in the AUJP. Compare this to the JMP E19A in the ANHT hac, and you will see what it's doing.

Anyhow, I set all three bytes to 01 (NOOP), which disables the actual call to unlock the TCC due to decel conditions.

Burnt the chip, drove it. PERFECT!!!! The TCC now stays LOCKED during decel conditions, but still unlocks normally according to the TCC tables which control the unlock speed in MPH. No more annoying engaugements/disengaugements going down the highway just because I let off the gas or cross the 75mph boundary. (yes, it will still unlock which I touch the brakes - this is NOT controlled in the computer). But I rarely need to use the brakes on the highway so it solves my problem.

DISCLAIMER. I am in no way responsible for any damage you do to your torque converter or transmission. this is a controversial change, which some people feel will damage my car. I feel that for how I drive my car, and with the way my car is set up, that this change is beneficial, but your mileage may vary...... this is for educational purposes only :lala:

Grim Reaper

Yes, there is a difference in the locations. This is the "extra code" that is in the AUJP. I thought it was 6 bytes, but I haven't checked that out in awhile.

John Millican

iTrader: (1)

I changed the 7E E1 A2 to 00 00 00 and got a SES light as soon as it went into closed loop. I could not pull any code, it did not flash only steady SES. I shut down and re-started, still got a SES light and cannot pull a code.

I then changed it to 01 01 01 like this thread said. As soon as I started it I got a SES light solid, could not pull a code.


I then went back to my orginal BIN and everything is fine.

What happened?

hectorsn

Did you update or disable the checksum?

91L98Z28

oops! yes, you have to correct the checksum. I used winbin to do it, but tunercat has a program to do this as well. or you could just disable it.

I'm still running the code with 01 01 01 to disable the TCC decel unlock and it still works great. BTW, you can't use 00 00 00 (i tried that myself by accident), 00 is a reserved opcode for testing or some such thing. 01 is a NOOP (no operation), which is ECM-speak for "do nothing here and continue on to the next opcode".

John Millican

iTrader: (1)

What do I do with the checksum? I have Tunercat but it doesn't mention the checksum at all.

91L98Z28

Here's tunercats' program for manually editing checksums:

http://www.tunercat.com/chksum_desc/cksum_desc.html

basically, you make the changes described above (the 01 01 01 stuff) and then recalculate the checksum on the entire .BIN with those changes in place. the new checksum is written into the .BIN and then the ECM is happy.

You could also use winbin to do this, it's free but probably less user friendly.

John Millican

iTrader: (1)

Ah, I understand now. Thanks!

Matt87GTA

Very nice work!

John Millican

iTrader: (1)

I did it. Works GREAT!

Man I love just crusing now, no unlocking and locking in traffic at speeds above the lockup settings just because I let up on the throttle.

91L98Z28

Which is exactly the same reason I did it in the first place. It's so much nicer not having to deal with the constant unlocking and whatnot.

The only downside is that the TCC will still unlock if you touch the brakes, and then relock as soon as you let off the brakes which can be a bit annoying. I am considering seperating the brakes from the TCC (so that touching teh brakes won't unlock the TCC), when/if I do that I'll post results.

RPOL98

I implemented these mods so that the TCC no longer unlocks when decelerating. works perfect, very happy in that category.

However, I notice that now, the engine maintains an elevated idle speed of 1200-1500 rpm, until all of the following conditions are met:
1. closed loop.
2. vehicle comes to complete stop. i.e. VSS says 0 and stays at 0.
3. shifter is in neutral.
4. and then idle will gradually decrease to the appropriate programmed idle rpm level set in idle engine speed vs temp (725rpm for my settings). and this takes a couple seconds, and you have to patiently wait for it to settle at the lower rpm before moving on.

folks behind at the stop sign wonder why I'm dawdling since there's no other cars to wait for.

after that, any time I come to a stop, the idle drops as it should. however, if I haven't done this complete drill for each engine run cycle, the idle stays high.

It kinda is annoying. makes me use the brakes more than I want to, just to come to a rolling stop, since the elevated rpm wants to maintain a higher vehicle speed. I still like the TCC decel lock more than this anomaly annoys me, though.

any thoughts anybody? did I miss something?

91L98Z28

well, there's one from the past.

I went to a T56 some years ago now, so I no longer have this change or an $8D auto-trans car to test it on.

I don't recall having the problem you describe, and I certainly never had to put it in neutral to "control" idle speed. I was running a 2400 stall 9.5" converter, so it's possible that if the idle RPM was elevated a few hundred RPM for a second or two, I may not have noticed it... but I certainly don't recall it.

With that said, I don't really see how this change could do what you're saying. The TCC is going to unlock when you cross the low-VSS threshold regardless of this code change, so there should be plenty of time for idle speed control to take over and manage the idle RPM. Are you sure it wasn't doing this before the change? try going back to the pre-change BIN.

Other things to look at:

How high are your IAC steps at idle?

Does the problem occur if you leave the trans in gear? Some of the rpm routines are a bit different depending on the P/N switch position (you do have it hooked up, yes?)

There is a setting for time between IAC steps, and decreasing this time might reduce your problem (i.e. the IAC will step out faster, and your rpm will return to idle faster).

There is also an IAC step adder for TPS position and I believe MPH as well, which can be played with. If there is an IAC step adder for MPH, maybe you can pull more IAC steps out at a higher MPH so you don't end up with a bunch of IAC steps as you come to a stop. Or maybe try disabling the whole thing.

Is the code running as it's supposed to? That is to say, you're actually getting into idle mode, idle timing, etc.? There are thresholds for max TPS position for idle, etc; perhaps setting the threshold for idle to a slightly higher MPH will help.

Finally - on similiar topic: I own an 88 fiero L4 5spd (2.5 TBI SD). Since the day I bought the car (nearly a decade ago now), it has two irregular habits, one of which is EXACTLY what you describe - the idle rpm will be high (1500-1800), and it won't come down until you come to a complete stop..then the IAC extends to the normal idle position and she idles fine for the rest of the drive. This happens maybe once in 10 drives and seems to typically occur when I start up and immediately drive away. If I let it sit for 30-40 seconds before driving, the IAC extends enough from the startup position that the problem basically is non-existent.

The other habit is that it will lose control of the idle altogether - the idle will end up hanging around 1800. ALDL stream shows TPS at 0%, and the IAC with a bunch of steps, but it won't extend the IAC (reduce steps) to control the idle, nor will it set a CEL. It just sits there fat dumb and happy at 1800 as if everything is just fine. Key off and then restart (with nothing else touched) immediately resolves the problem. It only happens on a warm start (such as when it's been 30-90 minutes since the last drive). Gotta be a bug in the code somewhere, tried to HAC it myself and only got about 1% of the way through...

RPOL98

thanks for the reply. lots of things to sift through. I do indeed have the factory gear selector sector switch hooked up (P/N, etc).

if I leave the trans in gear at a stop, the factory stock torque converter at a full stop prevents the car from idling too high, so not as noticable as in neutral.

I tried playing with the IAC steps/timers before to get that hot-rod "throttle blip", but that created some really weird idle issues, one of them was that idle "hunting" like on a roots-blown car, not drivable, so I went back to stock settings.

now that you mention it, this might've cropped up around the time I went to the SAUJP bin as a starter/template to get the benefit of the extended RPM map, of course throwing in my own VE map, idle/temp, etc. Probably imported some constant that didn't jive with something else I mod'ed. Otherwise, everything else is hunky-dory.

guilty of playing with multiple variables which hampers symptom/cause fault isolation.

thanks again.