Well, I'm finally stumped... v.WTF?

Well, today I went to tune an '89 Cavalier with a hybrid (frankenstein) 60 degree V6...

Engine looks to be well put together, and the owner has taken a lot of time to make it a very clean install.

It uses a '7730, with $6D (Due to vacuum EGR), which I'm quite familiar with.

The owner bought an Ostrich 2.0 and ALDL Xtreme, so we unpackaged those items, connected everything, uploaded the BIN, turned the key on, CEL blinks, good, checked ALDL connection, good. I changed the Mask ID to AA to disable the checksum for emulation. I made an adjustment to the injector constant, which when started seems to have good results, having the cold start about 12:1, instead of 10:1. So we let it warm up a bit, I make a few tweaks to the injector constant, just to try and lean it out a bit more, it does.

Once warmed up, I tried to adjust the idle and main VE tables, with ZERO response from the engine. So I made larger changes, still zero effect. Adjusted both tables drastically, both up and down, zero effect. After a bit decided to try $A1, with the EXACT SAME results. No changes in the VE or even timing tables had any effect on the WBO2 reading or how the engine ran at all. So I went back to $6D to try and figure out what is going on, since $A1 won't be proper anyway, due to the vacuum EGR. I tried for another 2 and a half hours or so with a complete failure on the ability to tune this car.

I can make any changes to a scaler and it has the expected effect, however changing any tables has zero effect. I tried 2 different Ostrich 2.0s, 2 different ECMs, with no change in the symptoms. I even unplugged the Ostrich while the engine was running to prove that it was indeed running off the Ostrich, engine stumbled and died (LHM will not be close enough due to injector size and fuel pressure), so it was for sure reading the Ostrich contents. I also downloaded the contents of the Ostrich a few times to make sure that the uploaded changes were indeed uploaded to the Ostrich.

So to recap:

ECM: '7730
Mask: $6D ($A1 also tested)
Ostrich 2.0 (tried 2 of them)
Tried 2 different ECMs
Tuner Pro RT
XDF that was downloaded from this site today, since I didn't seem to have one on my laptop.
Can change scalers with changes to the engine (though not as directly as I would have expected), but any table changes had zero effect.

I plan to setup my test bench and see if I can replicate the problem, while observing the outputs via oscilloscope, and any other tools I need to use.

Any ideas out there?

I would eliminate the Ostrich 2.0, for the time being. I would try a spare factory Original $6D Memcal. I would UV erase the PROM in the spare factory Original $6D Memcal and reprogram the PROM with the "Starter" programming. I would likely UV erase and reprogram the spare factory Original $6D Memcal PROM with same changes that were tried with the Ostrich 2.0.

I don't think there is a problem with the Ostrich 2.0 or ECM, maybe the Memcal is the problem part?

dave w

Quote:

---

Originally Posted by dave w

I would eliminate the Ostrich 2.0, for the time being. I would try a spare factory Original $6D Memcal. I would UV erase the PROM in the spare factory Original $6D Memcal and reprogram the PROM with the "Starter" programming. I would likely UV erase and reprogram the spare factory Original $6D Memcal PROM with same changes that were tried with the Ostrich 2.0.

I don't think there is a problem with the Ostrich 2.0 or ECM, maybe the Memcal is the problem part?

dave w

---

I don't have a way to UV erase. I had thought about programming an EEPROM, but that won't solve the problem I have. It runs off the MEMCAL, it just runs very rich, due to the larger injectors with higher fuel pressure and other engine changes.

The fact that scaler changes do have effect, but table changes don't (while emulating) is what is perplexing to me.

The only similar thing I've experienced previously was when I have forgotten to disable checksum and the ECM went into a quasi LHM while emulating, but changes still could affect the way the engine ran, just not as directly as when everything was correct. That was on a '7060 though on a very different application. Checksum was definitely disabled on the Cavalier, I had to manually edit the file with the hex editor, since it seem people that wrote the XDFs for $6D and $A1 didn't define the mask ID bit... *shrug*

At one point I changed the entire tables to 0 for both the idle and main VE tables and the car continued to run as it was. It wasn't in LHM, as the CEL wasn't on during any of this and we could trigger LHM manually by unplugging the Ostrich, at which point the fan would come on.

I also don't believe it's the ECM or Ostrich, since I know my Ostrich works just fine, his was new (which I know is not a guarantee that it's good, but 2 Ostriches being bad in that scenario is unlikely), Like I said originally I tried two different ECMs, and both worked the same.

I also don't believe that it's the G1, or the ribbon cable, because anytime I've had ANY issue with those the ECM would go into LHM, or throw codes (if the problem was on the MEMCAL side for the adapter).

The only codes that did pop up were for knock sensor which the owner says has happened before, and I'll be looking into that more, but is likely just because GM sets the knock sensors to be sensitive, and it's picking up the noise from the new valve train, and one for the )2 sensor, when we did try to drive, but the NB O2 sensor was unplugged at the tine, since my WBO2 was installed in place of it. So I'm not worried about either of those codes.

I have zero experience with emulating but what u describe is exactly what I get on tprt when the incorrect xdf is trying to read a bin.I have a cadillac bin but no xdf for it.when I tried all kinds of guessing to see if I could find an xdf that would read it I got most of the scalers but none of the tables would read.I have no clue if the scalers would even be correct but what the tables read was way negative numbers and way hi numbers and was very obviously wrong because of wrong xdf.just thought your symptom sounded a lot like my trying to read a bin with incorrect xdf.hth

I don't have a bunch of ideas. For fuel I'm wondering if there's a minimum PW value... but you should have been able to increase fuel. Maybe you can alter the datastream values in code so the PromID outputs fuel or timing values from memory? At least you could see the values change or not change.

Switch settings on the Ostrich 2.0 configured correctly for 28 pin chip? Reload the Ostrich 2.0 driver. Operate TunerPro RT in Windows XP Service Pack 2 or 3.

dave w

UV erase by sitting in direct sun with no metallic sticker cover for 12 hours. Almost always works for me-and if it's not totally erased when you do a read on it prior to programming, let it sit a second day.

Or go buy a plant light and that normally does it, takes 12-24 hours vs 15 minutes with a real UV eraser, but it works too.