Finally getting another Delco project done, or started....

[Six_Shooter]

I have a friend with an '85 Turbo Regal, that has some engine work done to it, cam, ported heads, some other miscellaneous things. It's never ran right since he did all of this (years ago now). A couple years ago I took the car for a couple weeks, and had limited success tuning the 1227148 ECM that is in it. So I decided that to best plan for this car would be to swap to a 1227730 or 1227749 and use $58 to run it. I want the transition to go smoothly, and there is still some question of how well this will work in mind, so I built an adapter harness, to plug into the existing harness and then into the new ECM. That way if it takes more time to get a base tune done, the original ECM can be plugged back in and he can at least drive it.

Like I said this was a couple years ago, I started the adapter then, but got stalled on a few things, because I wasn't sure how I was going to actually terminate a few things, and wanted to retain the original ESC, to make the swap not only easier, but replacement parts, if they are ever needed, will be easier to remember what to get, since they will be stock parts.

Anyway, I have some pictures of the adapter:

I pirated the connector to attach to the car side from a failed ECM, and then using some ECM connectors and pins I had kicking around soldered the wires to crimp on PC pins, which I then soldered to the pins of the connector. I did this, because it makes a stronger connection than just soldering the wire to the pin directly, like I've done in the past.





Added some heatshrink:


The green wire is for the MAP sensor, that I will need to tie into elsewhere, since the ECM only uses the MAF, the MAP sensor is for the gauge only, in stock form, now it will run both.


I also modified a MEMCAL to use in this, don't worry, it's just a common 2.8L V6 MEMCAL, ARYF, if you're interested, that I removed the knock filter from, and added a 4K resistor, that should help keep code 43 at bay, and removed the PROM for an EPROM socket, to make tuning and flashing easier. Eventually I want to reverse engineer my TGP MEMCAL NETRES, and replicate it for this car, but for now, this will work.



I should be doing an initial tune this weekend.

[1project2many]

Nice job on the harness. Isn't it's amazing how much time can pass between when you set down a project and when you pick it back up?

I believe I have a memcal here from a 3.8. That would provide and integrated solution for the knock filter.

[Six_Shooter]

I have a 3.8 MEMCAL as well, but I didn't want to use the internal knock filter, for ease of swapping, I want to keep the OEM ESC. I also didn't want to swap out the knock sensor, for a different one, and I'm not totally convinced the biasing resistor doesn't change the way the sensor works, or at least the signal from it.

[Six_Shooter]

Well, I got the TR in my possession this morning, I only had a few hours to play with it, due to father's day obligations (not my idea).

It started off not too bad, found the MAP wire inside the car literally inches from the ECM, it actually comes into the car in the same harness. Wicked, that was easy, which should have also been clue #1 that something wasn't going to go well. So I added a length of wire to connect into the MAP wire in the adapter harness. Everything plugged in easy and seemed to work.

I plugged in my AutoProm (first time actually using it), uploaded a $58 bin, turned the key on, started the datalogging, everything is good. Start the car, get it to run, albeit not well, but it ran, on a '7749 and $58. After a few minutes of changing a few things, it was starting to improve and then bam, no ALDL connection. The wouldn't start either. WTF? Everything pointed to a corrupt bin, especially when simply loading a bin to Tuner Pro without changing anything, it would give me a "Save modified bin" pop-up when I wanted to switch to a different bin, I CHANGED NOTHING! It's odd. Shut down and restarted Tuner Pro several times, even tried V4 with the same results.

The odd part was I could load any $59 or $8D bin, with the appropriate XDF and ADX files and they would connect without issue. I had what seemed like corrupt bin results when I tried $8F bins. So the two codes that would work and connect fine, I couldn't actually run the car with, due to the hardware not matching what those codes were meant to be used with. Frustration set in.

By this time, I had to go and I took the car back on the original ECM, explained a bit about it to the owner, and said I would figure it out.

So on the hour trip to the Father's day celebrations, the several hours we were there, I researched this problem online. I found several posts indicating the same issue, and that there was a fix, great. The problem was that all of the links to the fix were dead, or at least the links to the patch utility and the actual patch. So close.

Well, I did more searching, and just before we were about to leave to come home I stumble upon this thread: (ALDLPatch for $58 on TGO). So, I tried using the ALDLPatch doing it without a patch utility, like I've done with other codes (E-fan control on $OD), and it works perfectly. I tested this on my bench.

FWIW, I did test the stock bin with an Ostrich and a MAX232 based home made serial cable and Moates Xtreme ALDL, which it worked fine with, on my bench. This issue seems to affect the AutoProm mostly. In my research it has been said that it will help when using other devices as well, like other serial cables and programs.

I'm attaching the patch here, to make sure it's preserved.

So now to get the car back and try again.

[EagleMark]

Why don't you post that patch in your $58 Info thread as well... Nice work on the adapter!

[Six_Shooter]

Why don't you post that patch in your $58 Info thread as well... Nice work on the adapter!

Good idea, I did include a link in the FAQ.

[Six_Shooter]

Well, I got the car back again this evening...

This car hates me.

While I did make some progress, I just can't seem to get it to idle nicely, or as nicely as the '7148 ECM does, even though SA is in line, and fueling seems to be inline (just using the NBO2 at this point due to limited time to be able to swap the WBO2 in). While fueling changes will effect the way the engine runs, it doesn't make any drastic changes, at least no head way. Adjusting the F1 table seems to do virtually nothing, but adjusting one of the bias tables does seem to effect the SA as measured at the crank with a timing light. This is with $58. I have issues with using $58 it seems. Sometimes it will upload and connect to the datalogging side without issue, other times it won't. I ended up going back to my standby equipment, ALDU1 and Ostrich 2.0, since the AutoProm doesn't seem to want to work in the car.

I did try an $8F bin, and it seemed to start and idle well, much easier than the $58 bins I tried, but I just don't have a well defined XDF or ADX for $8F, so I haven't pursued using $8F for this application.

I tried using setting that 1project2many posted in a bin on the code59 website, and using a different technique, that I used in my Datsun, for the DIS, I think there's something I'm missing from 1project2many's settings, because the only drastic change I see there is spark reference angle, but at the setting it has, there should be 60* of timing removed from the over SA. Min and Max advance/retard have huge values, which I don't quite understand how those values were arrived at. I had the best luck using the bias manipulation like I used in my car, but I'm trying to understand this other way of making it work. I have also referenced a couple other DIS $59 bins and they use other settings yet again, only adding to the confusion, of settings that work, or are supposed to work.

I've tried in vain to find out what Grumpy did in his GN years ago, as far as ignition settings go, but it seems like he didn't publish those settings, at least not to any website I've searched. I have a feeling he did post to the DIY_EFI news list, but many of those posts were lost, or seem to be inaccessible.

Does anyone have any ideas, other than pulling my hair out at the roots? I've never had a car fight me this much. I had more success running a 365 with a home made MPFI intake off a '7747 than I have with this car so far.

[EagleMark]

Well, I got the car back again this evening...

This car hates me.

I've got a loaded $8F XDF and a pretty good looking $8F ADX imported from ADS so may need work... also attached a $8F hack just inn case...

Not sure why you can't get $59 working?

[Six_Shooter]

I've got a loaded $8F XDF and a pretty good looking $8F ADX imported from ADS so may need work... also attached a $8F hack just inn case...

Not sure why you can't get $59 working?

I'll have to look at what you've uploaded there.

To run $59 would require a 3 BAR MAP sensor, and for this application, the resolution of $58 should be fine, though I am seriously considering getting him a 3 BAR MAP sensor.

[EagleMark]

I'm not sure what it is? But the XDF is huge. I've been collecting and been donated so much stuff I needed a web server to store it all...

Hope it helps. BTW I didn't have any bins in that folder...

[34blazer]

how is the engine acting with $58? surging, stalling, etc?


is this a hot-air setup?

[1project2many]

I've tried in vain to find out what Grumpy did in his GN years ago, as far as ignition settings go, but it seems like he didn't publish those settings, at least not to any website I've searched. I have a feeling he did post to the DIY_EFI news list, but many of those posts were lost, or seem to be inaccessible.

Grumpy didn't post the settings. After I worked out the values for my car I gave him a set of values for his and asked him not to let them out. Maybe a couple of years after we left DIY_EFI I told him it was probably OK so he might have put them on gnttype.org. I also gave him the voltage range to use for the stock GM knock sensor so the ecm wouldn't set a KS code. Happy to help if I can, in any way I can including digging through old emails to look for those settings. I've had problems with $58 and Moates hardware/software combinations. Never did get the thing to emulate correctly over long term. Might have been laptop, or software, or ?? I'd get two, maybe three uploads then it would stop emulating and I'd have to reboot the laptop.

When I worked out settings for Bruce I didn't have any stockers to look at and the ecm hacs for some of the GN ecm's weren't public knowledge. Plus, Bruce had been playing with different ignition modules and crank sensors/pulley between the fast start types and the older versions. Rather than try and piece together what Doc Plecan was running, it's easier to give stock settings. Right now I have them for the 86-87 cars. If you need an older one I may have to do some disassembly work to pull them from a calibration.


Reference angle 70 deg
Max advance 15.1 deg
Max retard -70 deg
GM used a "fast start" module on later cars which may explain any difference between previously posted values.

Sleep is winning... gotta go.

[1project2many]

Six... Are you there?

Adjusting the F1 table seems to do virtually nothing, but adjusting one of the bias tables does seem to effect the SA as measured at the crank with a timing light.

Missed this yesterday. You're absolutely correct! Adjusting most cells of the F1 table at idle does absolutely nothing for spark advance at idle because spark advance is limited to using the first row of F1 with closed throttle! After many hours of studying this I believe GM did this with the intention of using the spark advance based idle speed corrections. Too bad all the cals I've seen have this feature disabled. I posted a code change to code59 which checked for vehicle speed and closed throttle before limiting SA lookup but it's easier to just eliminate the closed throttle test.

Using a hex editor manually change the following code

Address     Values           Change to:
$10B8         13 29              20 03

This will force the entire F1 table to be used at all times. HTH.

[Six_Shooter]

I know about the 600 RM row = idle SA. ;) It's the same as $59, and also documented in the P4 document.

Hmmm, I'll have to look into using that change. Any idea if it would work on $59? Location would likely be different though, eh?

I added a switch to the bypass wire in the harness adapter. On any other car, that I've used a switch in the bypass wire, I can hear or sense a change in the way the engine runs, from bypass enabled to not, but this car just runs the same no matter what position the switch is in. It's not adding up to me.

I'm also very tired right now, I was up until 3:30 this morning working on home work (upgrading high school math, to make college entrance easier), and have very little sleep, so I'll have to look at this some point soon, after more rest.

Thanks

[RobertISaar]

you guys don't have a flag to enable/disable the lowest RPM cell based spark advance option? i hate it and is the first thing i disable in any tune i play with.