http://www.moates.net/fileman/index....%20ECU%20Hacks

I've uploaded ecu files with all necessary documentation to patch a 160 baud ALDL datastream.
This is for the 7747 and 8746, sorry to the rest but I don't have commented RAM for the others.
In layman's terms, these ecu files will modify your ALDL datastream by having the eprom tell the ecu to spit out different data. Instead of battery voltage being datalogged with winaldl (as an example) you could have the ecm's desired AFR. Another good change is to use the eprom id msb and eprom id lsb for things like Final SA and the minor knock counter (PA3 lsb).
Keep in mind that if you use final SA that the actual physical dizzy advance isn't considered so you need to add that value in. Also make certain that in the eprom, your initial SA is set to the same as your dizzy's physical advance.

 

Jon,

Thanks. Trying to remember how you told me to do this during the ECM POW WOW.

Dominic

 

Do you still have questions?

 

Jon,

This is one thing I couldn't quite get my head around. I think I know how to change the battery volts to Synch BPW etc. BUT,
how would I send another item to an area that is currently not defined in the ECU? Should I just create it in the ECU and then change the address to the item I want displayed. For example, I have a 4th gear switch setting available on my EASE scantool.
I could send Spark Advance to it but the 4th gear switch is not currently defined in the 7747 ECU I have.

 

FYI, I moved this into the Bin Definitions (ECU) files section (Jon originally uploaded to the hacks/source code section on accident).

Now its here:
http://www.moates.net/files/3)%20Bin...DL%20patch.zip

 

Real nice work Jon, I'm gonna have some fun with this.

 

It is in the ecu file but since 4th gear is just a switch it's only a bit (0 or 1) it's included with 8 other switchs in the 16. mcu2 i/o
That looks at the hardware location L0801 which isn't RAM and isn't EPROM or ROM! What is that location exactly? I've seen it and I know how to work with it but I don't know where it is physically in the ecm.... anybody?
So if you don't want your switchs to work then change that to something else... but you're limited to the software you're using. If you're using EASE then the 4th gear switch won't display something that's numeric like final SA, instead it'll just display/record the # bit of that byte (8 bits). Is this still too confusing?

edit; Maybe Joby will release a new version of WinALDL that let's you define your own definition files or does TunerPro already do this? I can't keep up with these things. At one point I had source code for WinALDL that I was going to patch for this kind of a modifcation but then I got lockers and never needed the mod.

 

Hey Dominic,
I am almost finished creating a tunerpro aldl definition to monitor SA and BPW in those two locations. I still have to finish testing it yet..... and get tunerpro communicating properly with my laptop...
Tunerpro lets you bias the display also, so its dead easy to add your base timing to the transmitted SA to give a total, at the crank number to display ;-)
This will become redundant once ken/bill/bob get the speedreader available...
Ben.

 

Ben,

Jon and RBob :hail: spoiled me with Lockers while at the ECM POW-WOW. 17 frames a second versus 1/second makes tuning so much easier. So I anxiously await Ken and Bill's efforts with SpeedReader. After you test out that def file maybe send it my way. Monitoring SA and Synch PW is what I am looking for. We bumped the FP on my car from 10 to 16psi. Fuund that the injectors were static and AFR was 14:1 at WOT. Not good. It "feels" like it is running quite a bit stronger now but the BLMs are in the low 100 range. But that's what tuning is all about.

 

Thats awesome info Dom.. So your 80lb'ers were static at 10psi, but ok at 16?? I *think* my 90's are ok at 15psi, only because when on the dyno it didn't lean out up top.... That was with the X-ram.. When i fitted the SY-1, the blms jumped to 140's and a few 150's! So i don't know where i am at right now..
Lockers would be sooooo handy right now.. I cannot wait for the speedreader to be available.
I will send the aldl definition to you as soon as i am happy it is right.
Ben.

 

Cool stuff, Jon! Thanks!

 

I was in the middle of trying this same idea out with the new 8747 ecu file I am making for Tunerpro, and was hoping it would work. Going to try a prom 2day, i hope.

But the question i have is..

I was looking through the hac trying to find the actual xmit code for the ALDL routine (160 baud). What would happen if i replaced all the data with, say, the bpw and target afr, just over and over. Does the routine update just once a second, or does it update right before it sends the data? Could i possibly get 10 or 20 actual data points per second, on just one or two variables? Does the question make sense? Hard to explain. It would not be realtime data, it might be up to 1 second old, but I think that might be useful, in some cases

 

I see what you are getting at and I really think you are barking up the wrong tree. Everything I have seen/read on this board says that 160-baud is gonna get a certain frame of data each second and that the data is already old by the time you see it. Your idea is awesome and if you can manage to do it, make sure you let the rest of us 160-baud guys in on the secret.
If you get that 8747 ALDLdef to work, how bout sending a copy of it my way? I cannot get mine to work for the 8746.

 

7Point,

This may be relevant to your idea. I use EASE for scanning and they say that you can increase update rate by limiting the number of parameters you are scanning for.

 

You can replace the data stream and put just the two parameters in every other location. Except: for the first byte. The first byte of the data stream is hardcoded into the ALDL routine. The ALDL address list is in the EPROM and can be changed.

The data is sent real time. IOW each byte is grabbed from memory when it is time to transmit that byte of data.

So it is easy to xmit the BPW (2 bytes) and the O2 reading (one byte) in alternating fashion.

RBob.

 

Its amazing the amount of info and knowledge that is around.


Now if i can just figure out tunerpro's aldl quirks, i can use some of this great info. Was using winaldl, and it worked, but thinking some of the bytes are different between the 7747 def i have to use and the 8747 i actually have. Along with if i can do what we are talking about, it will confuse the heck out of WinALDL (A nice program, no doubt).

 

Has anyone made the changes Jon has provided. I have tried using the Prom ID for Total SA but all I get are either the base or max timing. Same goes for Synch PW. What am I doing wrong?

 

What is being captured in a logged run?
If you can upload your bin file I'll look at it and see if there's anything wrong.

 

Jon,

How are you? Thanks for responding. I am not sure where to put the SA, Synch PW etc. The SA for instance has an lsb and msb address. Which do you use? Remember, I am using a 7747. BTW, trying to rid myself of that dreaded oscillating idle with the increased FP.

 

Try both, I can't remember which worked for me, I remember having a problem with it and only one was working. Same with Synch PW. It should be msb but you never know.
The oscillating at idle, is that when you've gone closed loop? If so, I'd check out the prop gains, that's what fixed my idle oscillations with the new o2 sensor. The old sensor was so slow my idle would oscillate but it took it a WHILE fore the idle to be "fast" and then "slow."

 

Jon,

On the SA, I tried both. One gave me Total advance reading of a constant 41.8. Sounds like the max advance. The other gave me a constant reading of 24.7 which was my min timing. Neither of them changed during the course of the scan.
I think I have the oscillation fixed now. I set the IAC controls to what they were when I drove back from the POW WOW. My suspicion is that my Stall saver was set to exit at 600rpm while the idle was at 850. Then the Stall saver decay coef was set at 19. Very slow. So when I would add load to the motor while sitting at a light for example, the motor would drop down enough to trigger the stall saver. Since that was slow to decay, it would keep bouncing of each end until finally the motor would shut down. I am hoping to scan tonight and see if I am close to the 128 Holy Grail. I first roughed it in using just the INT. Then I unlocked the BLMs so I am using that now.

 

That's odd that you're getting those values. I wonder what your bias' are set at. The worst case senario is that one of the bias values is off by ~20 degrees and you really ARE running max timing all the time, or vice versa. I'd look at all the tables and do the calculations by hand to be certain.
One other thing with tuning the idle timing, if you're right at the limit of too much timing the car will feel like it's going in and out of stall saver. It usually will idle really strong and then nose over itself as an indicator and get to about 500rpm before stall saver. That's juts another thing to try. With a stick you almost want to remove some timing at idle because when you're on the edge, just a flick of the throttle and the revs go up fast. Take some timing out and it'll be easier to launch... that's what I did with a friends Ford 5.0 with just headers and a TKO. That might not be universal but I'm just stating things that have seemed to help with getting a nice idle.

 

If I could suggest sending your calibration to somebody that has a test bench to test your calibration. I think Grumpy or RBob might be able to do it since I know for a fact they still have c3's around .
I would do it but my test bench is in Ohio and I'm in NJ right now.

 

Jon,

Thanks. Its been an interesting tuning exercise to say the least.
I started with your bin and locked the BLMs to rough in the tune using the INT. I had gotten it to a pretty good point where the INT was close to 128. Then I unlocked the BLM's and found that it was running richer than the INT indicated. RBobs comment about the BLMs being more accurate came to mind. However, I experienced a stumbling at idle which appeared to take place whenever the BLMs would reset due to the idle crossing over from one BLM cell to the other. So I disabled the BLM reset which then caused the idle to seesaw. Normal cruise was fine although still rich. I also continued to experience a slight stumble at very very light cruise throttle or decel. Especially in the 2000rpm range.
I had reverted to the lower FP and original cal just to eliminate a vacuum leak as the cause. Motor ran fine. So I put FP back up again. Last night I unplugged the IAC's and I was still experiencing a seesawing of the idle. Increasing the rpms with no load on the motor, it will not keep a steady rpm. Checked for any obvious vacuum leaks and didn't find any. The motor will rev up to 1600rpm, once I free wheel it rolling to a stop. It'll stay there and seesaw up and downto 1200rpm. After several minutes it will idle down to 900 rpm where it will sit until I hit throttle again. Then cycle will begin again.
I have forwardrd you my latest bin file.

 

That really sounds like a prop gain problem. It's overcompensating the fuel and causing it to go up and down, although that's an awefully high rpm to have that problem it still might be an issue.
SA might also be too high causing it to stop itself, I'll look at the bin later tonight and get back to you.

 

It may be having a pulsewidth size problem. With higher fuel pressure, it takes a shorter pulse to deliver the same amount of fuel. If the PW gets shorter than about 1.3 ms total (including injector turn on time), it gets nonlinear in a hurry. The ECU normally combats this with Quasi-Async where the pulses are roughly doubled but injected only half as often to deliver the same amount of fuel with consistent PWs. Maybe there is some parameter that deals with this that is now incorrect for the higher pressure and such. Also, maybe the block learn boundaries are different bringing one near the idle point, in which an oscillation feeds itself. Move them away from the idle MAP and RPM.

 

Red,

RBob uses the same size injectors (BB 80#) on his SBC. He's running 22psi while I am at 16psi. I do have the BPW set at 96 which is where Jon had it when we first bumped the FP.

Jon,

I have checked the bin I drove home with versus the one I sent you. I obviously have taken out a ton of fuel and bumped the timing by at least 2d at the 30kPa level. I may go home tonight and drop the timing and increase the fuel in that area and see what happens. I set the Prop Gains to where they were at the POW WOW.

 

One question,

When your timing reads in, it will display 10.5 deg, as 105 in the data stream??

I'm confused because I get values that move up and down with the rpms like they should, but there like 70 to 90, when My spark output should be 15 to 25 deg.

 

70 * (90/256) = 24.6 and 90 * (90/256) = 31.6
Sounds like it's working to me, you just forgot to do the math.
Remember, the ecm is looking at things in binary, not real-world values. Another example, RPM, the ecm looks at RPM in a few different ways, the 2 easiest to understand is the RPM = raw value / 12.5 and RPM = raw value / 25 . So remember that RPM is in the ram as an 8-bit value. So that means it can only be $FF large, that's 255 in decimal. 255rpm is rather small so the ecm multiplies by 12.5 and 25, in other words the ecm reads rpm with resolution values no smaller than 12. So 255 * 12.5 = 3187rpm which is still too small but good for controlling idle speed (higher resolution). For fuel calculations it only makes sence that the ecm willl look up to the max VE tables (duh) which go to 6400rpm. So 255 * 25 = 6375rpm . Is it making more sence now?
The comment ram hacks have equations in the comments side to convert the machine numbers into real world numbers.

 

Thanks Jon, thats the answer I was looking for.

I knew about the rpm coversion, and the pw, but this is the first I've seen of the spark conversion.

 

As a followup, it turned out to be the injectors. I had noticed that the spray pattern of at least one injectors was not fan shaped. After countless times tuning and changing other sensors, I finally sprang for the injectors. It worked. The injector must have been on the edge. They worked well enough at cruise, but where the PW really needs to be controlled, idle and decel, they couldn't handle it.