I was finally able to play around with this again and have some interesting results.
I started out with the above bin with just changing the injector size to 19 multiplying the crank pw by 1.57 as instructed. With this it would pop and sputter but not come alive and it really seemed like it wasn't firing on all 8. I then played around with the crank pw, first adding more pw, which made things worse, then taking it away. It actually worked best after dividing the original crank pw values by 1.57, so smaller than the original AJON values. This may have been because of an excess of fuel sitting around from the previous attempts, but this couldn't be the entire problem either. With the smaller crank pw numbers, we were able to make it keep running, barely, by pumping the throttle. Any constant throttle wouldn't do it, but pumping it would keep it alive. I wanted to keep it running long enough to get some heat in the header to verify which cylinders were firing, because it was obvious at this point that it wasn't running on all 8. Turns out, only the front four cylinders were warm, so it was definitely not firing all of them.
From there, just to see what would happen I changed the number of injector output ports to 8 and number of injectors per output port to 1 (originally 2 and 4) with no discernable difference in results. I did this since this is a sequential injection ecu and I was curious. I didn't expect this to work considering the xdf is for batch fire I believe, but I wanted to cover my bases.
The last thing I did to play around was to mess with the number of pips between injector outputs, just for fun. Again I didn't expect this to help, I just wanted to see what would happen. It was originally at 4, first I set it to 2, which made it so only two cylinders were firing, though it did stay running with throttle pumping. Oddly enough only cylinders 1 and 2 were the ones firing. Then I tried setting it to 8, which just made it garble around, backfire and not run at all.
After that we called it quits and unplugged the quarterhorse to verify that the car still functioned without it, which it did, so no harm no foul even with the fooling around. Needless to say I didn't play around with the ve or anything else since I couldn't even get it to idle.
I'd like to dive deeper into how all the functions and tables work together, but besides reading the notes, I'm not sure how to do that. A flow diagram would be great, but I doubt that exists as public info. So, could anyone share what the best way for me to learn the inner workings of all this is? Thanks for all the help so far, I'm looking forward to any responses and getting a better understanding of what’s happening.
Definition for SFI/SD3
-
Dashing ZX2
- Posts: 8
- Joined: 2022 Mar 10, 13:04
- Location: College Place, Wa
- Vehicle Information: 1999 Ford Escort ZX2, stock drivetrain
-
decipha
- Posts: 7101
- Joined: 2021 Feb 15, 12:23
- Location: Metairie, LA
- Vehicle Information: 2x '19 PU Explorer Sport 3.5L Eco's AWD
Re: Definition for SFI/SD3
its not firing the other bank because the port assignment is wrong
Thats typical when using an incompatible hardware code for that strategy.
Thats typical when using an incompatible hardware code for that strategy.
-
whitfield_72
- Posts: 5
- Joined: 2026 Oct 20, 10:57
- Location: Berlin, Germany
- Vehicle Information: 2005 Subaru WRX
Re: Definition for SFI/SD3
decipha how do you actually tell it's the port assignment though? is that the active injector mapping table in the xdf or is it baked into the code somewhere. if its the table then could he just remap it and see which cylinders wake up. cheers anyway.decipha wrote:its not firing the other bank because the port assignment is wrong
-
decipha
- Posts: 7101
- Joined: 2021 Feb 15, 12:23
- Location: Metairie, LA
- Vehicle Information: 2x '19 PU Explorer Sport 3.5L Eco's AWD
Re: Definition for SFI/SD3
Thats literally the dffference between strategies which are based on a given hardware code.
The hardware code is what dictates the port assignment and which memory registers they are assigned to.
For example some strategies in the 80s used register 47 as a BIDI port while others used it as a scratch register.
If you load a strategy that was compatible 'enough' to actually run you could literally be inadvertently turning outputs on and off simply by the code doing basic math going through its routines.
For example, you wouldnt want the diverter to actuate everytime the ecu computed a background value that just so happened to be a 2s complement divisible by 8.
Yeah, you could theoretically go through and reassign each reference, but by that time your probably better off just using the compatible strategy.
The hardware code is what dictates the port assignment and which memory registers they are assigned to.
For example some strategies in the 80s used register 47 as a BIDI port while others used it as a scratch register.
If you load a strategy that was compatible 'enough' to actually run you could literally be inadvertently turning outputs on and off simply by the code doing basic math going through its routines.
For example, you wouldnt want the diverter to actuate everytime the ecu computed a background value that just so happened to be a 2s complement divisible by 8.
Yeah, you could theoretically go through and reassign each reference, but by that time your probably better off just using the compatible strategy.
-
GrahamW78
- Posts: 4
- Joined: 2026 Oct 20, 06:11
- Location: Berlin, Germany
- Vehicle Information: 2005 Subaru WRX
Re: Definition for SFI/SD3
not got a clue about eec stuff really, my mgb still runs on a pair of su carbs and setting those up is not rocket science lol, but this thread has got me curious cos the hardware code seems to be the whole ballgame here. is there actually a list somewhere that maps each strategy to its hardware code, or do you just have to find out by flashing it and seeing what happens? can you read the hardware code off the ecu itself or the part number before you go loading anything on it, cos that would definately save alot of faffing about. is that neccessary info even findable for the older 80s boxes or is it all lost to time, anyway just my two pennies worth