This is from the ASDX bin for 7747. I can only guess so much because they are so abbrviated. I guess that 0x2FC is some kind of multiplier for ??? to correct ??? at idle. Yeah that's real vague, but I'm curious as why they would be different from auto to manual trannys.

...still trying to figure out my closed loop idle surge. I have WinALDL set to display the injector PW. It's unconverted, but I've gotten used to where it idles. In open loop the unconverted number is ~13000. Closed loop it's around ~16000 but then it starts to creep up and eventually it'll spike to ~22000 and that's when the surge happens

 

Capt'n, you do ask some good questions. Lets see, the term at LD2FB reduces the O2 error when rich. This is required because the O2 sensor is asymmetric in its response. No big deal.

The other term, LD2FC, is used to reduce the magnitude of the O2 error whenever in idle. I must say that the value above is quite small. The smaller the value the greater the reduction in O2 error.

So what happens is that instead of the ECM acting upon an O2 error of say 200mV's, it is first multiplied by the Idle Corr Term of 32 and divided by 256, then used as the error term ( (200 * 32) / 256 = 25). Quite a reduction in error.

The end effect is a greater variance in the allowed O2 value while in idle.

RBob.

Have a DIY-WB yet?

 

Hehehe, if you think his questions are good then my question is real bad. What the bloody A is an o2 error?

 

An o2 error can either through an error code or change the BLM/INT when in closed loop. If the o2 sensor is below 200mv for x seconds an error code will be sent. If an o2 error could stand for a boundry for moving the BLM or INT, it's really hard to tell. You'd have to read more into the code to figure it all out.

 

Hectorsn, from what I understand the O2 error is the difference (or percentage I'm not sure) of how far off the O2 sensor is from the ECM wanted.

Say the ECM wanted 14.7 AFR, and O2 sensor saw 14.3 AFR...that would be ~3% error. The ECM then does some voodoo science and it decides how much to move the INT based on this.

I hope I'm not misleading...(i.e. please correct me if I'm wrong).

Thanks RBob for the response (and for being a good "coach"), and J for your input.

I haven't built the wide band yet. I found a gem at the junkyard tho. It's the WB from some year Civic. The dummies that took the exhaust manifold off just tossed it aside. It looks like it's seen better days, but I'm gonna give it shot before I buy a new one. I'm still reading up on all the WB O2 info. Call me paranoid but I'm uncomfortable with sending someone I don't know money in hopes that they will send me something back.

Awesome. In the morning I will try raising the value at 0x2FB and see what happens.

 

hey let me know how that works out. also i will have a bin for you 2weeks form now for the accord project.

 

I had to haul some stuff this weekend so I used my old open loop idle bin.

I never noticed how stabil the injector PW is in open loop. It never came out of the 15000 area. Mostly 15120 and 15400. That's it. In closed loop it would be anywhere from 13000-22000.

Now I'm convinced it's a tuning issue....

I didn't the chance to try adjusting the 0x2FB parameter yet but I'll post back my results.

 

try raising yr idle a bit in closed loop. could be exhuast backwash pulloting the o2 readings. ive seen this in numerous vehicles with large diamtere short pipe systems. raising the idle a bit keeps the exhaust flow heaed int eh right direction. if the ve table is correct of course.

 

Note: 0x2FC is the one to change.

One reason for the varience in PW while in closed loop are the proportional gains. They are adding and subtracting from the BPW in accordance to O2 sensor feedback. Reducing these values tends to stabilize things.

It can be difficult to tune once the O2 sensor is moved further from the engine along with cam/heads/etc that make the engine more responsive. Sometimes drastic measures are required.

RBob.

 

Oh, hehehe, thanks.

Funny you should mention the the distance of the O2 sensor from the engine. My O2 sensor (3-wire) is about 2.5 feet down the drivers side pipe, if you measure from the #7 exhaust port.

If I understand your definition of "responsive" then I would say the roller cam (although small) and roller rockers make it very responsive...as well as the 2100 stall converter and the gears.

:hail:

 

FYI, I have the 2FC bit set at 1, the lowest it can be, besides "0".

I also have the idle Prop Gain factor in idle (supposed to be .0547, or bit = 14) to bit = 1, lowest besides 0.

This has helped my idle quite a bit, BUT, it still causes issues. What happens is that if your VE tables are not close to perfect, and, say you are driving at MAP = 40 and 1500rpm and have a high blm = 145, when you drop to idle, you will run rich for quite a bit as it takes a long while for the now slowed idle correction to take place to get to stoich. Vice versa on the lean side, takes a while to get to rich from lean. Slight hunting when lean but overall, worth the change.

I still am having difficulty getting "perfect" idle control with 7747 MPFI control. But surely livable and never stalls, now, with the above done. I guess I could/should get the idle VE block VE tables absolutely perfect ( a few cells could probably use a bit of change, but rarely operates in those cells) to help more and perhaps set one or both of the above values to 0. Anyone set those values to zero yet? What happens???

I also raised the R, R/L, L thresholds to straddle the .451mV O2 switch point, but I did that quite a bit ago. That helped drivability a lot.


Overall, I think I can get 425hp at the crank with this 7747 ecu setup. Not shabby. You gotta hear the Wrangler with the 2.5" true duals, o cats and Flowmaster 2-chambers. NICE at 5 grand...