Hi Guys,

I have done much tuning in the last year, so i'm a bit rusty.
I am sorting a car for a friend running $5d (Australian Holden).
It runs great, and idles fine in open loop, but hunts in closed loop.

The timing is rock steady, as is the IAC. I can see that it is fuel/MAP related.

Is there a way to force it to idle in open loop?? THe .xdf I am using doesn't have a parameter that will allow it.. Alternatively, are there some prop/integral constants I can tweak so the BLM's don't work so hard in closed loop?

Thanks!
BEn.

 

In order to pinpoint the cause, can you list a few more specs of the engine.

Specifically:
Injector Size
Cam durations and lobe separation angle
O2 sensor type (1-wire, 3-wire, 4-wire)
O2 sensor location (way back at the end of long tube headers, or close up with cast iron manifolds, or something in between).
MAP seen during idle
Desired Idle RPM
Spark advance
IAC counts
Injector pulsewidth

Do you have a commented hac or source code for $5D? You can add your own xdf items if you can find them in the source. I don't think you'll find an "open loop idle" mode or RPM qualifier, but there's probably a few things in closed loop that can be tweaked to get it to idle in closed loop if you want it.

 

I don't know the full specs.. I've just been dragged in to help finish off a 'not so good' mail order tune for a friend of a friend..

The tune is actually pretty good, drives nic, WBo2 is just about spot on etc.. Its just the idle in pk/n that needs work.

Its a 304 cuber, has 50kpa at idle.. apparently has a 280 adv. duration cam. Single wire sensor, it is getting hot and working, as it goes closed loop nicely.

I have tuned $42, $61, $32/b, $6E, $E6, $8d before, but I am not to familar with $5d, and cannot see anything to force O/L idle or dampen the blm/int response..

Any clues?

Ben.

 

Ben,

I'm just about to burn a badly idling car with a $5d binary in it - do you need a light? Mine only has 65kPa at idle and I think it now has a vac leak as well, but it's also had crook lifters and so on. I'm receiving counselling for it now .

I note what you said about idle spark, but some of the $5d F1 maps put too much extra spark in as the MAP drops. I did better with all 10 degrees on mine.

You might also try the kludge where you set the base air screw to give you near zero IAC steps.

Lastly, you could ditch the whole $5d binary and drop a 16183082 in it (or a '165) and use a $12 binary. They have good enough idle code to idle fine even without a VSS.

Let me know if none of this works and you find something that does!

John

PS, I've also found that you need to be careful of the 34 ("low MAP") fault code with a bigger cam. You can adjust the threshold easily enough. If it's ignoring the MAP at idle because of this, you may have trouble, and you won't see the CEL in ALDL mode.

 

Thanks JohnL,

Looks like the $5d code is a bit limited in this area maybe...
I already have the throttle cracked open to get it to idle at all! It now hunts ~600-800rpm when in n/pk. The extra load from the converter helps when in drive and then its not so bad.. Although it will stall most the time when you pop it into drive if you don't feather the throttle at the same time..

I've been told that I can run $12 on the '808 if I put a bigger chip in the memcal (i already put a 29c256 in there) and disable the vats and set the VSS PPK to 1250.. So I may have to do it that way.. I know that I will lose the aldl comms also..

I didn't want to go to the effort of copying all the tables from $5d to $12, but It looks like that is what I'll have to do.. I was hoping for a quick fix

Thanks again,
Ben.

 

I've never experimented with KPNDELAY, but you might try shortening the IAC delay time from N to P.

You could also try increasing the idle speed in drive in F13D, so that it wants to increase speed going into drive.

John

 

Got a $5D hack with a fair bit started (not much C/L though) if anyone wants to continue with it PM me your email and can send it to you. Its not mine, was from diy-efi i am led to believe but dont think it is still there.

Have been playing with $12 code for awhile and have most of the C/L stuff done but the rest is driving me a little PM me if ya wanna take a look and lend a hand

 

Ben
If you raise the "Stoich idle minimum coolant temp threshold" enough it should prevent it going into closed loop.
Have a look for KIDLECLT in your XDF.
It's generally set to 80C, raise it to 160 and it should keep it in open loop.
Adrian

 

Hi Adrian,
I saw that parameter and thought the same, but it didn't do as expected...
Have you used it before and been successful in forcing open loop idle?
Thanks,
Ben.

 

Ben,

I suspect you've already played with the Prop Gain at Idle? Or does the .xdf not have that? That seems like a healthy duration for a 304. Is it possible for you to play with the Idle O2 Constants?

 

Hi Dom,

The xdf doesn't have much idle specific stuff at all..
I'll have to get hold of a hack and see if I can find anything else worth adding to the xdf..

 

Ben
You'll find that every variable that can be calibrated is already in the XDF your using.
Just about every byte from the start to end of the data section is accounted for.
Adrian

 

Have had a quick look in the $5D code and what Adrian suggested should work providing the F57B table does not have a target of 14.7:1 and the AF option flag is to check for stoich before alowing closed loop (B0 of 0xC009).

The closed loop routine checks if the target AFR is stoich, if not then skips all C/L correction. The Closed Loop Flag (0x0038, B7) will be clear if closed loop has been disabled - you can change the ALDL stream to include this byte.

The INT and proportional gain (or step value in $5D) are held at 0x80 (128d) while the closed loop flag is clear.

Have you tried running open loop fuel by increasing the min temp for closed loop and had the surging stop?

 

Thanks UNEEK,

you have highlighted a few things there that will help me! Yes it will run fine when forced to open loop, I just couldn't manage to do it only at idle..
The car is not available to me at the moment, but I will try it as soon as possible.
MANY thanks!
Ben.

 

Ryan, the XDF I'm using has the notation for $C009, "Bit 0 1 = BPW Mod (0 = A/F Mod selected)".

Does this look right?

John

 

Sorry John, no idea where i got that from - that flag does effect C/L but has nothing to do with checking for stoich. Have been working on $12 code heaps lately and think i have got confused as it does have the option to still use closed loop even if the target is not stoich, sorry for the mis-information.

PS may have a some new $12 based code for you to try on the pulsar if you want, have changed the spark output routines to allow for any reference angle between 0 and 90deg, standard it does not support the small ref angles with reluctor dizzies.

 

Uneek
that flag is only used when the HUD is connected, so affects nothing in normal operation.
Adrian

 

yeah thats correct