DIY PROMDo It Yourself PROM chip burning help. No PROM begging. No PROMs for sale. No commercial exchange. Not a referral service.
Welcome to ThirdGen.org!
Welcome to ThirdGen.org.
You are currently viewing our forum as a guest, which gives you limited access to view most discussions and access our other features. By joining our community, at no cost, you will have access to post topics, communicate privately with other members (PM), respond to polls, upload content and access many other special features. Registration is free, fast and simple, join the ThirdGen.org community today!
P730 - Octane Retard Routine and how to control it better
This routine is currently only applicable to the SD TPI system (8D). A similar routine has not been found in the 89 MAP TPI (6E) BIN. As for existing in the TBI systems or earlier 32 or 32B MAF systems is not known. If you have any of the other systems, you should review the hac for them and report whether such a routine exists in them.
The Octane Retard routine was designed by GM as a safety factor in case you got some “low octane” gasoline. When knock is detected over the “threshold”, it will lower the Maximum Spark Advance you can have and thus your “Total Spark Advance” display on your scan tool will mysteriously become less (4* stock). The routine is capable of “recovering” this “stolen spark advance”, but the way GM set it up, you basically will never get it back until you turn off your engine and restart the motor.
You basically have two alternatives, either modify the parameters to allow the Octane Retard routine come in only when higher levels of Knock Retard is detected, decrease the amount of retard it puts in as each occurrence of knock retard is detected, lower the maximum retard level and alter the speed it recovers at.
After much experimentation, I found it best to just disable this whole silly routine. Now if you live in an area where the grade of the gas is questionable (like taking a trip to some country south of the United States which will remain nameless), you may very well wish to invoke this routine and alter the way it functions. But if you are confident in the quality of gasoline in the area you live, I believe disabling it makes tuning far easier.
Without disabling it, I found that when I checked my “Total Spark Advance”, “Relative Spark Advance” and reconciled it to the total Spark Advance I SHOULD be getting; I would lose spark advance that made it very difficult to tune for Wide Open Throttle.
Play with the various Constants that I define below, but also try disabling it and see which you prefer. The following are the various parameters that affect the Low Octane Retard routine (and how to disable them or things to try with them).
1) Low Octane – MAP Threshold (address x’8229). The MAP value that this field is set to is the minimum that is necessary to enable “recovery”. If the MAP sensor is above this threshold and no knock is being detected, the recovery routine will begin and attempt to give you the retard it took away. Conversley, if you have a load higher than this threshold and knock is detected above another threshold (described below), it will begin lowering your Total Spark Advance.
2) Low Octane – Knock Retard Threshold (address x’822B). This is the level of knock retard detected that will initiate the retard routine. If you wish to slow down the rate of retard or disable it, set it to a higher value.
3) Low Octane – Knock Retard Increase Amt (address x’822C). This is the recovery rate of the knock retard. The larger the number the quicker the recovery. For this routine to occur, you MAP Threshold has to be over the amount set and no knock retard must be occurring.
4) Low Octane – Knock Retard Decrease Amt (address x’822D). This is the rate at which knock retard is pulled out of you spark advance. Lowering the value will make the rate of retard slower. To disable it, set it to 0.
5) Low Octane – Retard Amount (address x’822E). This is the amount of retard pulled from the system as each occurrence of knock retard happens over the Threshold Amount. Set this to 0 if you wish to disable it.
6) Max Spard Advance Retard (address x’8028). This is the maximum knock retard allowed for the Low Octane routine. It is a Double Byte field and is a negative field. (ie, x’FFFF is –1, x’FFFE is –2 and so forth). To disable this routine, I found setting it to x’FFFF allowed the ecm to only retard it the least amount possible.
So that is the Low Octane Retard parameters and how they work. If you disable it, you will find that your “Total Spark Advance” will always remain constant (or only be affected as much as your wish) and should you wish to retain it, have the ecm recover it quicker. I found a moderate KPA of 54 kpas for my MAP Threshold with a lower Knock Retard Decrease and larger Knock Retard Increase allowed it to at least recover it fairly quickly. But I still prefer to just disable it, unless the quality of gas where I live are I am travelling to is questionable. Then, I would just control it better.
[This message has been edited by Glenn91L98GTA (edited April 16, 2001).]
------------------ TRAXION's 1990 IROC-Z
Best Time = 12.244 @ 112.51mph (1.778 60' / 7.819@88.32mph in the 1/8) All Natural. No Force. No Drugs. Stock Bottom End. Stock Body Panels.
Gunning for NA 11's with bigger cam, bigger stall, and bigger exhaust.
-=ICON Motorsports=- Moderator: PROM board at thirdgen.org
This isn't even in TunerCat's next version. I just sent TC the info and hopefully they will include it as part of their next release or the one after. Between Tim and myself, I think we've been keeping them pretty busy. Hopefully, TC will consider my idea of a pure Metric and pure English version of TunerCat too. I just informed them about the conversion factor to make Fahrenheit display properly.
I use TC's TDF Editor to define these. Since you get a 30 day trial period, why not give it a whirl? I love TDF Editor. That package is worth it's weight in gold.
Alright glenn, i am severly cornfused now. you said @ the beginning that the timing that was taken out was not put back in until the car was turned off & then back on again, BUT you are talking about retard increase amt & decrease amount like this is a fluid live action. as i was reading this I thought that this was something that would retard the timing based on some input (maybe knock counts or something & it would not reset until the car was turned off & then back on again.
but then as i read on this sounds like ESC attach rate & ESC recovery rate & the "normal" knock sensor retard routine. the one that listens to the Knock sensor & pulls or adds timeing depending on the knock counts.
So, i guess the question is, are we talking about the same thing or is this something that is on top of the knock routine? I am using GME Pro, & these are available things & think that they must be also in Tuner cat & that this is something that is on top of the knock retard routine.
BW
------------------
Bob Wooten
71 Camaro, 91 EFI
r71chevy@earthlink.net
www.r71camaro.homestead.com
Bob, these are not part of the normal "knock retard". These are in addition to, the regular "knock retard" routine.
Yes, you CAN modify the fields I have identified to make the "Octane Retard" routine to recover. But under its stock settings, it does not recover except at an extremely slow rate - such that you don't really recover this retard except by turning the engine off (or modifying the identified parameters).
This routine works by reducing your "Total Spark Advance/Relative Spark Advance" and unless you reconcile these fields, you will never notice it.
Ohhh I see. so this is an active thing it is just that is moves so slow that you dont see it happening. turning off the car & then back on again resets it so you see an instant change & not a slow moving firmware filter, & tuning in the atack/something else rates to get decrease the filter length. This makes more sense, got it.
However,I am not sure that i am following you on the total spark advance/relative spark adavnce thing. when i look @ a data master run i get spark advance, spark retard & knock count. I am guessing that spark advance is the same as total spark advance but what is relative spark advance the same as?
thanks Amigo,
BW
------------------
Bob Wooten
71 Camaro, 91 EFI
r71chevy@earthlink.net
www.r71camaro.homestead.com
Without any occurance of Knock Retard (the normal one), Total Spark Advance = Relative Spark Advance + Base Timing. When you have Knock Retard occuring (the normal one displayed on your scan tool), Total Spark Advance - Knock Retard = Relative Spark Advance + Base.
If you had 40* Total, 0* Knock, 34* Relative and 6* Base, it would be 40 = 34 + 6. Now if you had 6* Knock Retard it would be 40 - 6 = 28 + 6. Your Effective Total Spark Advance (not displayed, must be calculated) would have retarded 6* to 34*.
If (on top of this) Octane Retard also occurred, and it pulled 4* away, the numbers would look like this 36 - 6 = 24 + 6, you would only have 30* Effective Total Spark Advance. The 4* would just disappear from the Total AND Relative and not show up in the Knock Retard field at all.
Assume that you were previously getting 40* Total Spark Advance (and 34* Relative) displayed as your MAXIMUM values while cruising on the highway (while no Knock Retard was occurring), after the Octane Retard Routine kicked in and did its dirty work, your MAXIMUM Total Spark Advance you would only reach would be 36* (with only 30* Relative). Without modifying the constants for Octane Retard, you would remain at 36* as your maximum Total Spark Advance for a VERY LONG TIME until you turned turned off your engine and restarted it.
This is how I first observe this, when I was reconciling my MAXIMUM Total Spark Advance (to my various Spark Tables) and where I was previously getting 42*, I was unable to get more than 38*. After further investigation (and reviewing various Diacom data captures), I noticed that once this occurred, it happened for ALL ranges of the engine operation (WOT, part throttle, all rpm, etc)...they all became 4* less and never returned (except after a VERY LONG TIME).
[This message has been edited by Glenn91L98GTA (edited April 19, 2001).]
I am good with teh first sentence but the second one does not jive. Total = Relative + Base (Cool). Total - Knck rtd = Relative + Base (Not cool).
maybe i am not understanding what "total spark advance" is. my understanding is that total is just that the total number of degrees of rotation that the the spark occurs.
if that is the case then I think that it ought to be total = Relative + Base - Knck Rtd. if that were the case the it would be total = 34 + 6 - 6 (for 6 Deg of retard).
and if you are also getting octane retard you are going to have Total = 34 + 6 - 6 - 4. which would only be 28 Deg which would make HP take a huge hit in ths shorts.
I can see where when looking @ the maps & the scan tool you would have to do some math to finger out what is going on with them timing.
I think that when you are looking @ the timing on the data stream you are seeing the true timing regardless of how much is being pulled out, although what you are not seeing is what the table is asking for AND how much is being pulled out via the low octane retard, but you are seeing the knock retard. so when you look @ what the table is asking for & how much knock retard you have compared to what the actual timing is, is infact where the discrepance is coming from, & what lead you to the ocatane retard issue.
let me know if i am WAY out to lunch here, if not I get it & i totaly see how this would screw up the tuning, 'specially if you are tuning WOT cause it could change every time you went back into the garage & burn a chip.
Tanx again.
BW
------------------
Bob Wooten
71 Camaro, 91 EFI
r71chevy@earthlink.net
www.r71camaro.homestead.com
Duuuuuuuuuuuuuuuh, i swear i passed 3rd grade. guess i am going to have to break out the calculator next time, that higher level math, just KILLS me every time.
than you for pointin that out.
Bob (going back to 1st grade with his kid) Wooten
------------------
Bob Wooten
71 Camaro, 91 EFI
r71chevy@earthlink.net
www.r71camaro.homestead.com
Bob, not so hard on yourself. I bet you just had a bad hair day. Are you sure you didn't leave the equation on the table and one of your kids came by and finished it for you? Hope you have fun making new friends at school, also be very carefull when crossing the road. Oh ya, don't forget to pack your lunch, school food stinks.
Bob, the main thing I am discussing when we get into the math is that I am trying to show you how to calculate your Total EFFECTIVE Spark Advance.
In all cases, the equation is Effective Total Spark Advace = Total Spark Advance (displayed) - Knock Retard = Relative Spark Advance + Base Timing. (Assuming your Base Timing at your distributor is the same as your Base Timing defined in your Constants).
What I have discovered is that the 2 different types of retard are displayed differently and must be calculated differently. The "regular" Knock Retard that is displayed on your scan tool must be deducted from your Total Spark Advance displayed on your scan tool to get your true Effective Total Spark Advance. The Low Octane Retard (which is not displayed) is already accounted for in your Total Spark Advance.
I have found the "Relative" Spark Advance (which is what the ECM really is adding to your ignition system) is the only "true" spark advance that can be relied upon.
But it is a good exercise in calculating what your Total Effective Spark Advance should be and actually reconciling it. Just remember that if you think you should be getting 40*, you are NOT having any knock retard occurring but you DID have a knock incident that more than likely the Low Octane Retard routine has probably "done it's thing" and will account for why you probably are not getting 40* when you should.
Thank you gents. I will check both directions before i head off across the street & remember to not run with sizzors @ craft time.
I think that i get it, I am going to whip out the Data master run & see if i cant follow along & see what i come up with. gotta do it or it does not sink in (hard headed right brained male & all).
what I find interesting is. what is the aldl data stream for? diagnostics right? if that is the case then why does the general not put the actual total spark advance as the one that is displayed? & then have the different retards shown as well (or knock retard & a bit flag for the low octane retard). maybe one of those bit flags that is not defined is for that. This seems kind of silly to me. if someone is trying to figure out why the timing on a car is so out to lunch you would think that the firmware writer would want to give the Tech a heads up?
In any case, gotta go chew on this for a an hour (or 10) & let it digest
thanks again.
BW
------------------
Bob Wooten
71 Camaro, 91 EFI
r71chevy@earthlink.net
www.r71camaro.homestead.com
Bob, I believe the reason the "Genera" did it the way did is because the "Total Spark Advance" is used as a control within the eprom for a "maximum". It reflects a "cap" of the "maximum allowable". Whereas the "knock retard" (not the Low Octane Retard) is a temporary retard and has no effect on the "Maximum Allowable". I believe it is more of a terminology problem than anything else. The REAL thing to watch is Relative Spark Advance and add your base to get the Effective.
Guess you've learned more about knock retard than you really wanted to. Just think of the "Total Spark Advance" as really meaning "Maximum Allowable" and life becomes simpler.
Here is another post that guys with SD should be aware of. I have been discussing this subject with some of the "gurus" and it appears that GM used a similar routine on other vehicles BUT it is more of an "Octane Advance" routine instead of an "Octane Retard" routine as used in the $8D.
I am looking at ways to modify this routine to work in "reverse" to make it add extra spark if the ECM can safely do it, rather than "retard" it.
Just wanting to know, I'm new to this and been trying to read every thing I can. I have been playing around with winbin using the aujp.bin and in the values tab it has six different enters for the low octane knock retard and in the arrays tab It has two. One for low octane knock muti vs rpm and one for low octane knock multi vs map are these the routines you are talking about.
For the person that e-mailed me...here is the info on Octane Retard on the 7730.
It has been suggested that with a little fiddling, you could modify the routine to be an "Octane Advance" routine that would see if you can handle additional timing.
I have seen very similar activity from my LT1 8051PCM $EE mask. Once knock has been detected, The PCM just starts is own little plan of removing timing in the amount and time of its choosing. I haven't yet determined exactly how to get rid of it yet but its gotta go. It's driving me crazy. Your post was very helpfull to me and confirms my observations about a different operating mode.
Thanks
Steve
__________________ Project Blown 455 Olds is underway!
There were a number of other ECMs after the 7730 that also have a similar routine. The Sy/Ty is suppose to have a similar routine, but it works more as for "Advancing" instead of "Retarding". Given the $EE is a more sophisticated routine, I would expect it to have something like the Octane Retard Routine (but more sophisticated).
IMO the ultimate solution is to find out what the source of the knock is and eliminate it. But it is sometimes not possible to eliminate knock at all times in all circumstances.
I suggest to not outright disable the Octane Retard Routine (unless you never plan to drive away from your local area and sure of the quality of the gas you burn). I would made it "recover" in a more reasonable period of time. Ori the 7730, the recovery time never occurs in the stock setup until you turn off the ignition. It is one of the culprits that affect the way the engine feels when "hot restarted" on a hot day (doggy).
Originally posted by 87_TA Does knock in PE mode have any influence on this what so ever ?or just under normal closed loop operation.
For the 7730, all knock of any significant magnitude seemed to affect it. There is a "threshold" value, but it didn't seem to work as expected. I found that any knock exceeding a couple degrees would cause it to engage.
So best way to disable it is to set octane retard amount to 0?
Also does setting the pe knock retard to a set number only allow it to retard that amount actually?
Or does it just fool the scan tool?
__________________ 93 Firebird, 7730 ECM, most other 3rd gen running gear, 406, Ported AFRs, T56, 10.41 @ 132 N/A, 9.4 W/200 Shot 93 oct
I'm still confused......
So Glenn, does that mean to disable it you would:
1. set the Low Octane knock retard mult. vs map to "0" ?
2. set the Low Octane knock retard mult. vs rpm to "0" ?
Thanks for the help,
Don>>>
Originally posted by Dyno Don I'm still confused......
So Glenn, does that mean to disable it you would:
1. set the Low Octane knock retard mult. vs map to "0" ?
2. set the Low Octane knock retard mult. vs rpm to "0" ?
Thanks for the help,
Don>>>
I never found a need to manipulate those two tables. I was able to control the Low Octane Routine solely through Constants. I will include the name of the constants, it's hex address location and the value I set them to to disable it. But do feel free to use the Low Octane Multiplier Tables and report on how well they work. I found I just prefered to disable it entirely as I drive my car only in a local area and only use the "good stuff" (Chevron 94 Octane is the best available in my area and very consistent)....I don't use "No Name Brand" or less than 94 Octane.
These were the settings that minimized the Low Octane Knock Routine for me, and gave the results I was seeking (I wanted it disabled). As I said, feel free to do more experimentation and report on the results. But make sure you read my original post for the explanation of what they do. It's been almost 2 years that I worked with this.
In fact, these values didn't even exist in TunerCat's $8D tdf at the time. I had to add them to my own "special tdf" that I created myself. I did report these to TunerCat and they have added these constants and those two tables after I explained them to TC.
I guess I should get the most current $8D tdf, but I am doing my "own thing" and working on other things. That is why I listed that actual hex address that the constant is located so you can make sure you are working with the right constant. Unfortunately, my version of $8D is quite unique and it is too much work for me to add the items I am working with at this time.
Last edited by Grim Reaper; 03-31-2003 at 04:19 PM.
