11sORbust

Could someone explain the function of the o2 constants? Which one would be used to "help" command different a/f ratios(in closed loop)?

11sORbust

730 ecm....

11sORbust

Well...... Once I figure everything out I'll share how to command different a/f ratios in closed loop...

Grumpy

The O2 constants are the voltage levels the ecm has to cross when swinging for the PID to work.

The code goes for averaging an AFR, not running AT a specific AFR.

For the engine to run AT a given AFR, means running Open Loop.

It was a delicate balance of changing commanded AFR and O2 voltages to truely arrive at a different commanded AFR.

11sORbust

.450mv is the point that the ecm uses to get the average a/f ratio of 14.7:1,right? So how does it work? Could you share a snipet of code?? Maybe we should take this off list, don't want someone to make money off this one

TRAXION

No. There is nothing in the code about 450mV. Everything I see is about 0.600mV at idle. Off-Idle is a different story and the swing points change with MAP.

Tim

jeepguy553

Interesting. Couldn't you specify a different AFR to run in closed loop? Seems like there would be a way to tell the ECM that 14.7:1 is great...UNLESS a certain condition (or set of conditions) is satisfied. Would that involve a custom table of sorts? Sounds like a cool way to go with this idea.

11sORbust

So....is that a mv vs map table(of some sort)? I thought there was just o2 constants??

11sORbust

you say "swing points", not swing point. So does that mean the ecm is using two mv points to average a/f ratio?

Is this subject ok to discuss? I don't want to "go there" if this infomation is being held to protect the board from "certain types".....

Grumpy

The problem is the sensor.
The Skew rate is almost vertical and trying to tell 15.5:1 on it is very diffficult. Try watching a WB and a stock O2 and you can see where the problem is.

Now if you want to use a WB then you can pick and chose any AFR easily. If you go to the PLX Devices Site, you can see what voltages you'd want to get a specific AFR. Then just build a Comparator to toggle at those WB output voltages, and feed that to the stock ecm, and code. I'll let the EE hash out the particulars.

11sORbust

I was planning on using a wb in place of the stock O2. That way commanded a/f ratios will be acurate....

I think that I figured out how to keep the integrator from "pulling" the blm in line with 14.x:1 a/f ratio. My car will be running soon. So I'll test everything out. I should be able to command any AVERAGED a/f ratio in closed loop. I can give you guys a update, if you like.....

TRAXION

You can't command any AFR because AFR has nothing to do with INTs and BLMs. You can only command a different O2 swing point.

I have my car out today. I am currently running with O2 constants that are 100mV lower than stock. I chose lower than stock due to the fact that all of my overlap is allowing fresh O2 to come past the O2 sensor at idle. Thus, the O2 sensor will read leaner than it really should be reading. I figured that since the ECM does this 100mV for AIR then I'll be safe doing it too as a starting point. Car is running great. The off idle O2 constants in the table of Fast O2 vs. MAP are still stock (which I don't think need to be changed). I also finally disabled AIR and did some other tweaking of the hardcode (removed ability for ECM to use lean BLMs at WOT, removed ability for ECM to use open loop AFRs at WOT, etc)

Tim

11sORbust

could you explain??Cause I thought the entire ecm was "designed" around the idea of running stoichiometric,generally....

Grumpy

How is more overlap allowing more O2 past the sensor?. At overlap the exhuast side pressure is higher then the intake side, and causes a more self EGR'ing.

Yes, dropping the values will often seem to improve things, but that's also changing the rich to lean ratio of times spend at either side of Stoich.

Just as an alternate strategy, some of the GN guys patch to a fixed BL of 128 at a given % TPS.

TRAXION

As per all the posts I have made recently I have showed how the stoichiometric AFR constant has nothing to do with making a car run at that AFR. IF you want the car to run at a specific AFR then you can either run open or you can change the O2 constants / tables.

Tim

TRAXION

I rethought through this process and think I see where you are coming from. Thanks! I guess I was just thinking about how stinky my exhaust is. My previous thoughts about the fresh intake charge sneaking past the O2 sensor would sort of explain this. But, given what you said ... now I can't explain it. My exhaust is way stinky ... not a 'fuel' stinky. It doesn't smell like fuel. It smells like some type of byproduct? Not sure. It very easily will permeate your clothes and make them REALLY nasty. I was hoping that toying with the O2 constants and thus changing the average true AFR at idle would make a change in this stinkiness. I am not running converters and I use pump gas ... 93 octane. The car idles great and runs awesome. Just stinky at idle and I'm getting old so things like that bother me ... specially when I show up for work and my clothes stink of nasty exhaust fumes.

Tim

RBob

iTrader: (1)

This stinky fuel seems to be a regional problem. Where I am it is the same way, it stinks! I hate being behind vehicles that have just been started, I hate being behind vehicles that have been in PE mode to get their vehicle up to speed. Their exhausts stink.

From what I can tell if the cat is up to temperature the problem goes away. Once a little 4 cylinder vehicle tries to hold speed on a hill, forget it. Gas mask time.

This past Fall I have been in other areas where this 'stinky' exhaust wasn't a problem. In my area it is disgustingly bad. To the point where on an upward grade people are racing each other to stay in front.

This past summer doing idle tuning with the car in the driveway was bad. Afterwards all my clothes went right into the washing machine while I took a shower.

Even my daily driver as I start it to go to work. Unless I am in the car with the doors shut I don't even start it. Used to be when it snowed folks would start the car and then brush the snow off the windows. No way, not any more.

So, I'm with you Trax, I tune my idle as lean as I can so I can stand to be in the same room as myself. I recently posted on another board that I tune my idle by smell. This is why.

Again, from what I can tell and know about fuel mixes, this is a regional problem.

RBob.

TRAXION

Wow. Thank you. I was feeling sissy-fied complaining about how the exhaust stinks. I feel better now that I know that somebody else hates it too. I TOTALLY understand the "right into the washing machine". That's exactly what happens when I am in those fumes for awhile. Right now I tend to make sure that I am in the car before I start it. It is rediculous how well and how quick this stuff permeates your clothes. It's unbelievable. Now that you mention it ... my Subaru has a similar smell on first startup! It's definitely not as bad ... but it's definitely stinky on first statup. Thanks RBob - I feel better knowing that others hate this too For the heck of it I am going to try toying with the O2 constants to see how lean I can get the idle and how well I can get the stinkyness to go away. If you have any luck please let me know.

Tim

11sORbust

if you had a catalytic converter there would be alot less smell. I noticed "the smell" when I bought the bastard t/a. It doesn't have a converter and smells very,very bad . My daughter gets a headache when riding in the car. No exhaust leaks, just alot of city driving.

I agree that the gas is a big factor . But it doesn't seem regional to me(I mean in my area). Sice the t/a doesn't have a cat I can smell the exhaust ALOT more. I have noticed that certain gas stations have different additives and general quality. In turn some gas stations have stinky(er) gas. That's just my neck of the woods....


I'm not trying to command a certain a/f ratio. Maybe I worded that wrong. I want to control the a/f ratio at idle and part throttle. I'm going to keep the iac and learning ability. Just because I have a stout engine doesn't mean I have(or want) to run open loop.

I still think that the ecm is based on 14.7:1, generally. Everything I know about the ecm supports that. For example, the nb o2 is designed to be very acurate ONLY at .450 mv. All my text books tell me that .450mv= 14.7:1. Closed loop uses the o2 as one of the main sensors. Please correct me

TRAXION

I would word it like this ... someone please correct me if I am wrong ...

The ECM isn't based on 14.7:1. GM chose 14.7:1 due to emissions and the ability of that AFR to work well with catalytic converters. Thus, GM incorporated 14.7:1 into the ECM by using BOTH a stoichiometric constant and O2 constants/tables. Fuel Calculations are based off of the stoichiometric constant. HOWEVER, BLMs and INTs are based off of the O2 constants/tables. If you were running just open loop then the stoichiometric constant would satisfy everything. But, once closed loop takes control then the resulting BPW (as derived using the stoichiometric constant) is modified according to the INT and BLMs through the use of the O2 constants and tables. You constantly mention 450mV. Well, I keep saying that is not what the ECM is using most of the time. Here is what it is using ... see attached pic.

Tim

11sORbust

Well.. it looks like GM put more stock in the NB o2. I was thought that the stock o2 was very acurate at .450mv . Then it lose resolution very fast on either side of .450. That snipet of code has many different mv reading. Could someone correct me if I'm wrong. Maybe .450mv is not 14.7:1????

JP84Z430HP

Is it that 450 mv is just the theoretical 14.7:1?

I agree in that everything I have been taught, in classes, books, etc.... says that anything over 450 mv is considered rich, and anything less is lean. Now, it's looking like the ECM is doing some stuff that's quite a bit different.

My understanding is also that the ECM will change the pulse width slightly to get the O2 mv to cycle just above, to just below 450 mv, in order to make the catalyst happy.

Now, with that being said, I'm confused about this 600 mv stuff, and now I have something else I need to look at as I learn the code........

As for the "Stinky Gas", I could live with it if it was the same gas that I got my best fuel mileage from. In my daily driver, I was getting 24-25 MPG about a month ago, for a couple weeks, then it dropped to about 21.5-23. Not a big drop, but enough. I think I was getting higher numbers at first, because the A/C wasn't being used hardly any, and now, we're running on the winter blend gasoline. (This is my theory and I'm stickin to it until someone tells me different!!)

11sORbust

I have the latest tdf and do not see o2 mv threshold vs map table. The "super8d" ecu doesn't have it. The only thing I see is a few o2 constants.....

TRAXION

I have the latest TDF and it has the appropriate constants ...

Fast O2 Rich/Lean Threshold at Idle
Fast O2 Rich/Lean Hysteresis at Idle
Upper Zero Error Ref Slow O2 R/L at Idle
Lower Zero Error Ref Slow O2 R/L at Idle

The table isn't in there yet. I haven't had a need to modify any values in the table yet. Just idle values. Thus, I haven't added it to my TDF nor have I asked TC to add it to the standard TDF. Of course, you could always to ask him to add it and he would.

Tim

11sORbust

I would imagine that's the correct/only way to really command different part throttle a/f ratios.

Is there a way I can add the table myself? Could somebody help teach me how to do that

Grumpy

.45, or .5 depending on sensor and who you talk to, are STOICH..
Lemme say that again, STOICH.

The AFR for different fuels are different for STOICH.
With gasoline, it's ASSUMED, that 14.7 means STOICH.
So if you add say 10% Alky, the ecm will correct the AFR to STOICH.

Setting the min and max O2 volts set the values that the ecm will try to swing thru to AVERAGE, STOICH.

If you try to change the commanded AFR to other then STOICH, then remember all your WOT commanded AFRs are also effects.

And if you go playing with the min/max O2 values you may also have to tinker with the code 13-44-45 gualifiers.


There's also been a bit of todo at CARB about what the stock O2 is really sensing, and how other exhaust gases may be effecting their accuracy.

TRAXION

Yes, AFAIK, that is the only real way to change part throttle AFRs (via changing the O2 mV in that table). My part throttle 'feels' fine and the plugs look good so I left the part throttle O2 mV points stock. I did change the idle O2 mV significantly in an attempt to make my exhaust less stinky at idle.

As far as adding the table, if you are using TDF editor then it is pretty straight forward. Just open up the $8D and look at the existing tables and then you'll see how to add this table. If you run into a problem then ask a specific question so that I know what to specifically tell you.

Tim

TRAXION

ANHT and AUJP use 600mv for idle and 530-580 for part throttle cruise depending on the load.

Tim

Grumpy

Your mixing issues.
My numbers were for what the sensor generates at STOICH.

I was talking hardware.

11sORbust

Another unresolved issue. I have been taught that the nb o2's stoich point is .450mv. Trax has shown me in the code the ecm doesn't use .450mv for anything. So why is there a difference between hardware(.450mv) and what the software is using (.5x0-.600mv)? How can the ecm overlook the only mv reading that the o2 is acurate at?

TRAXION

... and there ya have the million dollar question that nobody has answered yet (as far as I know).

Tim

11sORbust

What about a.i.r., you said the ecm has compensation for that(.100mv),right? Could that be the reason or do you think the ecm is "compensating" for something else? Maybe stoich is not .450mv when the engine is operating. Maybe the "stoich mv" is not a constant but varies based on engine conditions. Guess I'm going to start changing those constants/table and log the results(sometime soon). Wish I had an ecm bench......grrrrrrrr!

(poking in the dark......as usual)

91L98Z28

I believe it has been mentioned that a typical NBO2 will have signifigantly varying output dependant upon (for example) EGT. So yes, is certainly possible that the stoich mV could vary depending upon fuel used, EGT, load, etc.

My personal take on a NBo2 is as follows. It's like a spotter in the exhaust stream. S/he has NO CLUE what s/he's doing except whether or not s/he sees oxygen. S/he doens't know how much, or how little oxygen there is; gets paid minimum wage, is sometimes unreliable and has mood swings based on the weather. In fact the only promise you have is that there will be a some kind of yes/no signal as to whether or not oxygen is being seen.

that's the only thing i trust a NBO2 for. and then, only sometimes.

to me, the voltage values being equated to AFR's is kinda like (back to our spotter analogy) saying that because the spotter has his arm REALLY low, there must be more oxygen in the stream (or vice versa). and that if his arm is only a little raised, there is only a little oxygen, but if it's raised really high, that there must be a lot of oxygen.

instead i submit that maybe the spotter is a lazy bum and you'll be lucky if he raises his arm in a timely matter at all, let alone assigning any value to how high (or low) his arm is.

stoich. is a chemistry definition, and for the fuel we run on the street is supposed to be around 14.7:1. it's the point where all the fuel and all the oxygen is consumed, and there is nothing left over (except the byproducts). my (limited) understanding of chemistry says that stoich should always be consistent (for a given quality of fuel and given type of air). just because the engine is spinning really fast, shoudln't change the laws of chemistry, you still need X amount of fuel to burn Y amount of air.

the above is completely ignoring anything to do with enriched AFR's (below 14.7:1) which are used for making power and have very little to do with ensuring that all the fuel and all the air are burned. the goal with an enriched AFR is (given that air takes up so much space, and fuel so little) that we add a tiny bit more fuel (which will displace only a tiny bit of air), in order to ensure that every last bit of oxygen is burned (which overall is a power gain).

example

you have 14.7 grams of air, and 1 gram of fuel
lets say you can make 14.7 hp from 14.7 grams of air (completely bogus number i'm sure)

and lets say that due to inefficiencies/bad mix, you lose 20% of your power (another bogus number)

so we have 11.76 HP, and 2.94g of unburned air. that ~3g of air is worth 3HP, if only we could burn it.

so, if we enrich by 10%, (now using 1.1 gram of fuel), we only have 14.6 grams of air left in the same volume (theoretical max of 14.6hp now), BUT we can burn a lot of that 2.94g of unburned air, and might gain 1 or 2 hp out of it the deal. we'll be spitting raw fuel out the exhaust too, because there is too much fuel now, but we gained power.

hope that makes sense. if it's utter nonsense, please correct me!

finally, the reason for stoich is emissions. if you consume all the fuel and all air air, then there is no reactive materials (i.e. fuel) going out the exhaust, which is supposed to keep the air cleaner.

TRAXION

Hey RBob,

I was curious as to how low in terms of AFR ... or, more specifically with regard to the bin, how low you have gone with the O2 Sensor mV constants? I have used 430mV (which is a huge difference numerically as compared to the stock 597mV). I am in the process of trying 390mV. Going from 597mV to 430mV didn't clean up the smell that much. I am hoping that dipping down into the 300's will. But, maybe the 200's are necessary?

Tim

RBob

iTrader: (1)

I'll take a look tonight, can't remember off hand. There are three terms for the idle O2 values in the $8D mask. Need to tweak all three. Can also change the idle O2 error mutiplier to tighten it up.

Also note that several ANHT hac tables in that area are mis-labled (such as the one at $84B4).

RBob.

11sORbust

What actual a/f ratios have you been working with? MV and tuning by smell has me wondering if you guys are using a wideband? My car doesn't smell anymore. I think it was bad gas that sat in my tank for 6 months. It was also way rich......thanks.

RBob

iTrader: (1)

OK, here is what I've been using:

486mV, mean R/L O2 val for idle
501mV, mean rich O2 val for idle
466mV, mean lean O2 val for idle

Shows about 16.1 AFR on the WB. However, I doubt that is the true AFR. The NB is a GM AFS-74 heated in the area right after the collector. The WB is located similar but in the other pipe.

RBob.

11sORbust

What did the wideband report with a 430mv nb reading?

11sORbust

are you guys using a wideband in place of the stock nb O2? If not then 4XXmv would net different actual a/f ratios based on the weather or how the nb O2 feels,yes? The furthter you go from .450, the more that stock o2 will skew the resulting a/f ratio....

Grumpy

[QUOTE]Originally posted by 11sORbust
Another unresolved issue. I have been taught that the nb o2's stoich point is .450mv. Trax has shown me in the code the ecm doesn't use .450mv for anything. So why is there a difference between hardware(.450mv) and what the software is using (.5x0-.600mv)? How can the ecm overlook the only mv reading that the o2 is acurate at? [/QUOTE

The ecm isn't looking to maintain 14.7, it's looking to average it.
The converter NEEDS periods of rich and lean to work. So they have things set so the ecm has to go so far rich and so far lean, hence the low voltage and high voltage O2 values. If they were looking to maintain exactly 14.7 the the would use a real tight range based on the Stoich voltage where the sensor toggles from reading rich to lean, ie if 14.7 was .45, they might use, .44 and .46, to tighten the range of variance. The are a bunch of different O2s. GM wouldn't inventory dozens of different senors if they weren't EGT, backpressure, sensitive, not to mention maybe to meet those parameters, the output voltage of Stoich might vary.

11sORbust

PURE GOLD! that explains alot to me. The ecm is looking for an average not at xxxmv for stoich.

69 Ghost

iTrader: (1)

All the latest Super 8d ecu will be released very shortly on TunerPros web page and I do have the settings below that you were looking for on this thread but they maybe named a bit differently. Look under the Idle constants. My latest ECU is mapped accordingly:

Fast O2 Rich/Lean Threshold at Idle = Idle O2 Rich/Lean Slow Zero Error Lower
Fast O2 Rich/Lean Hysteresis at Idle = Idle O2 Rich/Lean Slow Zero Error Upper
Upper Zero Error Ref Slow O2 R/L at Idle = Idle O2 Rich/Lean Threshold Fast
Lower Zero Error Ref Slow O2 R/L at Idle = dle O2 Rich/Lean Threshold Fast Hysteresis

BOWTYE8

Great Thread!- I still have to reread it a few more times

I also cant stand the smell. Five minutes on my car idle in garage and I have to shower and change clothes due to the exhaust smell.
Removing plugs shows a decent burn. Not black-a darker tan. I will post a W/B log from a idle.

11sORbust

That's great!!! Could you add the o2 mv threshold VS map table also? I can't wait to start twisting that table, find out how to command a/f ratios vs map

I'm suprised there is no o2mv VS map VS rpm table.....or some sheet. Just map wouldn't seem to be enough. Do you guys think that once you change from cell 4 (for idle) the o2 idle constants still work? My car idles at 70kpa so I wonder if the mv vs map table is screwing with things?

There is always more questions than answers...........

69 Ghost

iTrader: (1)

Guys let me know if you want these tables in Super 8d. Like I said I am getting ready to release the next version with as much or more work than the first. At first glance I see there are a number of tables starting at 4A2 in the ANHT_hac file. Please let me know if you only want 4AB (Fast O2 Threshold Vs. Map) or the whole list and be specific. I will get them in and release this within the next week.

3.8TransAM

Someone care to break down and explain the o2 constants to shed some lite on them?

Goofed with them a few times using the romulator b ut never quite figured them out or which would work most for my benefit

thanks
Jeremy

TRAXION

I did some more research and now take back what I said. Tons of fresh air makes it into the exhaust at idle. Although what Bruce said makes sense when you are just thinking about it. It just doesn't work that way. To further lend evidence as to what I just said just run Performance Trend's Engine Analyzer Pro and use the options to view the engine at a low idle RPM. I used 900rpms (where my car idles) and MOST of the overlap period shows fresh air being sucked right into the exhaust.

Check out the following image. Look at the top which shows the valve events. The overlaping section is the overlap. Fresh air is being pulled into the exhaust starting very early in the overlap period and continues for the rest of the overlap period. The red box was what I added ... and this is important because it shows directional arrows for the direction of port flow.

Tim

SMasterson

Could this be why Innovate's O2 Analog output that is factory programmed to cross at .45mv cause my otherwise great tune to run very poor?

I don't want to highjack this thread but I was very confused by this. Reading thier docs it seems that LM-1 is capable of actually replacing the factory O2 and is supposed to be faster, so I tried it. Not for long though, it was going to require a complete retune to get the BLM's in line. They dropped to 90 almost instantly and when the Cells changed it caused a major surge in the engine. Like I said, I quickly unhooked it and started reading, here first!

If anyone is interested I'm going to reprogram the LM-1 to swing at the average between Trax's above mentioned .53 and .58 and see how it runs.

Thanks for any input or suggestions,

PLANT PROTECTION

iTrader: (6)

I think I need a break down too. On the idle o2 R/L upper threashold and idle o2 R/L lower threashold, the ecm will give fuel until the upper threashold is hit and then pulls until the lower threashold, constantly bouncing back and forth? Also, what does the 'fast' and 'slow' in the titles mean? And what does 'hysteresis' mean in this case, delay, offset, or ?

11sORbust

Good question!!