Wide Band O2 + data logger for 84-89 Carrera

[scarceller]

Few more ideas and feedback needed.

First, I want to restate that I fully understand that in the stock DME the NBo2 sensor does NOTHING for adjusting AFR while in WOT mode. This means that once the WOT switch is closed at the Throttle body it puts the DME into WOT set of datamaps that are typically setup for 13.5AFR but the NBo2 signal is ignored.

However, these WOT maps are not always 100% across the entire RPM range while at WOT. For instance what if we go WOT in 3rd gear at 2000RPM and just let it ride till 6000RPM will we see 13.5AFR across the entire RPM range? Not sure yet on this but I will do a test run very soon and record my AFR vs RPM for this condition.

Now for the idea and questions: what does the DME do if we disconnect the WOT switch? in otherwords what if the switch is disconnected (no WOT signal available to the DME) but we still Full throttle? I assume the DME will just remain using the mid-throttle maps, I will test this but wanted some oppinions.

So if we stay in the mid-throttle maps and we provide a NBo2 signal tuned to 13.5AFR by the WideBand controller we would maybe get 100% spot on AFR at 13.5 across all RPMs at WOT?

The TechEdge 2J1 wideband controller I have has the ability to output 2 NBo2 simulated signals (it has 2 wires) that each can have custom maps. This means I could send a NBo2 simulated signal tuned for 14.7AFR and on the other line one tuned for 13.5AFR, I would use the 14.7 for idle and mid throttle and then the other for WOT. I would then re-wire the WOT switch to a relay that picked the 14.7 while not in WOT and then pick the 13.5 while in WOT. You get the idea.

The idea is to run the DME in closed loop mode at all times (even in WOT mode) with realtime AFR data being feed back to the DME from the wideband controller. One NBo2 simulated signal for idle/mid range (14.7 or so) and another for WOT (13.5 or so). Each signal fully configurable for a AFR of your likeing.

Comments? am I crazy?

[scarceller]

Here are the details of a WideOpenThrottle run I just recorded, the run was done in 2nd gear starting at 2000RPM up through 6000RPM. I just got the car in 2nd gear at 2500RPM then went WOT till I hit 6000RPM on a slight hill/incline.

This first chart shows RPMs vs AirFuelRatio, take note that while in WOT mode the AFR is not 100% rock solid across the RPM range. I ran the AFRatio run 2 times to be sure they where nearly the same.



In this chart I plotted the same RPMs against AirFlow, air flow was measured by recording the AFM barn door position 0-100% 0 being closed and 100 being fully opened.



So I conclude that the WOT maps alone are not spot on, and produce a AFR somewhere from 13 to 14 or so but not a steady value over the RPM range. So is achieving a spot on AFR (of say 13.5) across the WOT RPM range important? Not sure but I think it can be done. Will try few tricks soon.

[ischmitz]

hm, I think you need to get yourself an in-circuit EPROM simulator. That way you can directly change the map points while you drive and observe real-time results with your data logger setup. You could take it a step further and write an auto0-tune algorithm that changes bins towards a desired AF ratio. This is what Megesquirt does...

[Javers]

Wow, i'm amazed the barn door fully opens so early on in the rpm range at wot. I imagine it must make it quite tricky to maintain a decent a/f ratio past 4k rpm without having any real input from the afm. I notice you've got quite a lean peak in the graph that seems to start just as the barndoor flies up to maximum position. I half expected to see the opposite happen, the engine getting overly rich as the door shoots up.

Jon

[scarceller]

My mistake on the AirFlowGraph, I created the map to map the 0-5v signal to a value of 0-100%, the map was bad! I just updated the posting with the graphs to a correct version of the graphs.

Also, if you look at the graph for the AirFuelRatio it has me concerned because the AFR goes above 14 several times in the RPM range while at WOT. Values above 13.5 at WOT are most likely to lean. I thought maybe something was wrong with the run so I ran it 2 times and the graph shows the results of AFR against RPM with AFR shown twice, once for each run. They 2 runs are almost identical.

It is very obvious that for my car the WOT maps are not properly tuned, they may have been good at some point but they are not correct now.

I would love to see that AFR rock solid at 13.2 to 13.5 at WOT across the RPM range. Will work on my idea to run the car in ClosedLoop while in WOT and feedback a NBo2 signal from the Wideband Module that simulates a NBo2 at an AFR of say 13.2. Will keep you posted.

Originally Posted by Javers

Wow, i'm amazed the barn door fully opens so early on in the rpm range at wot. I imagine it must make it quite tricky to maintain a decent a/f ratio past 4k rpm without having any real input from the afm. I notice you've got quite a lean peak in the graph that seems to start just as the barndoor flies up to maximum position. I half expected to see the opposite happen, the engine getting overly rich as the door shoots up.

Jon

[Javers]

Ah that looks much better! A nice linear air flow all the way to the top. And I would say even turning for AFR of 13.5 is still rather conservative, I would reackon you could get more power by dropping to 13.0 or so.

It will be really interesting to see what happens when you disconnect the wot switch and start tuning.


Jon

[ischmitz]

Originally Posted by Javers

Ah that looks much better! A nice linear air flow all the way to the top. And I would say even turning for AFR of 13.5 is still rather conservative, I would reackon you could get more power by dropping to 13.0 or so.

It will be really interesting to see what happens when you disconnect the wot switch and start tuning.


Jon

I would not disconnect the WOT switch. Increase the bins of the WOT fuel trim map to increase the amount of fuel. This should give you better power. Get closer to 12.6 (ideal for maximum power). And make sure your AF meter is calibrated properly....

[scarceller]

I assume "Increase the bins of the WOT fuel trim map" can only be done via re-chipping?

Originally Posted by ischmitz

I would not disconnect the WOT switch. Increase the bins of the WOT fuel trim map to increase the amount of fuel.

[scarceller]

I have read several times, even other posts in this forum state that the AFM Barn Door bottoms out at about 4000RPMs if you are at WOT. However, my recording (of live real-time data) from the AFM signal against the RPMs at WOT shows diffrent here is the graph:



As you can see, my actual recording shows the AFM door still not bottomed out at 6000RPMs. If the door is fully closed the value is 0 and fully opened it's 100. I take the 0-5 volt signal directly from the AFM and record it. The TechEdge Module that collects the data then sends it to the software live and the software is setup to map the 0volts to value=0 and 5volts to value=100. The idea is to think of the 0-100 as a % door is opened.

So, is something wrong with my car? it runs just fine - or - is this idea that the barn door bottoms out at 4000RPM just wrong? I would think that what the graph from my car shows would make most sence since it allows the DME to know what/how much air flow is coming in right up to red-line.

Any comments?

[ischmitz]

yes - that is correct. As long as the WOT switch is not engaged the DME "thinks" you are cruising and it is trying to maintain closed-loop control using the narrow-band O2-sensor input. I doubt you would make serious power. The 2-dimensional maps (load/RPM) go only to something like 60% load. Load is defined by the amount of air going through the barn door.

Typically, 90% of the time (some arguably have lower figures ) you use part throttle during every-day driving. And here the DME tries to maintain a leaner mixture to help emissions and gas-mileage. Figures of 14.5:1 to 16:1 are comon. Once you put the pedal to the metal the system switches to open-loop and goes closer to the ideal-for-power mixture. It's sort of like the accelerator circuit on carbs. You use a lot more fuel but have a lot more fun.....

Ingo

[ischmitz]

Sal,

as far as I can tell the DME does not use the entire range of the barn door since the 2D maps only go to 60% load. Maybe this is because the response becomes more and more unlinear once it is almost fully opened. Others might chime in here and maybe disagree. I will post a typical map for you to see a little bit later.

[scarceller]

I was just questioning the notion that the barn door bottoms out at 4000RPMs? My real-time data shows this is not the case. But what you are saying is the DME does not use the AFM signal at WOT at some given point?

I'm going to test this theory: I want to know if the AFM signal has any effect on fuel mixture at WOT around the 5000-6000RPM range. I have reduced the spring tension on my AFM so that it can go further open than in the prior run. If it is true that the AFM signal has no effect on AFR between 5000RPM - 6000RPM it should have NO effect on the AFR. We shall see. Have the car ready for the run just need the time to road test and record.

Will keep you posted.

Originally Posted by ischmitz

Sal,

as far as I can tell the DME does not use the entire range of the barn door since the 2D maps only go to 60% load. Maybe this is because the response becomes more and more unlinear once it is almost fully opened. Others might chime in here and maybe disagree. I will post a typical map for you to see a little bit later.

[ischmitz]

OK, that will be an interesting data point. I am looking forward to your results.

[Javers]

Might also be worth verifying (if you haven't already) that opening the barn all the way by hand actually reads 5v / 100%. If its not performing as expected might be time to refurbish your afm, which is also a good oppertunity to upgrade it to the later 80's spec, as detailed here: http://www.frwilk.com/motronic/afm.htm As without that extra bridge wire I guess resistance builds up over time and reduces your voltage output.

Jon

[scarceller]

This has already been checked. With the ignition turned on but car not running I push the door fully opened (very slowly) and can see the signal swing nice and smooth with no dead spots from 0-96. My AFM outputs 5.8 volts when fully pushed in. Thanks for thinking of this but has already been checked early on.

Originally Posted by Javers

Might also be worth verifying (if you haven't already) that opening the barn all the way by hand actually reads 5v / 100%. If its not performing as expected might be time to refurbish your afm, which is also a good oppertunity to upgrade it to the later 80's spec, as detailed here: http://www.frwilk.com/motronic/afm.htm As without that extra bridge wire I guess resistance builds up over time and reduces your voltage output.

Jon

[ischmitz]

Here is the promissed example - it is a 14x12 part-throttle fuel trim map from a 964 engine. The bins go to 58% in this map. So once the load goes higher you can pretty much assume that the pedal was put to the metal and that the WOT curve is being used. That is why there is no "need" for load bins beyond 60%.....

The highest resolution is around the 1st half of the load range. That is where most of the action takes place and where the Motronic is trying to make the engine run smooth and at an ecconomic A/F ratio.

[Steve W]

Originally Posted by scarceller

I have read several times, even other posts in this forum state that the AFM Barn Door bottoms out at about 4000RPMs if you are at WOT. However, my recording (of live real-time data) from the AFM signal against the RPMs at WOT shows diffrent here is the graph:

As you can see, my actual recording shows the AFM door still not bottomed out at 6000RPMs. If the door is fully closed the value is 0 and fully opened it's 100. I take the 0-5 volt signal directly from the AFM and record it. The TechEdge Module that collects the data then sends it to the software live and the software is setup to map the 0volts to value=0 and 5volts to value=100. The idea is to think of the 0-100 as a % door is opened.

So, is something wrong with my car? it runs just fine - or - is this idea that the barn door bottoms out at 4000RPM just wrong? I would think that what the graph from my car shows would make most sence since it allows the DME to know what/how much air flow is coming in right up to red-line.

Any comments?

The AFM bottoms out somewhere around 5000 rpm on a 3.2. Here's a screenshot from an LM-1 logging a full throttle run in second, with a shift into third. From the markers you can see that the AFM voltage rises no further after 5200 in this example.

[Steve W]

Originally Posted by scarceller

I would love to see that AFR rock solid at 13.2 to 13.5 at WOT across the RPM range. Will work on my idea to run the car in ClosedLoop while in WOT and feedback a NBo2 signal from the Wideband Module that simulates a NBo2 at an AFR of say 13.2. Will keep you posted.

That would not be a good idea - the WOT switch is not only used to reference the full throttle fuel maps, but also the full throttle ignition maps for full power. Disabling the WOT switch will leave the DME in part throttle mode with a lower resolution, lower output ignition table.

[ianc]

Steve,

If you're reading this, what do you think would be the net effect of rotating the WOT sensor somewhat counter clockwise, by some small amount? To advance the moment when the WOT maps are referenced. Seems like your desired conditions for maximum acceleration would kick in sooner, in terms of pedal travel. Will the WOT maps extend to 4500 - 5000 RPM?

I set the switch a little early, thinking this would benefit me, since I extremely seldomly am at WOT.

What is your opinion Steve? Or anybody else?

ianc

[Steve W]

Ian, I'm not exactly sure what you are asking, but what you propose I don't think makes any difference. The WOT maps are referenced anytime the switch is activated, and has no rpm limitations - it's active from idle all the way to the rev limit.