Closed Loop Injection

[rjwoods77]

Is there really any point to doing this over open loop? Do the judges give more points for it? Other than the possibility of better gas milage is there any other benefits of the system. We are currently planning and open loop speed/density because of the possibility of future forced induction. So when competition and the cost report come around, can I eliminate the o2 sensor, the barometric sensor(motec said to use a map and a baro sensor)since they are only for tuning and baseline pressure?

[rjwoods77]

Is there really any point to doing this over open loop? Do the judges give more points for it? Other than the possibility of better gas milage is there any other benefits of the system. We are currently planning and open loop speed/density because of the possibility of future forced induction. So when competition and the cost report come around, can I eliminate the o2 sensor, the barometric sensor(motec said to use a map and a baro sensor)since they are only for tuning and baseline pressure?

[Erich Ohlde]

You are talking about Closed loop O2 sensor feedback I'm assuming?

[rjwoods77]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by jayhawk_electrical:
You are talking about Closed loop O2 sensor feedback I'm assuming? </div></BLOCKQUOTE>


Yeah unless there is something else that i am missing.

[MyDuster360]

I'm an ASE certified tech in EFI and have been running my 74'Duster with programmable MPI for a couple years now(Holley Commander 950 Pro.)when i started our team last spring, I was amazed that most teams run open loop during the comp. I still haven't heard a clear justification for it. I really don't know what i'd do if I didn't have O2 comp on the Duster. It's that valuable.

So I say Why not use O2 comp? Except for the minimal extra cost, there's no downside. I read better fuel economy as better engine efficency. On the dyno you usually tune the engine exclusively at WOT, but how often is your FSAE car at WOT? Not counting the accel event, I'd say, 10%-20% of the time at most. Let the O2 correct the part throttle and transitional throttle to keep the A/R spot on. Spot on, RARELY means STIOCIOMETRIC.

Your Motec should give you the option to map a O2 comp table realtive to your Fuel map. Map the target A/R suffiectly rich near WOT and lean it out a bit passed stiok near every else on the map(expect idle of course).

Any way, I'm planning to run wide band O2 comp on our rookie car using a PE ecu. I'm going to tune the base map for 5%-10% rich for safty and let the O2 take care of the rest.

This is the plan right now. Again this is our first year. So i might find out real quick, that my old ways of tuning need to be modified for FSAE.

[Erich Ohlde]

O2 feedback is good on cars that are running more of a steady state condition i.e. going down the highway at 70mph at constant throttle. For an autocross where the engine is constantly changing rpms and manifold pressure is changing. O2 sensors can't read fast enough for the closed loop to work properly.

[Garlic]

Don't worry about it. Too many transients. These cars have nothing to do with street cars. You'll gain nothing. Feedback is for cruising or long term steady-state WOT, and these cars see neither.

Without a wideband, you can't run closed loop at anything but stoich, anyway. And you'll always want more or less than that.

If you have the ability to log it, I'd run an O2 just to see what's going on.

Rob, please don't think the judges give points for stuff!!!! They give points for good reasoning and a good explanation of your choices. They could care less if you have something extra if you don't even think you need it.

[MyDuster360]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">the engine is constantly changing rpms and manifold pressure is changing. </div></BLOCKQUOTE>

Erich, you call it racing, I'd call city driving. Times of constantly changing rpms and manifold pressure is where O2 comp does the most good.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">O2 feedback can't work fast enough to work properly. </div></BLOCKQUOTE>

The sensor itself is plenty fast in response. Case in piont: closed loop WOT operation is common in OEM and drag racing. Transport delay adjustments in the ecu and simple common sense sensor positioning, negate any worry of this.

[MyDuster360]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Garlic:
Without a wideband, you can't run closed loop at anything but stoich, anyway. And you'll always want more or less than that. </div></BLOCKQUOTE>

No, you can a narrow o2 above and below stioch by adjusting the switching voltages up or down at appropriated point on the fuel map. Several ecus have that capability, one being my rather low brow C950. The percent of comp needs to changed too since the narrow o2s are much less accurate off stoich. By which I mean change the comp % at WOT for example to +20%/-0%, so no can be fuel taken out, only added. I've been running it for a year now. Works pretty good in lieu of constant wide band o2 comp.

[Charlie]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by MyDuster360:
The sensor itself is plenty fast in response. Case in piont: closed loop WOT operation is common in OEM and drag racing. </div></BLOCKQUOTE>

Proof is not in use of some other application, proof is 'the sensor has XXX response time, the ECU has XXX response time, the delay in fueling from injector is XXX, and the end result is aceptable because _______". I don't think you've proven anmything.

OEM using WOT closed loop? I'd have to see proof of that. Drag racing using WOT closed loop? Sure, why not. 8-10 seconds is plenty of time to have feedback corrections take place.

FSAE cars have crazy RPM acceleration numbers. Way faster than anything else, really. Much faster than anything geared to run quick in a 1/4 mile, or a real road course.

Come up with a good arguement with some numbers and I'll listen, but you can't just say 'I'm ASE certified in EFI (not sure what that means, I'm a master and L1 but never seen an 'EFI' cert)" and tell people how it is.

Obiously you're a pretty smart guy with a good idea of what's going on; but I've been there, tried that, and doesn't sound like you have, so to convince me, I'd need something more than what you're saying (and so would the judges!)

[MyDuster360]

Charlie, I don't get around here much, so i was trying to send the message that I have more than an "Internet education" in efi. For what its worth, I'm in no way a Master Tech but am ASE cert in Engine Performance,Advanced Engine Performance and Parts Specialist.

Erich hit on it earlier and I think your right also, about the drag racing example I gave, Steady WOT is different enough to make a differnce.

I'll say it again as I did earlier, I thoughts on this might change in a few months, but for now im sticking to my guns and say O2 comp has a place on our car.

I've tied to back up my opinions with good reasoning, practical examples, and real personal experience from outside FSAE but ill step it up a notch. No time now, more facts later.

[Mustang Mac]

I have to agree with Charlie. We have run the PE for 3 years and two years ago we tried applying wide band O2 comp. As charlie stated, the complete response time, ie sensor, ecu, injector, is to great and by the time it changes the A/F ratio, the engine is so far from that initial point that the current mixture is inadaquate. The engine ran like crap and many hours were spent testing and tuning. Our biggest step forward was actually removing the O2 loop, all of a sudden the engine ran smooth.

Some people just have to learn on there own, so you can take what has been said with a grain of salt, but trust some of the people here who have tried this and you will save yourself some headaches.

[KevinD]

i'll save you the time of learning it yourself. i got drilled in design semis last year because of the way our wideband 02 was implemented. in a 3rd gear acceleration, the car goes through 3000-13000 RPM in about 3-5 seconds (don't remember specifics here... but you'll get the idea). that is about 2-3k rpm per second. when you check the specs of the wideband you will see it has a lag, and just for arguements sake lets say .2 seconds from the time it is burned in the cylinder, flows through the exhaust, picked up by the sensor processed by the controller, and sent to the ECU to make the correction. in .2 sec at 3000rpm/s your looking at 600 rpm difference from the time it is measured until the soonest it can correct.

and as Charlie has already mentioned, autocrossing your RPM will change by large amounts very quickly throughout the entire run.
And from a tuners perspective, i was having problems with 1000 rpm resolution fuel tables not being enough to tune out bad resonances, so i would have little to no faith in a closed loop system with a MINIMUM of 600 rpm lag (remember, done in first gear, you can go through 10k rpm in a matter of a second or so).

take that for what you will, but i learned my lesson from the judge and wont make that mistake again http://fsae.com/groupee_common/emoticons/icon_wink.gif

[DaveC]

If you want to run closed loop, I'd do it similar to an OEM design, and go to open loop if the rpms or throttle position are over a certain point. I think you'll find it'll stay in open loop most of the time. I want a wbo2 on our car, but really just for dyno pulls. Just my $.02...

[MyDuster360]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Mustang Mac:
I have to agree with Charlie. We have run the PE for 3 years and two years ago we tried applying wide band O2 comp. As charlie stated, the complete response time, ie sensor, ecu, injector, is to great and by the time it changes the A/F ratio, the engine is so far from that initial point that the current mixture is inadaquate. The engine ran like crap and many hours were spent testing and tuning. Our biggest step forward was actually removing the O2 loop, all of a sudden the engine ran smooth.
</div></BLOCKQUOTE>

Ryan, I'm used to a much more sophisticated O2 comp control which is setup to cure the exact problem your describing. But without the abililty to program at least the transport delay in the PE ecu, your making alot of sense. I still think it could be done and done well, but most likely not with our PE ecu. I just wish we had the budget for a high dollar ECU.

[Garlic]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by MyDuster360:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Mustang Mac:
the complete response time, ie sensor, ecu, injector, is to great and by the time it changes the A/F ratio, the engine is so far from that initial point that the current mixture is inadaquate.
</div></BLOCKQUOTE>

Ryan, I'm used to a much more sophisticated O2 comp control which is setup to cure the exact problem your describing. </div></BLOCKQUOTE>

I'd love to know how such a system works! It can predict what your RPM will be after the delay, and also, predict what % difference fueling you'll need too?

If you need a high dollar system to make this happen, I guess you are talking about a higher dollar system then a MoTec, 'cause it won't work either.

Who carse if you can program your delay time, the problem is that you just don't know where you'll be next. Knowing your delay time only works if you are at least close to steady state, which these cars never are.

[VFR750R]

Nascar runs open loop

[VFR750R]

God I'm too funny

[Erich Ohlde]

I heard ur car last year Cornell. Sounds like u need to shim your injector needles.

Yea, that was as bad as urs http://fsae.com/groupee_common/emoticons/icon_razz.gif

[raska]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by VFR750R:
Nascar runs open loop </div></BLOCKQUOTE>

http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

With their steering too.

[murpia]

Closed-loop fuel control with a wide-band O2 sensor will save you a lot of time when dyno tuning. Just make sure you can easily import the feedback corrections into the open-loop fuel map. While you're at it, measure the transport delay and O2 sensor dynamics with some step inputs to fuel.

Then, do some steady-state (long straight) track testing, to verify the installation differences dyno -&gt; chassis. Again this can be closed loop, as long as you import the feedback corrections into the open-loop maps.

Then, turn-off closed loop but leave the wide-band O2 datalogging and go transient testing. Apply the transport delays / sensor dynamics and you can tune your transient fuel maps to compensate adequately for manifold filling, and port wetting effects.

Finally, specify to the cost judges that under article 4.3.7.4 your wide-band O2 is datalogging only and can be removed and be replaced with a bung. Then you can run it and double-check your transient fuel maps during the event.

Ian

[Chris Davin]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by VFR750R:
God I'm too funny </div></BLOCKQUOTE>

Wow, it took me almost two hours to get it.

[MyDuster360]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Garlic:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by MyDuster360:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Mustang Mac:
the complete response time, ie sensor, ecu, injector, is to great and by the time it changes the A/F ratio, the engine is so far from that initial point that the current mixture is inadaquate.
</div></BLOCKQUOTE>

Ryan, I'm used to a much more sophisticated O2 comp control which is setup to cure the exact problem your describing. </div></BLOCKQUOTE>

I'd love to know how such a system works! It can predict what your RPM will be after the delay, and also, predict what % difference fueling you'll need too?

If you need a high dollar system to make this happen, I guess you are talking about a higher dollar system then a MoTec, 'cause it won't work either.

Who carse if you can program your delay time, the problem is that you just don't know where you'll be next. Knowing your delay time only works if you are at least close to steady state, which these cars never are. </div></BLOCKQUOTE>


Garlic
transport delay has nothing to do with RPM predictions.?.? It simply will not execute the O2 comp until there has been suffiecent time for the "new" corrected A/R exhaust to get(transport) to the O2 for a new reading. Thats why I think Ryan was having such a problem. The PE has no adujstment for the appropriate lag, his sensor was trying to RECORRECT the A/F before the newly CORRECTED A/F got through the engine and to the sensor.

Holley 950 (http://www.holley.com/data/Products/Technical/199R10149-7.pdf)

Page 79-81 of this pdf has the best discription. This is the type of o2. control is what I'm used to. DFI GEN7 has somthing similiar. Im not for certain about Motec but are you sure they don't provide any sort of thing like this? You'd figure if they had on-board wideband controler they put in some kind of adjustment for this.

[DaveC]

Yes, but the Holley system still has a provision for open loop. If the feedback system could always work well and be of benefit, there would be no such thing as open loop operation. This is all fairly standard stuff as far as how modern efi systems work. As I said, give it a go. The car will probably stay in open loop most of the time, and if you program it correctly as to when to enter open loop vs closed, it wont harm performance, it'll even give you a fuel economy advantage for all the time the car is at low rpms with low throttle inputs http://fsae.com/groupee_common/emoticons/icon_razz.gif.

[MyDuster360]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by DaveC:
Yes, but the Holley system still has a provision for open loop. If the feedback system could always work well and be of benefit, there would be no such thing as open loop operation. </div></BLOCKQUOTE>

The provision is there because of the narrow OR wide band user option. Also open loop is still needed for warm up of course, but especially idle(big cams and long tube headers) where the sensor often cools off to a point where it gives a false rich signal. A lot of guys I know run closed loop at all other times but those.

Noe the less, I think I'm stuck with running open loop if I go with the PE ecu anyway. Unless any one has any other ideas?

[DaveC]

Maybe the AEM unit? It's a bunch more $$$, but nowhere near a Motec.

[Garlic]

Of course the MoTec has features like that, and more.

The point is there is never time for the closed loop to operate. The transients are too quick.

You can spend the time to program in enough limiting parameters in the closed loop system so it only operates when it is effective, and the result would be it would never operate!

Like DaveC said, it might go closed for just a second if your driver is tenative on the throttle, but totally off throttle you don't want closed loop because, in most instances, you shouldn't be feeding any fuel at all.

And even if it operates at some part throttle, it is not a guaranteed increase in fuel economy.

Map your stuff right, keep logging it everytime you run and tweaking it, that's how you get a good engine calibration.

[DaveC]

Also, the f4i doesnt use an o2 sensor stock (at least the older motors we have), except California models for better emissions test performance. I do believe a wb is an invaluable tuning aid, but as for runing closed loop all the time, maybe you can get it to work for you, give it a go and let us know how it works for you...