Closed Loop Mode Disengaging Too Early

[JimX]

I noticed something today when checking out some PowerFC map traces and monitor screens, and that is it seems my car isn't staying in closed loop mode when it should. I found that the ignition timings would drop down to 15-25 degrees or so even I'd set it for around 35-45 in that load area, but as soon as I squeezed on more than the tiniest amount of throttle it would jump straight up to what I had programmed into the map.

That means I can coast down hills in closed loop mode, and can't even maintain a steady speed on level ground if I want to keep it in closed loop. This is all well under 3000rpm obviously (usually 1500-2000). The load point areas don't seem to matter either. It can be on zero and <3000rpm and still not going to closed loop, just because I have my foot on the accelerator slightly.

Does anyone know why it would be doing this and what I could modify/edit to fix it? I thought maybe it might have been because the throttle position sensor was set incorrectly but it's showing up as 0.51V when off which seems to be ok from what I've read.

I'm 99% sure that not being able to get into closed loop mode is why I have such crap arse fuel economy. Closed loop seems to be working fine, it's just that I'm never able to stay in it long enough to save me any fuel. Any advice appreciated!

[paulr33]

as far as i've read closed loop mode is based on the oxygen sensor values:

That's between 0.6 and 0.1 volts on a slow/narrow lambda sensor. You should be able to see that voltage on the Datalogit at idle and cruise, that will confirm closed loop running. It's not instantaneous, it doesn't drop to 0.6 volts as soon as it idles, closed loop running takes a few seconds to engage. That's to allow for the slow part of the slow/narrow lambda sensor.

If you are seeing voltages around 1 volt after say 15 seconds, that means non closed loop running. In which case you will need to check the lambda snesor operation, the PFC settings etc. Make sure the O2 Feedback Control box is ticked on Settings 1.

if i check in my datalogit logs i can see mine going in and out normally. if you are seeing it drop the timing values quickly this will be based on the airflow meter "load" signal and this is normal. the airflow meter will be really quick to tell the ecu more air is coming in (even on light throttle changes) which is why you can see the timing drop instantly (well almost).

when you have all the map points tuned it should be instantenous with ign timing changes which gives a super nice crisp response, something i've never seen in a car before until I tuned my points

[blind_elk]

Closed loop has NOTHING whatsoever to do with ignition timing (in the sense that you have raised, at least).

Closed loop is where the ECU monitors the O2 sensor, and makes adjustments to the fuelling to create a stoichiometric mixture. It does this in a reactive cycle, so the A/F readings cycle up and down the range of A/F readings, oscillating around the stoichiometric point.

ie A/F lean, add fuel. A/F still too lean, add more fuel. A/F rich, reduce fuel. A/F still too rich, reduce fuel. A/F too lean, add fuel. etc.

If you are varying the load or throttle, the ECU should realise that it cannot make a stoichiometric mixture, and will drop out of closed loop until it senses a steady state, at which point it re-engages closed loop.

In most programmable ECUs, it should be possible to tell the ECU when closed loop is and is not active, how it should make adjustments, and how big those adjustments should be.

[N I B]

PFC does not allow for adjusting closed loop at the level most other ECUs do.

[RB30-POWER]

If the light load map cell figures aren't really closely tuned to 14.7:1, closed loop won't be able to alter the pulsewidth enough to acheive stoich ratio's.

Disable closed loop, tune all the light load cells in the map as close to 14.7:1 AFR as you can, then turn closed loop back on and presto.

Its just that closed loop operation can only alter the base map/cell pulsewidth value by something like + or - 5%

Either way how to do you know that it isn't in closed loop, do you have an air/fuel ratio guage that you can see not going back and forth?

[JimX]

I don't know if the ECU is going out of closed loop. I was seeing the ignition timing drop and move around as described above, rather than stay constantly at what I'd set for that particular load point. Similar to what it does at idle (moves around between 15 and 17 or so, instead of constantly at 20). Whatever the case, my economy sucks. I guess I'll just live with it until I can get it fixed by someone that has the right equipment to measure the AFR levels and knows what they're doing.

[Sydneykid]

I don't know if the ECU is going out of closed loop. I was seeing the ignition timing drop and move around as described above, rather than stay constantly at what I'd set for that particular load point. Similar to what it does at idle (moves around between 15 and 17 or so, instead of constantly at 20). Whatever the case, my economy sucks. I guess I'll just live with it until I can get it fixed by someone that has the right equipment to measure the AFR levels and knows what they're doing.

<{POST_SNAPBACK}>

Hi Jim, you can tune the light load runing using the standard slow and narrow lambda sensor. Trim the fuel to get ~0.8 volts for all load points from just above idle (say 1,100 rpm) to earliest boost build ( say 2,200 rpm). It is a slow process, but the datalogit logging will help.

cheers

[blind_elk]

I was seeing the ignition timing drop and move around as described above, rather than stay constantly at what I'd set for that particular load point.

<{POST_SNAPBACK}>

When you tune the PFC, you set a value at a particular load / rpm point. But the engine doesn't necesarily run exactly on those map points. So the PFC has a guess (it "interpolates") at what the value would be at that exact load / rpm

eg say you set fuelling at 1.40@750, 1.42@1000, 1.45@1250 rpm points in a particular load band. Now the engine is idling at 875, and you can't program a fuel delivery for 875 rpm, so the PFC has to calculate what fuel would be at 875, based on the values you have told it for 750 / 1000. That would mean a fuel delivery of 1.41, since 875 is halfway between 750 and 1000, and 1.41 is midway between 1.40 and 1.42.

Now, because the engine idle isn't perfectly stable (it might wander +/- 20 rpm), then the amount of fuel required will continually be being re-calculated.

The same is happening with the ignition, and you can't necesarily run the engine on the exact load point you have programmed. That is why you are seeing the ignition timing varying from the values you set.

[JimX]

Thanks for the info everyone. The reason I thought closed loop wasn't engaging is because lately I've been getting up to 17-18L/100km around town, with fairly normal driving.

Because of that I was worried that I was going to get crap mileage on my road trip yesterday but it seems I was mistaken. I ended up getting 10.4L/100km which is slightly better than what it was before I got the PowerFC. I'm completely happy with that mileage even though I know others are getting better.

I did some light load tuning on the freeway, but that was mostly to eliminate knock due to the heat. When I get a Datalogit I'll give it another go as Sydneykid suggested.

[paulr33]

have you had your afrs checked on the road with a wideband o2 ?

[JimX]

No, it's only had the AFR's checked on the dyno.

[paulr33]

wont hurt to get your tuner to drive around with you with a wideband 02 hooked up, as on road load / conditions can vary a lot from dyno cells. try plodding around in some local streets, then head onto the fwy for a quick cruise at 100km/h then go up a few hells, should sort it out fairly quickly

[JimX]

Well, it'll hurt my wallet Flooring it up hills is what I've been doing these past 2 days on a 700km round trip up north and back to visit relos. I think it's acceptable for what I've done for now even if around-town economy continues to suck. It was good that the weather was so hot, I backed off my timings a bit to get it at a "safe" knock level and I still got great economy.

I got surprisingly better mileage than I was expecting on the freeway, so I might just live with crap city mileage for now and maybe do a proper road tune once I get my new turbo next year. I'm still working on the low-end timing so perhaps I can still get a bit better economy around town without forking out more tuner $$$ for exact O2 levels.

[paulr33]

yeah know how u feel

im trying to save for a house and gotta resist spending more money

next on my hit list is water spary fmic kit