'Actual consumption' values...

[johnl]

OK, today I had 'actual consumption' displayed on the dash (I don't usually have this display mode on, but occasionally I have done). The readout values seemed wierd, quite intermittently erratic.

With an already up to temp engine, starting the engine and driving off sensibly I'd initially be seeing up to 25 L/100km (which seems bad...). This would then quite gradually creep down to more reassuring and expected numbers. Cruising at 100 - 110 kmh on the highway (fairlry level road) I noticed the readout showing over 10 L/100kms, up to about 12 L/100kms on an incline, which seemed abnormal and way more than I've previously seen in similar circumstance (which would be more like 6 or 7 L100kms, + or -). Again, the reading would gradually decrease, but any decent prod on the throttle would see it instantly rise to maybe 19 L/100kms (+ or -), and again take a long time to fall back to reasonable numbers. Backing off completely would see 2 L/100kms displayed (expected), but then even just touching the throttle would see an instant readout often over 10L /100kms (+ or -).

On the return trip the readouts were much more normal, until I experimentally floored the throttle and they shot up again, then started repeating the above behaviour. While this was happening the engine felt less powerful. This wasn't happening all the time, it came and went, and when it wasn't reading wierd numbers the numbers were more or less what I've see before and the engine felt better. Something seems to be responding slowly, at the very least.

So, I'm thinking there is an issue. Maybe the MAF? Maybe one or both O2 sensors?

Thoughts?

Regards,
John.

[bazzbazz]

"Ghosts in the code?" 

[baldrick]

Is it possible that your recent rear brake fiddling has left you with dragging brakes? Easy to check, go for a drive at speed along an open road for a while, coast a stop without using the brake hop out and feel the temps of the rear wheels. The wheels do function as a heat sink and if they are warm to the touch but the fronts are not... bingo. Out of interest is there a big difference with the average & average B fuel consumption figures? Its also possible your fiddling with the RbW throttle may be a factor.

[johnl]

Thanks for the input.

This crossed my mind when I first noticed the strangely high 'Actual Consumption' being shown on the dash. So I coasted to a halt (from 100kmh) and checked the brake temps, i.e. felt them with my hand. The rear brakes were just a bit hotter than the front brakes, but not particularly hot by any measure.

The rear brakes are dragging to some degree, but if this were causing the apparently high intermittent fuel consumption (assuming what the readout is saying is correct), then considering how much extra fuel seems to be being used (when the reading is displaying unusually high numbers, which is intermittent), then the brakes would have to get very hot. All that extra fuel would have to create a lot of energy, and if the high consumption were mainly a product of the drag then that energy would have to end up in the brakes as a very high temperature.

Later I actually stopped, jacked a rear wheel up, and rotated it by hand. I could hear the pads rubbing, but the wheel was easy to rotate. The other rear wheel was just as easy to rotate. I just don't think the drag that exists could be enough to account for the consumption numbers I'm seeing on the display. Put it in neutral and the car coasts a long way as well, and it will pick up speed coasting downhill. It just doesn't feel like the brakes are slowing it any more than is usual.

I can't reconcile the dragging brake idea with the way in which the displayed consumption will go very high with any substantial throttle input, but then only very gradually go down again after lifting off and even decelerating (normally it lowers quickly after lifting off), despite no change in speed (on more or less level roads). If the issue were drag, then I can't see why the high consumption displayed is intermittent, i.e. the displyed numbers can be high or low for periods of time in similar circumstances (i.e. speed, road conditions). It's just a bit strange, something is going on and I'm fairly sure it's something to do with the engine...

The throttle pedal mod is only a simple pedal travel limiter. When I floor the pedal (as far as the limiter allows) the car accelerates well (when it seems to be running properly), and I can actually lift the pedal off the floor by about maybe 5mm before I start to feel acceleration diminish. I can't see how merely preventing the throttle pedal from going all the way to the floor could do wierd things to the fuel consumption.

I still like the limitation on the throttle pedal travel, as it means the engine responds to any pedal movement right up until the pedal is nearly touching the floor, rather than having 'extra' pedal travel that doesn't seem to do anything.

The brake pedal still feels very good, much better than before I tightened the rear caliper piston to pad clearance. I'm quite sure that the previously 'sloppy' brake pedal is a problem associated with the rear caliper design.

Regards,
John.

[poohbah]

I think it's probably just how these devices work. Our Subaru has one that shows two concurrent figures, one being an average over the trip since ignition, and the other showing "actual" as you go consumption. If I floor it, it immediately spikes over 20L/100km but then if cruising steadily declines to around the usual average of circa 8/100. It spikes whenever you accelerate and drops when you cruise, but the trip average is usually roughly the same.

[johnl]

The display has two fuel consumption modes. Depress the end of the column stalk once and you get a readout of instantaneous 'actual consumption' (i.e. consumption over the last few seconds of driving), press it again and you get 'average consumption' (i.e. consumption over a much longer distance / time, i.e. since the readout was last cleared, or over up to the last X number of kms). If the average consumtion hasn't been cleared quite recently (by holding the stalk button down for some seconds) then it takes quite a lot of driving in a changed manner (heavier foot, or urban vs highway etc.) for the average consumption readout to change. The 'actual consumption' mode readout changes (should change) very quickly dependant on throttle opening.

It's the 'actual consumption' readout that is behaving strangely here. It used to be different than it is now, in that it used to be consistent with the manner in which the car was being driven at any point in time, but it has now become inconsistent. Sometimes it is behaving as before (as it's supposed to). Sometimes it gives numbers much higher than I was seeing when it was consistently reading 'normally', and then it seems to change back to normal, for a period of time.

When reading normally, the display is quite responsive to reducing throttle position after heavier throttle use, i.e. the numbers quickly reflect lower throttle opening (and apparent lesser fuel usage). When it's reading abnormally the numbers are not very responsive to reduced throttle, they do lower but only very very slowly. It seems as if the the ECU is still fuelling for a wider throttle opening than is actually the case, taking a lot of time to 'catch up' to the fuelling required at a much lesser throttle opening. It's weird.

The numbers I am seeing for larger throttle openings also seem much higher than they used to go. Previously I'd never see numbers over 20L / 100kms when flooring it, now I am easily seeing numbers that high. Note too that when the display is reading stangely that the engine seems a bit 'flat' and less responsive. It also sounds 'flatter' than when the display is reading 'normally'.

My chief suspects are the MAF and the O2 sensors, maybe intermittently reacting slowly to changed conditions(?). Today I think I'll pull the O2 sensors and have a look at them, maybe give them a blast with a blowtorch...

Regards,
John.

[MattK]

Ha, re brake dragging when we lived in Newcastle nearly 20 years ago, I drove a 33 which I think had discs at the front and drums at the back? Anyway, the proportioning valve clogged up and locked the rear brakes on, partially, in the latter part of a 2 hour drive to Forster / Tuncurry. As we got near the town, the terrible hot metal smell I thought was coming from the Land Cruiser in front, stayed with us. I went to brake at the first roundabout and pretty much nothing happened! When I got out to see what was going on, the rear wheels were too hot to touch and the car was pretty much undriveable - had to stay overnight in a motel and drive back in a rental the next day.

[poohbah]

To be honest the only thing I find of any practical use is the range indicator. Though I have my doubts about them too. Subaru low fuel warnings start piping up when gauge gets below one "bar" but range indicator still shows +80km. And I doubt it's even that low. I've driven round for days after the range indicator on my old EA Fairmont, and AU falcon was showing zero.

[johnl]

So blasting the O2 sensors with a blowtorch made zero dirrerence, not unexpectedly, it was a long reach at best.

I'm getting no brake 'lock on' of the kind that MattK has reported for his old 33, despite the pads lightly rubbing the rear discs. All this seems to have done is make a substatial improvement to the pedal travel, general brake feel and braking response. I once adjusted the pedal free play to zero (with my old Accord), and prompltly had very similar symptoms to what MattK had with his 33. The slight pedal free play is there for a reason (to ensure that the master cylinder piston seal can't obstruct the pathway between the cylinder bore and the reservoir).

Regards,
John.

[Citroënbender]

The "actual consumption" figures are always buffered, also I note the software doesn't allow such small closed throttle numbers as older ECUs. My '94 Pug will drop to 0.2/100 when coasting in top gear, the 147 never gets close. Many programs won't read higher than 25 or 30/100 so that can be why such a number is seen...

A weakened spark is one reason for possible higher fuel consumption under load, I am starting to think the coil packs don't age well. Have you checked hot, cold, wet, dry compression recently either?

[johnl]

I would expect some degree of buffering, or the readout would likely be unreadable due to moment to moment fluctuation (?). However, the readout is rapidly responsive on the upswing, very not at all on the downswing. The impression is that either the AFR is indeed very slow to lean off (why?), or the readout is not reflecting the reduction in AFR at the rate at which the AFR is actually leaning off.

The impression I have is that a reading of 25 L/100kms is the maximum dispayable readout, even if the momentary real consumption might actually be more than that. I've never seen more than a decimal point free 25 l/100kms displayed, even with a heavy throttle on a cold engine (don't say it, I know, rare occurance when in a big hurry). On cold start up I'm also seeing 25 for a few hunderd metres of light throttle driving (then a slow decrease), which I'm fairly sure it didn't used to do (high teens on cold start up yes, over 20, no...). And, even with an up to temp engine I've been seeing high readouts up to 25 L/100kms after restarting from a short stop (say filling up with fuel), but for a shorter distance.

It's also erratic, for example yesterday; driving at 70kmh on a longish straight flat stretch in fifth gear on a steady whiff of throttle, the readout was a steady 5.5 L/100kms, but later driving the other way on the same road in an identical manner the readout was saying 7.5 to 8 L/100km. This is a pretty big unexplained difference that is hard to reconcile with all working as it should (even if the road in question were on a very slight incline, which it doesn't appear to be).

Haven't checked compression, it doesn't seem like a compression issue. Havn't checked ignition spark strength, doesn't feel like an ignition issue, it never misfires (that I have ever noticed), nor struggles to start easily.

When I had the O2 sensors out they were white and dry with light surface deposits, when I was more half expecting them to be black and sooty (happilly they weren't oily or heavily encrusted with crap suggesting problems with the engine itself, though this would have given an imperative to change the sensors). I guess I'll just throw some sensors (O2 and MAF) on it when disposable funds allow (and keep scratching my head if / when it makes no difference...).

Regards,
John.

[Citroënbender]

Blocked front cats, compromised compression, suboptimal coils are my suite of proposed issues that beset your engine.

Consumption will always drop more slowly than it rises, you have alerted the ECU to a possible need for increased power and there is no way you can instantaneously tell it the request is 100% finished. A bit like spotlighting; see a roo, stop the ute and take aim... Roo hops away before you can safely fire - you don't immediately lower your gun, you keep it raised and steady while you scan around in case there's others.

[bazzbazz]

Anyone ever thought there is a fault in the circuit that calculates the fuel consumption and that there is nothing wrong with the engine.   

You would be amazed what the mind will do to try and make sense of what it is being told. When the screen tells us there is something wrong we start to imagine all sorts of things to justify what it is telling us. 

[johnl]

CB,
wouldn't such issues cause consistent power / consumption problems? My engines' apparent issues seem to be erratic. When my engine is running well, it runs very well, plenty of power for a 2 litre engine (though too much is never enough...), and quite responsive. When it is running badly, it's not hugely worse, just somewhat 'doughy' in comparison (feels somewhat like the ignition has been retarded to some degree).

The 'actual consumption' display can respond very quickly to lessened throttle, but only when the throttle is fully closed (from some degree of being open), when it almost instantly drops to "2L/100kms". I assume the ECU is then taking it's fuelling cue from the throttle potentiometer (whether this is actually a full shut off or only partial, I assume would be for emissions reasons), and ignoring other sensors?.

What appears very odd to me is that any lessened throttle not great enough to trigger effective fuel cut off results in what seems an extremely lethargic response on the display, so if the display is to be believed and the AFR is indeed remaining rich for some time after throttle reduction, then considering that the air flowing into the engine has been significantly reduced by the lessened throttle, then the AFR must (?) be going very rich indeed, i.e. get even richer, at least for a relatively short time until the ECU catches up with what is actually required for a lessened throttle opening...? I can't see why the ECU would have to 'overshoot' the reduced throttle fuelling by nearly as much for as long as it seems it might be doing...

It's my understanding that the ECU would be 'measuring' economy by counting the number and width of injector openings X distance being covered in Y time, so the display should be a reasonably good benchmark for momentary consumption.

Dunno, maybe I should stop thinking about this, for now...

Regards,
John.

[Citroënbender]

It's my personal experience that a failing cat can be intermittent.  Whether due to loose parts shifting around in the matrix (ceramic type cats) or raised internal temperature in metal substrate cats actually burning some of the soot to the point where it breaks down and allows partial flow in the affected zone again. 

The longer it's left untreated, the more this will impact on the rest of a motor... 

As to spurious consumption data, why or how could just that aspect be (digitally) wrong in isolation?