E36 328is Crazy high fuel consumption

[OptiManc]

E30M3SE said:

If you haven't got a mis-fire it's not likely to be a coil issue.

Coils can be visually inspected for cracks, and tested by measuring the resistance between the outer 2 pins (there are 3 pins), should read somewhere between 0.4-0.8 Ohms.

Thanks E30M3SE, my coils seem perfectly fine beside a small amount of carbon.

Something that had been nagging at me for some time was the idea of changing the sparks. When I got the car 3 years ago i had a pack of 8 Bosch quad spark spark plugs gathering dust, so i removed the single spark BMW branded ones (with 97k miles on the clock I assumed they were OEM) and upgraded to the Bosch quad sparks and these were replaced like for like during the service intervals.

So last night I down-graded the plugs to twin spark NGKs.

Result; a little bit drop in torque (sensation only not measured) but slightly smoother drive, but ultimately no difference in fuel consumption. However, I had disconnected the battery during the exchange, as I have done during every sensor changes. Resulting in early signs of improvement then after an approximate 10 mile drive the problem would presents itself again – fuel gage needle dropping, range counting down and MPG eye watering.

Disconnecting the battery, I assume is causing a reset or flush of configuration or stored data which is then being collated during running time offering 'normal' fuel consumption results, the data is then possibly becoming 'corrupted' and then resulting in high fuel consumption and poor performance.

Assuming this is correct, I'd assume again it’s the ECU where this data may be stored or used. Therefore, is an ECU replacement beyond the realms of reality? The ECU has been flashed only some months ago, but being from an IT profession I know computers don’t always behave as expected even from a flash or upgrade.

[vsonix]

It could well be the ECU being dodgy, I think it mightn't be a bad idea to look into it a bit further. The car running well initially indicates possibly all sensors etc could be fine. What would be interesting to do - drive somewhere with plenty of open road for testing economy, then, reset the ECU again, and with the engine as close to operating temperature drive around for a bit, keep resetting and checking MPG1 or MPG2 so you get a rough idea of your consumption at various speeds on the OBC then see when it changes/gets worse, and by how much by. If you can pinpoint a certain amount of time every time, that could be a big clue too...

I must admit I am enjoying this collaborative diagnosis, it's what I love about car forums, BMW ones in particular

[jz325i]

Have you checked the resistenace of the crank sensor with a digital multimeter?

I'm pretty sure yours would be the rear mounted crank sensor. Should be part 25 on this page ...

http://www.realoem.com/bmw/showparts.do?model=BG12...

The front mounted crank sensors should have a reading of around 1280 ohms.
But I can't remember exactly what the rear mounted ones should read - somewhere around the 500ohm mark (but double check the correct reading).

If yours reads correctly then I would next test the cabling between the crank sensor connector and the ECU connector.

While in the ECU compartment, check for water ingress.

OptiManc said:

I have a '98 E36 328is with an M52 (M52TUB28 I think) auto. I've always wanted one since I was about 16, around the time the final evolution of that shape was introduced.

Anyway, problem is for the last 12 months or so I have been battling with its high, very high fuel consumption, it was never like this when I bought it in 2008.

I always fill up with between 20 or 30 litres of V Power, and I used to get a good 90-100 miles for my 30 litres depending on how I drove, more miles on the motorway of course refuelling on reserve light. Now I'll get around 50 and that’s on a good week, the worst ever was 38! Refuelling and driving off the forecourt, the Range will countdown instantly, by the time I’m off the forecourt I’ll have ‘lost’ around 4-6 miles already – but the Range will stay static while the car is stationary

I've had it serviced - inspection 2 TWICE with engine flush, changed the MAF, ran engine cleaners through it (Forte Fuel Treatment), pumped up the wheels to 44psi, tried different petrol stations, emptied the tank and refilled up to 20 litres until empty again to check fuel gage for faults, checked emissions numerous times, reset the ECU, flashed the ECU, checked for air leaks (no rough idle btw) changed numerous filters and other little things like checking brakes aren’t sticking, accelerator isn’t weighed by mats. Nothing has made a difference, cold or hot it’s the same. Furthermore, performance is laboured from take-off and acceleration at any speeds; it takes ages to pick up revs, however! Occasionally it’s perfectly fine, quick, sounds smooth and prompt and fuel consumption stabilises. After a restart it'll go back to being lazy and gulping down the fuel.

No sign of leaks or smells of petrol.

Any ideas what so ever on how to resolve this would be greatly appreciated.

Many thanks
OptiManc!

[vsonix]

OptiManc said:

The ECU has been flashed only some months ago, but being from an IT profession I know computers don’t always behave as expected even from a flash or upgrade.

from what I know about the E36 ECU I think they are of a similar power to an 8-bit Z80 chip. Probably so simplistic that a bit of corroded solder somewhere is making it think 0 = 1 @ location 255 or something. Not enough to break it, enough to fool it into thinking something's closed when it's really open, or something...

[shouldbworking]

Interesting read. At this point I think i'd be considering a compression test just to see if the engine is mechanically sound or not.

Can you swap the ECU with another from the same model?

[E30M3SE]

^^^^



Compression and leakdown, before shelling out on a new ecu swopsies with a know good ecu would be my prefered route.

[OptiManc]

vsonix said:

It could well be the ECU being dodgy, I think it mightn't be a bad idea to look into it a bit further. The car running well initially indicates possibly all sensors etc could be fine. What would be interesting to do - drive somewhere with plenty of open road for testing economy, then, reset the ECU again, and with the engine as close to operating temperature drive around for a bit, keep resetting and checking MPG1 or MPG2 so you get a rough idea of your consumption at various speeds on the OBC then see when it changes/gets worse, and by how much by. If you can pinpoint a certain amount of time every time, that could be a big clue too...

I must admit I am enjoying this collaborative diagnosis, it's what I love about car forums, BMW ones in particular

Agreed, its great to have such valuable knowledge share, so thank you to everyone who has contributed so far, your efforts are very much appreciated.

I have in the past compared MPG1 and 2 over different speeds. Since it’s an auto, I even switched on the hazards and driven only in creep on relatively level ground, the Range was counting down at its normal excessive rate, however MPG1 from what I recall was around the 40 or 50 mark. Shortly after the reserve light came on I continued to creep to the petrol station, say around 1 mile away, upon arrival the needle was well and truly past half way in the reserve.
So there must be a motion related cause.

[OptiManc]

shouldbworking said:

Interesting read. At this point I think i'd be considering a compression test just to see if the engine is mechanically sound or not.

Can you swap the ECU with another from the same model?

Thats a bit past my abilities now, so have booked her in for a check up tomorrow morning regarding the wiring and I've booked her for a compression test with BM Tec in Manchester for Tuesday 19th.

Out of interest where is the ECU located on the M52? Perhaps it just needs a

[The Nur]

OptiManc said:

Thats a bit past my abilities now, so have booked her in for a check up tomorrow morning regarding the wiring and I've booked her for a compression test with BM Tec in Manchester for Tuesday 19th.

Out of interest where is the ECU located on the M52? Perhaps it just needs a

I like the thought that the ECU just needs a good telling off

Go outside, open your bonnet and start shouting at it in a condescending manner

[OldSkoolRS]

The Nur said:

I like the thought that the ECU just needs a good telling off

Go outside, open your bonnet and start shouting at it in a condescending manner

Then snap a branch off a tree and whack the bonnet with it to really show it who's the boss.

[OptiManc]

OldSkoolRS said:

Then snap a branch off a tree and whack the bonnet with it to really show it who's the boss.

Followed by a

[OptiManc]

The telling off didn’t work, however I have noticed an improvement in fuel consumption AND performance if I try one of the following (not sure of the correct pattern exactly as I only came across this last night)
- cold start, turn ignition ON and OFF after a few seconds then start engine again immediately. The fuel gauge is steady, Range static and only drops as the miles climb driven carefully but eventually it returns to being faulty again

- turn engine ON, run for a while and switch OFF before reaching optimum temp (not recommended I know) listen out for fuel tank vent and start engine again. Same result and same outcome eventually

- when OFF but hot, turn ON ignition listen out for fuel tank vent type sound turn ignition OFF and start again immediately. Same result and same outcome eventually

I'm thinking the initial ON and OFF clears a fault, causes a confusion in the ECU which is actually a correction or the charge and voltage going through the circuit a second time helps the data travel around the system better.

Hope ive got this across accurately enough to make some sense.

Had to cancel the compression test for Tuesday due to work obligations.

[OptiManc]

Did some more testing this weekend regarding the OFF and ON theory, initiating the ignition for a few seconds, turning OFF and then firing up the engine means I get at least 5 miles of stability. The fuel gauge doesn’t move, Range is steady, performance is good and MPG1 gets to a nice 22-ish driving casually. Then suddenly it all turns brown again.


I have a compression test tomorrow at 10AM but I’m more drawn to a dicky ECU now, so I will be looking into ECU remapping services which might be able to diagnose faults with the current configuration.

[OptiManc]

Compression tests are in, 174psi from all 6 chambers.

I'm sure its down to the ecu, so remapping is Plan J.

[Rollcage]

Have you tried some decent laptop based diagnostic software and a diag cable? Many will give live feeds so you can observe differences during driving/idleing, etc.

[OptiManc]

Rollcage said:

Have you tried some decent laptop based diagnostic software and a diag cable? Many will give live feeds so you can observe differences during driving/idleing, etc.

No I haven’t actually, I have tried to get engine mod guys to come out and do that but they’ve been reluctant. I've seen the cables online on eBay but have been warned against getting cheap ones and the 'proper' ones seem to fetch into the £300 mark. Spoke to a BMW remapping 'expert' on Friday morning, long story short after describing the issue he said it’s not something he thinks he can solve with software and wouldn’t want to get involve with it!

However, yesterday while driving home it began to smell of melting plastic soon I saw smoke wafting from the engine bay and the temp gauge shot up to the red mark instantly! I pulled up and saw coolant purring out and a water pipe to the front of the engine ruptured (will try and get pics actually). So, got towed home, replaced the pipe immediately after having my tea. Topped up the coolant with just 1 litre of water

Tomorrow she’s going in for a closer look, I suspect a blockage somehow and cross my fingers it’s not a gasket. Wondering though if this could actually be the phantom problem all along and has now finally caused the eventual outcome.

[OldSkoolRS]

FWIW you don't need to spend £300 on a cable. I know there are some dodgy ones about, but I went for one that was recommended on various BMW forums and it has proven to work very well. If your recent overheat doesn't lead to a solution, then I'd recommend trying these (no connection to them myself):

http://bmwcables.moonfruit.com/

He even responded to some email questions on a bank holiday, so I got my issues sorted out quicker than I expected. I got the £34.99 version as I need to use the 20 pin underbonnet connector on my Z3 to access all the items, though the OBDII connector under the steering column works for engine errors, resets, etc which would help for scaning whilst driving (with a passenger on the laptop of course ).

[OptiManc]

OldSkoolRS said:

FWIW you don't need to spend £300 on a cable. I know there are some dodgy ones about, but I went for one that was recommended on various BMW forums and it has proven to work very well. If your recent overheat doesn't lead to a solution, then I'd recommend trying these (no connection to them myself):

http://bmwcables.moonfruit.com/

He even responded to some email questions on a bank holiday, so I got my issues sorted out quicker than I expected. I got the £34.99 version as I need to use the 20 pin underbonnet connector on my Z3 to access all the items, though the OBDII connector under the steering column works for engine errors, resets, etc which would help for scaning whilst driving (with a passenger on the laptop of course ).

Thanks for that, picking her up tomorrow after a water pump and thermostat replacement. Thermo was stuck closed and pump has some play in it.

Will invest in these cables post-haste if the current work hasn’t solved the issue in anyway.

[vsonix]

Fingers crossed this will solve the issue! Thermostat + housing is known for being an issue but I thought we'd eliminated it a while back... If it does, all the other work you've done should see you right for years

[OptiManc]

Right, shes sprung a coolant leak now . I suspect its either a hose or the rad itself. However, since the thermo and pump were replaced i did notice a gradual improvement in fuel consumption and performance was sharper, not as good as before but certainly more prompt with room for improvement.

But now its somewhere in between, not as bad but not as good as before either.

My theory is that the fault was intermittent due to the faulty thermo, when it would eventually open the system would cool sufficiently and performance would improve. The new kit had, temporarily solved the issue, but now with the rad leak, inefficient coolant level means the same symptoms yet again but this time caused by something else.

Am I right or along the right path at least?

Going to avoid leak stop stuff and look at replacing the hose if thats broken else replace or repair the rad.