SaabCentral Forums banner

1 - 20 of 26 Posts

·
Registered
Joined
·
883 Posts
Discussion Starter #1
40mpg and 440miles on a single fill of fuel!!! - that's at least 100miles more than I ever got before for a long M-way cruise at 75mph. A few weeks ago I reset my AMM to a higher resistance and noticed better MPG but also higher running temps and slight knock when I boosted hard - indicating it was running lean? Earlier this week I had to replace the battery (which was very old and I had destroyed by running it flat by leaving the interior lights on) and the high temps and knock have disapeared. Possibly the discharged battery had reset the ECU - or could it be that a poor battery was part of the problem all along? The car is a 91 T16 with Lucas EFI + Lambda, 22 deg BTDC idle ignition setting (10deg on boost), No Cat, stock ignition components and fule pressure. BTW the car feels very smooth and strong pulling through the power - like a new car:lol:
 

·
Registered
Joined
·
204 Posts
Nice one!

I got 34.5 mpg(imperial) out the other day running at 70 along the motorway. I have also had 33s and 31s. I've been really impressed because when I first measured it I got 19. Must have messed it up somewhere.

In thinking about it I did think it must be possible to get 40. I thought that by replacing all the bearings and greasing them nicely, by changing the fluids, and maybe running higher primaries might do the trick. By the sounds of it you haven't done this, so doing so should improve things even more. Maybe doing all these changes and runing the motorway with a cleaned and waxed car on a day with a lower than average barametric pressure and a small tailwind maybe this could be stretched to 45. That is diesel territory.:cool:
 

·
Banned
Joined
·
7,270 Posts
the only guess I can come up with would be that the ECU has some sort of adaptivness in it which starts-off rich and straightens itself out as you drive the car. what your saying about the AMM makes sense. Lean = knock and better fuel economy. I always thought the 900 uses way more fuel than it would require to move a modern car of the same spec. Can't wait to see what MPG I get with the Motec. I'm expecting it to be much better than standard MPG.
 

·
Registered
Joined
·
883 Posts
Discussion Starter #4
I wish I had done this years ago - i can't imagine how much fuel I've been throwing away over the years;oops: The upside is that I'm now more minded to do some restoration work and keep the old girl rather than buying a 9-3.
 

·
Registered
Joined
·
604 Posts
trackside said:
40mpg and 440miles on a single fill of fuel!!! - that's at least 100miles more than I ever got before for a long M-way cruise at 75mph. A few weeks ago I reset my AMM to a higher resistance and noticed better MPG but also higher running temps and slight knock when I boosted hard - indicating it was running lean? Earlier this week I had to replace the battery (which was very old and I had destroyed by running it flat by leaving the interior lights on) and the high temps and knock have disapeared. Possibly the discharged battery had reset the ECU - or could it be that a poor battery was part of the problem all along? The car is a 91 T16 with Lucas EFI + Lambda, 22 deg BTDC idle ignition setting (10deg on boost), No Cat, stock ignition components and fule pressure. BTW the car feels very smooth and strong pulling through the power - like a new car:lol:
I have been playing with my AMM too, allso with reseting the ECU!
Can you tell me what your specs are on the AMM? Do you messure on the first and the last pin?
 

·
Registered
Joined
·
883 Posts
Discussion Starter #7
Resistance between pin 1 and 6 should be 336 ohm - I set mine to 450ohm. I heard somewhere that 10 ignition on and off sequences resets the ECU - easier to unhook the battery for 10mins or so to be sure though.
 

·
Registered
Joined
·
17 Posts
Interesting stuff. I guess your naff battery could have been reducing the spark voltage a bit, leading to poor combustion? Or maybe the weather has changed and air density is better?
Are you thinking of putting the car on the dyno - or connecting up a wide-band lambda sensor soon? And why did you chooses 450 ohms, any particular reason? So many questions... :)
 

·
Registered
Joined
·
883 Posts
Discussion Starter #9
Could well have been the poor battery - I was noticing a steady decline over the years of the economy. Air density should make no difference - it's an air mass meter and it will take account of atmospheric changes. No intention of testing it on a dyno - the economy is good, the power is good and there is no knock on boost. I chose 450 as it's about 100ohms more than stock - more resistance = less voltage to the ECU = leaner mixture as the ECU thinks less air is entering the engine (so the theory goes).
 

·
Registered
Joined
·
883 Posts
Discussion Starter #11
There is an adjustment screw on the body of the AMM - often sealed with a tamperproof plug. The LH2.2 units had an adjustment which as far as I can remember adjusted the whole range - not sure about the later LH2.4.
 

·
Moderator
Joined
·
4,637 Posts
Two words: Tread carefully...

More words: I'm not claiming to know anything bad about this setup. Quite the opposite, I don't know what problems might arise so I'm suggesting you be sure you fully understand the issues...

Lots more words:

This is pretty interesting stuff. I don't have the knowledge to comment coherently so I'll just ask questions. Maybe some of our more sophisticated engineering types can comment...

trackside said:
more resistance = less voltage to the ECU = leaner mixture as the ECU thinks less air is entering the engine (so the theory goes).
Is that how the AMM works? Is this the hot wire AMM? This heats a wire which is cooled by the passing air. More air cools it more, less cools it less. The resistance of a wire usually increases with temperature. Does the AMM pass a monitored current through the wire which will vary according to the resistance {which supports the more resistance=less air hypothesis}?

How does an engine react to running lean? Presumably there's a tradeoff or we'd be all running as lean as possible. Is there damage done? It is only knock that causes damage?

On an APC system, knock will back off the boost. On an LPT, there's no knock sensor. What can this type do to avoid knock? Is it simply a matter of advancing {retarding?} the ignition? What about an EZK ignition system. This adjusts timing in response to knock, advancing it as far as possible and backing off if knock starts. Will running lean cause it to advance the timing too much and negate the benefit of the less fuel?

How is power / performance affected by mixture? What about emissions? Presumably CO will go down but, if the engine runs hotter, won't NOX emissions increase? When's the next MOT due? Have you checked the emissions?

Oooh! Questions, questions... See, I get only 280 miles on a tank in the 900i auto {EZK ignition} and I'm itching to pull out the ohmmeter...
 

·
Registered
Joined
·
883 Posts
Discussion Starter #13
I've got APC and a knock LED hooked up so I can monitor any leanout under boost. Lean running usualy also causes hotter running - which I initily had - but it wasn't overheating even in the hot spell we have had. The AMM sends a voltage to the ECU - the voltage increases with airflow therefore I presumed increasing resistance decreases the voltage - I may be totaly wrong but so far it's working ( I've asked several times on this forum for answers to this but got no response). Detailed workings of the Lucas EFI are pretty hard to find - even the official Saab workshop manual is thin with facts about it's actual functioning - it's just a fault diagnosis list. In my experience - if it pulls smoothly and strongly throughout the rev range, has good throttle response without hesitation, if it does not knock on boost, if it does not run hot and if it uses fuel at an economical rate - it's OK. This adjustment may well be specific to my particular setup? - 6 deg advanced timming over stock, no CAT etc so I can't say it will be worthwhile (or safe) on any other car - I'm just reporting my experience.
 

·
Banned
Joined
·
7,270 Posts
I think an engine that runs rich is a safe engine and that engines are always set to run rich just in case they fall out of tune between services. If you are able to keep an eye on your air/fuel ratio then you will probably find you can run leaner than stock. This lambda graph from the Emerald rolling road day last year shows that most of the cars run quite rich. The red reference line was one I added to illustrate what I had discovered to be a perfect lambda reading although because so many of the cars showed such a rich read-out it was argued that the line should have been lower on the graph. Nutcase (99 turbo) has fully manual mixture controls and no lambda sensor. He believed his mixture setting was the best setup. As you can see however, there are some lines running closer to what I had found to be the ideal setting, especially the newer clio although that is N/A so it would be eaiser for that car to keep an even mixture. We also speculated over the mixture changes. Ultimately it is just a bunch of lines on a graph and what might work well for one car might not be any good for another (say different climates or running conditions) This graph shows the difference between the cars tested on that rolling road on that day.

Something else I just noticed is the spike in Ylee's chart - guess that's when his humongous turbo starts hurling air into the combustion chambers!! :lol:



 

·
Registered
Joined
·
883 Posts
Discussion Starter #15 (Edited)
Thanks for that data graph - very interesting. My epic mileage on a tank was done with economical driving very much in mind - no hard acceleration and as near to 65-75 as possible. this means on my car it will have been using the Lambda to control mixture most of the time - which is odd as this should be independent of AMM signals. On the other hand if the lambda is constantly trimming a set fuel map that is rich it must be easier and more economical to do small or no adjustments to a map that is near to the 14.7? Or possibly again the bad battery was messing with the voltage signals and ECU? The only way to know is to reset the AMM back to stock and test again - I'm reluctant to change it now it's running so sweet.
 

·
Registered
Joined
·
9,170 Posts
I'm the 900 Aero with the blue line on that graph. It has since emerged that my fuel filter was partially blocked and restricting flow. The a/f ratio would very likely be richer now.
 

·
Moderator
Joined
·
4,637 Posts
Jezzadee said:
It has since emerged that my fuel filter was partially blocked and restricting flow. The a/f ratio would very likely be richer now.
Interesting point. The fuel pump has an internal relief valve set at 6 bar. If you're boosting 1.5 bar with a 3 bar FPR, that leaves a spare 1.5 bar to overcome restrictions in the fuel lines/filter. I reckon the filter would want to be fairly badly blocked {given its large cross section} to start reducing the pressure at the injectors...


ejenner said:
I think an engine that runs rich is a safe engine and that engines are always set to run rich just in case they fall out of tune between services. If you are able to keep an eye on your air/fuel ratio then you will probably find you can run leaner than stock.
So, how do you keep an eye on the ratio? I've no apc but I do have EZK on this car so there is a knock sensor. Is knock a reliable indicator of too lean? Is it possible to connect up a knock LED to the EZK system?
 

·
Registered
Joined
·
2,220 Posts
Yes, you could theoretically save fuel by leaning out the mixture.

The reason the mixture is unnecessarily rich has to do with emissions: Hydrocarbon and Carbon Monoxide emissions increase with mixtures slightly rich of stoichiometric, but nowhere near as much as Nitrous Oxides increase with mixtures even a tiny bit leaner than stoich.

This means that a richer mixture is more likely to be emissions-compliant; minimum emissions are a trade-off against maximum economy.

Also, the Classic 900 is no Honda Civic--it's a heavy car, and there's a fuel penalty for that, too.
 

·
Banned
Joined
·
7,270 Posts
The knock sensor is an electronic component so it is possible to use it's output for some sort of in-car gauge. I can't tell you how to do that but I'm sure someone has done it already. These figures were acquired using a lambda probe in the exhaust. You can use the same technology for tuning your own car - you can buy cheap kits from people like Gunsons. The other way to do it is by bolting a wide-band sensor into the downpipe. There is obviously an advantage to have the permanent install in that you can drive about and take readings while your on the road and log them on a laptop.

This is an example of a wide-band sensor installation.

 
1 - 20 of 26 Posts
Top