Discussion
1.5 or 1.8
blitzracing said:
Let us know the answer. That adjustment controls not only the amount of fuel at tickover, but also makes a huge difference to the on / off throttle fueling amounts throughout the rev range, like a scaling factor for the output of the air flow meter.
im not really any wiser 
im told its within the normal range but its not as straight forward as turning the co screw to increase the voltage... i would need a co test thingy..I`ll wait to my next mot which is due soon.
Meatball said:
blitzracing said:
Let us know the answer. That adjustment controls not only the amount of fuel at tickover, but also makes a huge difference to the on / off throttle fueling amounts throughout the rev range, like a scaling factor for the output of the air flow meter.
im not really any wiser im told its within the normal range but its not as straight forward as turning the co screw to increase the voltage... i would need a co test thingy..I`ll wait to my next mot which is due soon.
in saying this mine where out by quite a bit and it as improved pick/up, this car as the HC chip installed.Edited by That Daddy on Tuesday 24th July 08:35
The problem you have is over 2500 rpm the lamda sensors dont have any effect, and even below this you only have + /- 10% tweek available to you. If the settings out you could exceed the available compensation, leading to poor running. I think to be spot on, you need to look at the Lambda probe output with an osciloscope. I would think you would be able to clearly see the change in waveform at the too rich and too lean ends of the pulse as you exceed the 10% compensation available as you tweek the control. This would give you two (screw) points of adjustments for too rich and too lean, that you could then take a mid point between the two so minimum compensation would be required. Hopefully this point would also tie in for the best fueling when the ECU goes open loop above 2500 rpm. Ive only however expermented on a non cat setup that runs best at 1 volt in my case (with a rechip as well). If however anyone fancies letting me loose with my 'scope on a cat system for some readings I would be most happy to help. Near Newbury, Berks.
At last Ive fitted a couple of Range Rover lambda probes to my non cat car to do some testing with a Chimeara 3.9 TVR ECU chip, and thanks to people from this forum who have helped so far.
Setting the base CO trim voltages.
On a non catalyst system the trim value makes a big difference to the idle mixture, (to the point you can foul your plugs, or generate very poor idle) and the problem is the published value for these settings will be affected by the fuel map being used and by the fuel pressure, so really can only be accurately set with a CO meter.
On the TVR 3.9 catalyst map with a standard fuel pressure I tried a range 0 - 2.5 volts on the AFM trim value, and the ECU held the mixture within the correct range, although the recommended 1.8 volts gives about the best lambda probe response. At no point during the adjustment did the probe(s) show .3 or .7 volts steady to show they had run out of range on too rich or lean, so the setting did not make a huge difference, and the ECU fuelling adjustments coped within this range. The lambda probe output responds quite slowly, that as Greenv8s says can be easily seen on an analogue voltmeter. I’ve been told by "experts" that the ECU goes open loop above 2500 rpm, but on switching to the catalyst map the fuelling constantly is tweaked between rich and lean THROUGHOUT the rev and load range, as you would expect with a constantly closed loop system. On the non cat map, my fuelling frequently dropped out of the range of my narrow gauge sensor and gauge less than 15:1- too lean. Problem is the non cat map appears to be the original Range Rover one, so it’s impossible to back to back testing against the TVR map. I can only conclude that either the TVR fuel mapping is incredibly accurate above 2500 rpm, or its still closed loop, and I’m being told porkies by the experts….. The car runs much better at low speed on the cat map. Both non cat and catalyst maps show similar fuelling at around 12:1 A/F under hard acceleration, although the cat map reduces the amount of fuel more readily after a few seconds.
Setting the base CO trim voltages.
On a non catalyst system the trim value makes a big difference to the idle mixture, (to the point you can foul your plugs, or generate very poor idle) and the problem is the published value for these settings will be affected by the fuel map being used and by the fuel pressure, so really can only be accurately set with a CO meter.
On the TVR 3.9 catalyst map with a standard fuel pressure I tried a range 0 - 2.5 volts on the AFM trim value, and the ECU held the mixture within the correct range, although the recommended 1.8 volts gives about the best lambda probe response. At no point during the adjustment did the probe(s) show .3 or .7 volts steady to show they had run out of range on too rich or lean, so the setting did not make a huge difference, and the ECU fuelling adjustments coped within this range. The lambda probe output responds quite slowly, that as Greenv8s says can be easily seen on an analogue voltmeter. I’ve been told by "experts" that the ECU goes open loop above 2500 rpm, but on switching to the catalyst map the fuelling constantly is tweaked between rich and lean THROUGHOUT the rev and load range, as you would expect with a constantly closed loop system. On the non cat map, my fuelling frequently dropped out of the range of my narrow gauge sensor and gauge less than 15:1- too lean. Problem is the non cat map appears to be the original Range Rover one, so it’s impossible to back to back testing against the TVR map. I can only conclude that either the TVR fuel mapping is incredibly accurate above 2500 rpm, or its still closed loop, and I’m being told porkies by the experts….. The car runs much better at low speed on the cat map. Both non cat and catalyst maps show similar fuelling at around 12:1 A/F under hard acceleration, although the cat map reduces the amount of fuel more readily after a few seconds.
blitzracing said:
If however anyone fancies letting me loose with my 'scope on a cat system for some readings I would be most happy to help. Near Newbury, Berks.
Didn't you ask for volunteers to have go at mapping a cat car a few months ago ??(i got out volunteered by someone closer, in newbury.)
The guy from Newbury never responded to my emails, so I gave up, and thought Id be better of messing about with my own car. I did go to the effort of building a 20 LED diplay to monitor the output voltages graphically, but need not have bothered for idle settings as an analogue meter works well enough. I will use it for full load testing however, as its easily viewed in the cockpit. Next thing to find out for SURE is each banks fuel tweeked on the fly, or just one setting for both banks, plus is it open or closed loop above 2500 rpm? Answers on a post card please
I did get an answer from Mark Adams that the map goes open loop between 3000 - 3500 rpm, or anything approaching full throttle. These values are preset in whatever fuel map you are using, and so are not written in stone. I dont yet know about the individual banks fueling, but am on the case by using an LED stereo audio power meter to display the Lamda outputs, one channel for each probe, so I can compare them side by side in real time. I think it must be able to tweek each bank individally, or why use 2 probes ?
spend said:
I seem to recall that in the past someone (probably MA) hinted that the dual lambdas just worked like a safety valve ie if too far out of sync then ECU assumes the worst. The ECU doesn't see each bank & lambda as 2 unique state machines.
Dave
It goes without saying that what Mark Adams says is right, but my understanding was that each bank is controlled by a separate feedback system.Dave
I know that both lambdas have simular voltage outputs, so that gives you two options.
The 2 lambda output values are averaged out by the ecu, and that then controls the amount of correction for each bank from the fixed part of the map, so the mixture will be the same each side. Without lamda feedback, the mixure is much leaner on one side on mine, and Im told this is normal so its very unlikely to be this system. More likely, its simply 2 closed loop systems adjusting the two banks. With the amount of correction available, it must constantly adjust a base setting to keep it in range. These ECU's arent half as dumb as people think!!
The 2 lambda output values are averaged out by the ecu, and that then controls the amount of correction for each bank from the fixed part of the map, so the mixture will be the same each side. Without lamda feedback, the mixure is much leaner on one side on mine, and Im told this is normal so its very unlikely to be this system. More likely, its simply 2 closed loop systems adjusting the two banks. With the amount of correction available, it must constantly adjust a base setting to keep it in range. These ECU's arent half as dumb as people think!!
Ok, one stereo audio power meter modified to read 0 -1.5v, strapped to my lambda outputs to I can see the left and right bank fueling side by side.......Right bank (as in offside)toggling lambda nicely. Left bank only toggling once in a while, tending to run slightly rich most of the time.(under any load / rpm range) Replaced lambda probe and ECU (with a new fuel map), still the same. Strange thing is if I switch to the non cat map, I can tweek both sides to to be within spec' (separately) so I know the probes and ECU mixture control (as in CO base setting) is good. I think (???) the left bank should be adjusting itself as the right, given the adjustment range available. Its possible I have a leaky injector, but the fault would be less apparent under full load, but its not. I need to try this setup on another car and see the results before I rip my injections to bits.....
The left and right banks are controlled independently, so it is possible to see different pulse widths between the two banks. The ECU learns these corrections as an adaptation over time.
In the event of failure of one of the sensors, the ECU will identify the remaining good one and use that for mixture control.
A significant difference between the banks does indicate an underlying problem, because they should be very close in normal operation.
Assuming there are no mechanical issues, any imbalance is usually caused by fuel injector spray problems. A poor spray pattern (i.e. jet instead of spray) can cause the appearance of a weak mixture on the affected cylinder. The Lucas injector is a particularly nasty specimen of the breed, and around 30% of the cars I see have injector problems.
The control range of the lambda sensors is +/- 22% which is vast in fuel terms. The lambda sensors are meant to trim the fuelling (from a reasonably accurate map), not do all the work. Once the corrections exceed 15% you will start to feel the action as you drive the car.
Altering the adjuster on the AFM does actually change the amount of compensation that is applied to the basic map, before correction. Therefore if you examine the exhaust gas whilst changing the adjuster, you will see no change. The correction exists as a software value only inside the program, so cannot normally be seen. The correct trim value for a lambda-controlled car is 1.8 Volts.
In the event of failure of one of the sensors, the ECU will identify the remaining good one and use that for mixture control.
A significant difference between the banks does indicate an underlying problem, because they should be very close in normal operation.
Assuming there are no mechanical issues, any imbalance is usually caused by fuel injector spray problems. A poor spray pattern (i.e. jet instead of spray) can cause the appearance of a weak mixture on the affected cylinder. The Lucas injector is a particularly nasty specimen of the breed, and around 30% of the cars I see have injector problems.
The control range of the lambda sensors is +/- 22% which is vast in fuel terms. The lambda sensors are meant to trim the fuelling (from a reasonably accurate map), not do all the work. Once the corrections exceed 15% you will start to feel the action as you drive the car.
Altering the adjuster on the AFM does actually change the amount of compensation that is applied to the basic map, before correction. Therefore if you examine the exhaust gas whilst changing the adjuster, you will see no change. The correction exists as a software value only inside the program, so cannot normally be seen. The correct trim value for a lambda-controlled car is 1.8 Volts.
Gassing Station | Chimaera | Top of Page | What's New | My Stuff