ECU - closed loop operations - how fast can they react?
Discussion
Been doing lots of tinkering recently - built myself a air/fuel ratio indicator that runs off the standard (narrowband) lambda sensor, with an LED display. Obviously, narrow band sensor cannot provide an accurate enough signal to show the differencs between 13:1 and 14:1 etc - but - the display does show you quite well how the car reacts from a steady throttle, then to a WOT, then on overrun etc, flicking back and forth between full rich, stoich, lean, etc etc.
(This is all building up to me possibly buying DIY PNP Megasquirt kit and building at home - I'm getting my head round things and furthering my knowledge first)
Bear with me on this - My question is - at a steady cruise with a fully warmed up engine, the ECU should be running in closed loop mode, to maximise power/economy, and hovering as close to dammit to whatever figure it decides, right?
Now, I understand how a closed loop operation/system works, so in reality the values you get out will go up, down, up down, etc, the 'mean' value being what you need. The higher the 'resolution' (not the right word?) and the faster the system reacts, the close the peaks and troughs will be to the mean value (E.G. in this case, ideal Stoichiometric combustion ratio).
My car - (Mk1 MX5) seems to be running at around 1 adjsutment per second. The rate at which the adjustments are made to the mixture, and thus the rate at which the AFR hovers above and below target, seems to be about 1 hertz, or maybe even slower.
The ECU in my car is known to be pretty basic, and its late 80's electronics I guess.
Still though - I am really surprised by this. it seems incredibly slow to react? Is this normal? What causes it?
I would expect that the hold up must be
- The time take for exhuast gases to reach the lambda (minimal, they must be moving opretty fast!)
- The rate the lambda can react? Are they slow to react? Its only a teeny tiny voltage change in the output that is measured?
At 4krpm, thats 60 odd crank revolutions per second, 30 odd 'squirts' of the injectors. seems like a big gap between each adjustment? How fast can injectors and things react?
What are more modern cars like?
A faster reacting closed loop system has got to be better than this, surely?
At 4krpm, thats 60 odd crank revolutions per second, 30 odd 'squirts' of the injectors. seems like a big gap between each adjustment? How fast can injectors and things react?
This fascinates me at the moment, I'm knowledge hungry!
(This is all building up to me possibly buying DIY PNP Megasquirt kit and building at home - I'm getting my head round things and furthering my knowledge first)
Bear with me on this - My question is - at a steady cruise with a fully warmed up engine, the ECU should be running in closed loop mode, to maximise power/economy, and hovering as close to dammit to whatever figure it decides, right?
Now, I understand how a closed loop operation/system works, so in reality the values you get out will go up, down, up down, etc, the 'mean' value being what you need. The higher the 'resolution' (not the right word?) and the faster the system reacts, the close the peaks and troughs will be to the mean value (E.G. in this case, ideal Stoichiometric combustion ratio).
My car - (Mk1 MX5) seems to be running at around 1 adjsutment per second. The rate at which the adjustments are made to the mixture, and thus the rate at which the AFR hovers above and below target, seems to be about 1 hertz, or maybe even slower.
The ECU in my car is known to be pretty basic, and its late 80's electronics I guess.
Still though - I am really surprised by this. it seems incredibly slow to react? Is this normal? What causes it?
I would expect that the hold up must be
- The time take for exhuast gases to reach the lambda (minimal, they must be moving opretty fast!)
- The rate the lambda can react? Are they slow to react? Its only a teeny tiny voltage change in the output that is measured?
At 4krpm, thats 60 odd crank revolutions per second, 30 odd 'squirts' of the injectors. seems like a big gap between each adjustment? How fast can injectors and things react?
What are more modern cars like?
A faster reacting closed loop system has got to be better than this, surely?
At 4krpm, thats 60 odd crank revolutions per second, 30 odd 'squirts' of the injectors. seems like a big gap between each adjustment? How fast can injectors and things react?
This fascinates me at the moment, I'm knowledge hungry!
I'm pretty sure it must be proper, sequential injection - theres timing sensors on the cams, it lectronically advances the ignition timing as revs rise, etc etc, so I would assume the injector timing is linked to this? Still learning about it allthough. Its not a completely basic system.
As for the display, no idea, I'm speaking to the people who designed it aswell to see what they suggest! Its purely an electronic device - I would have thought that was a pretty slow reaction time ? Again - still learning abotu the electronics side of it all too!
Its all very interesting though.
As for the display, no idea, I'm speaking to the people who designed it aswell to see what they suggest! Its purely an electronic device - I would have thought that was a pretty slow reaction time ? Again - still learning abotu the electronics side of it all too!
Its all very interesting though.
Edited by snotrag on Thursday 24th June 12:00
Gassing Station | General Gassing [Archive] | Top of Page | What's New | My Stuff