Engine Management Books/resources?
Discussion
If you want hardcore then you need this
Internal Combustion Engine Fundamentals
John Benjamin Heywood
http://www.amazon.com/Internal-Combustion-Fundamen...
Also check out Prof Gordon Blair's stuff
Not directly engine management, but engine operation and control.
Engine management is the means you put the above into action.
Internal Combustion Engine Fundamentals
John Benjamin Heywood
http://www.amazon.com/Internal-Combustion-Fundamen...
Also check out Prof Gordon Blair's stuff
Not directly engine management, but engine operation and control.
Engine management is the means you put the above into action.
ringram said:
If you want hardcore then you need this
Internal Combustion Engine Fundamentals
John Benjamin Heywood
Thanks ringram, actually I already have this. It's interesting to compare it with Ricardo's "High Speed Internal Combustion Engines" to see how relatively littel things have changes in the last 70 years or so.Internal Combustion Engine Fundamentals
John Benjamin Heywood
But I'm looking more specifically for info into the theory and practice of engine management through ECUs - especially MBE ecus. MAybe MBE do a proper manual, certainly the online help in the EasyMap software isn't much help
I'm not sure an MBE book is going to be very long, they are failry simple EMS's, with basic control structures and a small number of calibratable parameters etc. (Compared to say a typical production car BOSCH EDC or MED system).
OEM systems have moved to a high degree of freedom "model based" architechture (which uses torque as the primary control parameter ("torque based"), in which the basic physics of the control laws are replicated by the ems (usually multidimensional equations driven via a mixture of "Look up tables" and calibratable constants)
Bosch publish a large number of guides to their EMS systems, and these cover most of the control requirements etc, under their "yellow jacket" series
( http://www.amazon.co.uk/s/?ie=UTF8&keywords=bo... and http://www.tec-stop.co.uk/body_bosch1.html )
In all cases, the first step is to understand the relationships between the control inputs (air, fuel, ignition timing etc) and the engines response. Once you understand how an engine works then the control of it becomes much simpler! (and Heywood is a great place to start ;-)
Editted to add: In EMS systems nowdays, the diagnostic requirements now consume something like 50% of the processor load and 80% of the Calibration data RAM/ROM ! OBD system calibration and optimisation is an entire subject in itself......
OEM systems have moved to a high degree of freedom "model based" architechture (which uses torque as the primary control parameter ("torque based"), in which the basic physics of the control laws are replicated by the ems (usually multidimensional equations driven via a mixture of "Look up tables" and calibratable constants)
Bosch publish a large number of guides to their EMS systems, and these cover most of the control requirements etc, under their "yellow jacket" series
( http://www.amazon.co.uk/s/?ie=UTF8&keywords=bo... and http://www.tec-stop.co.uk/body_bosch1.html )
In all cases, the first step is to understand the relationships between the control inputs (air, fuel, ignition timing etc) and the engines response. Once you understand how an engine works then the control of it becomes much simpler! (and Heywood is a great place to start ;-)
Editted to add: In EMS systems nowdays, the diagnostic requirements now consume something like 50% of the processor load and 80% of the Calibration data RAM/ROM ! OBD system calibration and optimisation is an entire subject in itself......
Edited by anonymous-user on Tuesday 28th December 14:50
Edited by anonymous-user on Tuesday 28th December 14:54
Max_Torque said:
I'm not sure an MBE book is going to be very long, they are failry simple EMS's, with basic control structures and a small number of calibratable parameters etc. tnote]Edited by Max_Torque on Tuesday 28th December 14:50[/footnote]
Blimey, and there was me thinking it was pretty complicated! I'm starting to wonder if I'm trying to find out information which is commercially confidential - hence the lasck of detail availableEdited by Max_Torque on Tuesday 28th December 14:54
Specifically, I'm trying to understand how the throttle actuator responds to the available inputs. I have a pedal position sensor and associated voltage/site map as it's a fly by wire setup, plus all the usual revs, airflow,etc. I can log all the parameters as I drive, and see all the various traces for everything going into and out of the ECU
What I don't understand is precisely how/why the TP actuator reacts to this information. All I can find out so far is relatively general statements like "the ECU uses the inputs from the sensors to determine the optimum throttle opening".
My system was set up by a Very Well-known and Respected Person Indeed, and a bloody good job he has made of it too, it makes stonking amounts of power over the stock system, whilst retaining pretty reasonable driveability. The problem I have is that there are a few final tweaks I'd like to be made in the area of off-throttle response, but I can't afford any more of his time to play about with it. So my theory is that if I learn much more myself about what might be going on, I can give him much more detailed information about what I'd like done - thus hopefully he could get the car on his RR and sort out the last few glitches in a sensible amount of time before my wallet runs dry.
You dont need to understand the how or why to achieve that.
You tell the person setting it up exactly what you arent happy with, and let him sort it !!
he is the one who has defined the DBW parameters, so anything you need to know, ask him !
There is no generic information on how DBW should be setup. It can be different for every application depending on the drivers needs, or driving style, and the power delivery of the engine.
You tell the person setting it up exactly what you arent happy with, and let him sort it !!
he is the one who has defined the DBW parameters, so anything you need to know, ask him !
There is no generic information on how DBW should be setup. It can be different for every application depending on the drivers needs, or driving style, and the power delivery of the engine.
stevieturbo said:
You dont need to understand the how or why to achieve that.
You tell the person setting it up exactly what you arent happy with, and let him sort it !!
Yes, I realise that I don't _need_ to understand, but I want to, I find the whole subject fascinating and want to learn more about it! You tell the person setting it up exactly what you arent happy with, and let him sort it !!
People tell me that it's possible to set up DBW systems to give just as snappy a throttle response as cable systems, and I'd like to know how it's done. Am I asking too much? There must be some way of learning this stuff, surely? Or does everyone just start from scratch and figure it out for themselves?
I'd be amazed if the DBW throttle was running anything other than a PID controller (probably with feed forward pre-emption based on target position) Effectively, the system takes the voltage output from the pedal position sensor(s) (PPS) (usually 2 seperate channels, with inverse characteristics, so you can diagnose failures by simple channel summation). This voltage is then converted to a pedal demand (say 0 to 100%, probably with a "ratchet" adaption for the mim and max position).
Depending upon the operating mode (idle, idledrive, drive, overrun etc) the throttle position target will be calculated (generally for simple systems by a linear (or broadly linear) look up table (allowing some lead or lag to the PPS). This base throttle target will then be modified as required by things like the idle speed control strategy, or the "dashpot" demand during trailing throttle etc.
The Throttle plate position (TPS) is measured (again with 2 sensors for redundancy /safety, and this measured position compared to the "target" position and the "error" calculated. The Throttle motor (std DC brushed motor) will be PWM controlled (to effectively vary the supply current and hence the motors torque). The throttle position PID controller will use the position error and the integral of that error to generate the PI segments of the control output. This will probably be updated at approx a 300Hz to 1KHz rate. When the throttle plate is moving fast, a derivative term is used to limit the magnitude of the output gain to help prevent overshoot (The D term).
System response will be mainly controlled by 2 main calibration systems:
1) the overall gain of the PID controller; significant target position errors can exist if the PID values are poorly optimised
2) the control strategy that calculates the target position; depending on the strategy used and how it is calibrated this is likely to be the biggest decider on engine response feel etc
OEM EMS systems have at least 1 further control loop on throttle angle, where the torque control structure generates a "mass airflow" target, and a physical "throttle plate choking" characteristic is used to convert this to a throttle angle target.
For the MBE system i would be suprised if it uses anything other than an open loop throttle angle target based primarily on PPS demand.
If you can log whats going on, first step is to log "actual throttle angle" vs "target throttle angle" and check that the PID control is delivering precise and repeatable throttle plate positioning. If it is, then you will have to look deeper into the higher level "target" values to see how they are being generated etc
Depending upon the operating mode (idle, idledrive, drive, overrun etc) the throttle position target will be calculated (generally for simple systems by a linear (or broadly linear) look up table (allowing some lead or lag to the PPS). This base throttle target will then be modified as required by things like the idle speed control strategy, or the "dashpot" demand during trailing throttle etc.
The Throttle plate position (TPS) is measured (again with 2 sensors for redundancy /safety, and this measured position compared to the "target" position and the "error" calculated. The Throttle motor (std DC brushed motor) will be PWM controlled (to effectively vary the supply current and hence the motors torque). The throttle position PID controller will use the position error and the integral of that error to generate the PI segments of the control output. This will probably be updated at approx a 300Hz to 1KHz rate. When the throttle plate is moving fast, a derivative term is used to limit the magnitude of the output gain to help prevent overshoot (The D term).
System response will be mainly controlled by 2 main calibration systems:
1) the overall gain of the PID controller; significant target position errors can exist if the PID values are poorly optimised
2) the control strategy that calculates the target position; depending on the strategy used and how it is calibrated this is likely to be the biggest decider on engine response feel etc
OEM EMS systems have at least 1 further control loop on throttle angle, where the torque control structure generates a "mass airflow" target, and a physical "throttle plate choking" characteristic is used to convert this to a throttle angle target.
For the MBE system i would be suprised if it uses anything other than an open loop throttle angle target based primarily on PPS demand.
If you can log whats going on, first step is to log "actual throttle angle" vs "target throttle angle" and check that the PID control is delivering precise and repeatable throttle plate positioning. If it is, then you will have to look deeper into the higher level "target" values to see how they are being generated etc
For the record, a properly optimised DBW system can outperform a cable system! The system i developed for my car actually pre-empts a WOT throttle demand from the driver (by looking at rate of change of PPS) and manages to get the throttle plate fully open approx 30ms BEFORE the drivers foot hits the boards.......... (system gets from 0 to 90deg throttle angle in <50ms (compared to approx 80 to 100ms measured by my foot ;-)
(regarding your throttle off issue, remember that the base engine mapping makes a significant diffence to the negative torque magnitude as well (ignition angle, fuelling etc)
(regarding your throttle off issue, remember that the base engine mapping makes a significant diffence to the negative torque magnitude as well (ignition angle, fuelling etc)
One concept, worth remembering (and unusually confusing!) is that of "effective wide open throttle angle" , where at low engine rpm, the thottle angle at which the throttle plate is no longer choked is very much less than the geometrical wide open angle.
Intake mass airflow is proportional to engine rpm, so throttle plate (or plates) that are sized to be only a small restriction at peak power rpm, will not need to be fully open at lower rpm to deliver full torque.
For some DBW systems, effective WOT is used instead of true geometrical WOT. (mainly to linearise engine torque output with pedal position across the rpm range, but also to allow faster interaction for traction control dsc etc (if throttle plate angle is only just above the choke point, you don't need to move it very far to start controlling airflow)
Hence, you may find that at say 2000rpm, the TPS only reaches say 40% even though the PPS is 100%
Most drivers familiar with cable throttles find this linearistation of engine torque to pedal angle "un sporty" For a cable system, say 30% pedal postion would be full torque at say 1500rpm, but only 30% torque at 7000rpm. Whereas with a BDW system using effective WOT, engine torque would be linearised to 30% at all engine speeds. As a result most DBW system have a degree of "torque gain" calbrated into the system at low engine speeds to put some of the classic "cable throttle" feel into the system.
Intake mass airflow is proportional to engine rpm, so throttle plate (or plates) that are sized to be only a small restriction at peak power rpm, will not need to be fully open at lower rpm to deliver full torque.
For some DBW systems, effective WOT is used instead of true geometrical WOT. (mainly to linearise engine torque output with pedal position across the rpm range, but also to allow faster interaction for traction control dsc etc (if throttle plate angle is only just above the choke point, you don't need to move it very far to start controlling airflow)
Hence, you may find that at say 2000rpm, the TPS only reaches say 40% even though the PPS is 100%
Most drivers familiar with cable throttles find this linearistation of engine torque to pedal angle "un sporty" For a cable system, say 30% pedal postion would be full torque at say 1500rpm, but only 30% torque at 7000rpm. Whereas with a BDW system using effective WOT, engine torque would be linearised to 30% at all engine speeds. As a result most DBW system have a degree of "torque gain" calbrated into the system at low engine speeds to put some of the classic "cable throttle" feel into the system.
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff