For those of you that have been looking into a bit of tuning or need an easy to find reference point for understanding how the Bosch boost control works I drafted this. Although some of the function names may not match your particular ecu, it is the way in which the Bosch boost control developed so should explain a little more what you are looking at on the information on the net. Credit to the Bosch engineers who developed it. BOSCH BOOST CONTROL: Bypass valve for exhaust gas serves as actuator for boost pressure i.e. wastegate. The manipulated variable for the bypass valve is generated by a PID controller which generates a proportional component, a differential component and an integral component for the manipulated variable. It defines for the integral component a limit value which is derived from operating parameters of the engine. A device that changes the turbine geometry can also be used as the actuator. There is a none linear correlation between the boost pressure generated and the manipulated variable for the actuator which can be ignored as long as the working point of the boost controller shifts only within very narrow limits, in which case the variable can be regarded as almost linear. If, however, the working range of the boost controller is large instead of small, i.e. a large variation range for the manipulated variable is required, then with a greater displacement of the working point on the one hand the controller becomes too slow and on the other hand overshoots occur in the system. With the wide range of engines and outputs, boost pressure controllers are designed for wide working ranges. Due to this fact, there needs to be a way in which this can be achieved whilst not allowing slowing or overshoots in the system. The manipulated variable for an actuator which acts on the exhaust flow guided through the turbine, or one or more other variables constituting the manipulated variable, is/are transformed in a characteristics map into values such that, after transformation, there exists an at least approximated linear correlation between the manipulated variable and the controlled variable i.e. the boost pressure. This action yields a linear control curve which allows rapid and stable boost pressure control regardless of the location of the working point. The manipulated variable can be formed from a proportional component and/or a differential component and/or an integral component. It is advantageous to define for the integral component a limit value which is ascertained from a basic value that depends on multiple operating parameters of the engine and from a correction value superimposed thereupon which is determined adaptively as a function of engine speed, multiple speed ranges being defined. Instead of the manipulated variable itself, it is also possible for the proportional component, the differential component, the integral component and/or the limit value for the integral component to be transformed into a characteristics map. PARTS: Throttle valve and sensor for sensing the opening angle of the throttle valve, pressure sensor for sensing actual boost pressure pvdk. A rotation speed sensor for sensing engine speed nmot on the engine. Bypass valve on turbine activated via a spring loaded pressure capsule with an electro pneumatic timing valve. A further controller, detailed below, which receives as input signals the throttle valve opening angle a, the measured actual boost pressure pvdk, and the engine speed nmot, generates a manipulated variable ldtv, for the bypass valve. The manipulated variable ldtv, in the form of a pulse-width modulated signal, controls the electropneumatic timing valve, which for it's part generates the pressure for the spring loaded pressure capsule, which in turn acts on the bypass valve. A reference boost pressure plsol, is read out from the characteristics map KFLDPS as a function of the engine speed nmot and the throttle valve position a. In addition, the actual boost pressure pvdk is measured using a pressure sensor front of the throttle valve. The difference between the reference boost pressure plsol and the actual boost pressure, pvdk, is determined at a node V1. This difference is referred to as the system deviation lde. If the condition B_ldr for activation of the boost pressure control sytem exists, a switch, S1, is set to the output node V1 so that the difference between the reference boost pressure plsoll and the actual boost pressure pvdk is present at the output switch S1 as the system deviation lde. If the boost control system is not active, i.e. if the condition B_ldr does not exist, switch S1 is set to 0 and therefore the system deviation lde is 0. A threshold value decider SE1 applies a logical 1 to input S or an SR flip-flop FF if the system deviation lde exceeds a threshold UMDYLDR. Input R of SR FF is connected to the output of a comparator K1. K1 delivers a logical 1 if the system deviation lde is less than or rqual to 0.0. Under these conditions a logical 1 is present at output Q of SR FF if the deviation lde exceeds the threshold UMDYLDR i.e. if a transition is occuring from steady state to dynamic operation. If a logical 1 is preset at input R of SR FF i.e. if the system deviation lde is less than 0 the actual boost pressure is greater than the reference boost pressure FF is then reset and a logical 0 is present at it's output Q. The output signal B_lddy at output Q of FF indicated whether dynamic operation logical 1 - or steady state operation logical 0 - is present. In function block R1, a proportional control perameter ldrkp, a differential control perameter ldrkd and an integral parameter ldrki are ascertained as a function of the operating condition B_lddy and the engine speed nmot. Forming the product of the proportional controller parameter ldrkp with the system deviation lde in muliplier V2 yields a proportional component ldptv for the manipulated variable ldtv of the turbocharger. A differential component ldrdtv of the manipulated variable ldtv results from forming the product, in multiplier V3, of the differential control perameter ldrkd and the offset between the current system deviation lde and the system deviation lde 1-1 ascertained one cycle -approx 50ms earlier. The difference between the current system deviation lde 1-1 is determined in node V4. A delay element VZ1 furnishes the system deviation lde 1-1 delayed by one cycle. The integral component lditv of manipulated variable ldtv is formed by an integrator INT which calculates the product of the integral control parameter ldrki and the system deviation lde1-1, and superimposes that product on the integral component lditv 1-1 determined in the previous cycle. Lastly the proportional component ldptv, the differential component ldrdtv and the integral component lditv are added at node V5, yielding the manipulated variable ldtv for a bypass valve of the turbocharger. The integral component lditv has an upper limit imposed on it in order to prevent overshoots in the boost pressure control system. The limit value ldimx for the integral component lditv is ascertained in circuit block R2, specifically as a function of the system deviation lde, the integral component lditv, the reference boost pressure plsoll, the engine speed nmot, and the ratio vrlsol between the refernce charge and the max charge of the cylinders. Functional block R1 contains three families of characteristic curves LDRQ1DY, LDTQ1ST and LDRQ2DY depending on engine speed nmot. If the condition B_iddy for dynamic operation is present, the integral controller perameter ldrki from characteristic curve LDRQ1DY for dynamic operation is switched through to the output by switch S2. The differential controller paramter ldrkd is switched through by switch S3, from characteristic curveLDRQ2DY to the ouput. The proportional controller parameter ldrkp is created by taking the difference, at node V6, between a fixed value LDRQ0D that is switched by a switch S4 to node V6 and the differential controller perameter ldrkd. If the condition B_lddy for dynamic operation is not present, but instead the machine is in steady stae mode, then the integral controller parameter ldrki is taken from the characteristic curve LDRQ1ST; switch S2 is now correspondingly set to the characteristic curve LDRQ1ST. The differential controller perameter ldrkd is set via switch S3 to 0.0 and the proportional controller parameter ldrkp is set by switch S4 to a fixed value LDRQ0S/ The fixed values LDRQ0D, LDRQ0S and the characteristic curves LDRQ1LDY, LDRQ1ST and LDRQ2DY are applied, by way of experiments on the engine test stand, so that boost regulation is optimized in the dynamic and steady state modes. Functional block R2, which derives the limit value ldimx for the integral component lditv from the engine speed nmot, the reference boost pressure plsol, the system deviation lde, the ratio vrlsol between the reference charge and max charge of the cylinders, and the integral component lditv of the manipulated variable. The limit value ldimx is made up of a basic value ldimxr and a correction value ldimxak superimposed on it at node V8. In addition, another predefined value LDDIMX can be added to the limit value ldimx at node Vnine. This value LDDIMX corresponds to a small fraction, approx. 0-5% of the limit value ldimx, which ensures that the value in no circumstances falls below that small value. If the current integral component is greater than the limit value without the value LDDIMX, which represents a safety clearance, then the boost pressure can be spontaneously controlled even without raising the limit value, provided that the boost pressure deviation to be controlled does not require values greater than LDDIMX. A limiting stage BG1 limits the limit value ldimx to a predefinable value TVLDMX which corresponds, for example, to ninety fiver percent of the duty factor of the manipulated variable for the boost pressure control system. The current correction value ldimxak for the limit value ldimx appears at the output of a summing unit SU. In this summing unit SU, the correction value present at its input 1 is, under certain confitions, either increased or decreased in steps. The following conditions must be met in order for a stepwise decrease in the correction value to be performed in the summing unit SU: The boost control system must be active, i.e. B_ldr must be present; and the current limit value ldimx must not lie at the upper or lower end of limiting stage BG1. Both data are present at the inputs of an AND gate AN1, which delivers a logical 1 to a further AND gate AN2 if these two conditions are met. A further condition is that the absolute value of the system deviation lde must be less than a threshold LDEIA. For this, the system deviation lde is conveyed to an absolute value generator BB and then to a threshold value decider SE2, which delivers at its output a logical 1 to AND gate AN2 if the absolute value of the system deviation lde is below the threshold LDEIA. This threshold LDEIA is almost zero. In addition, a check is made in a threshold value decider SE3 as to whether the ratio VRLSOL between the reference charge and maximum charge pf the cylinders lies above a threshold LDRVL. If so, the engine is operating at full load, and threshold value decider SE3 delivers a logical 1 to an input of AND gate AN2. The last condition to be met would be that the integral conponent lditv be less than the limit value ldimx. A comparator K2 accordingly compares the integral component lditv of the manipulated variable and the limit value ldimx prior to node Vnine. A logical 1 appears at the output of comparator K2 if the integral component lditv is greater than the limit value ldimxr. The output signal of comparator K2 passes via an inverter NOT to an input of AND gate AN2. A logical 1 is thus applied to this input of AND gate AN2 if the integral component lditv is less than the limit value ldimx. If all these conditions are met a logical 1 is present at the output of AN gate AN2. This condition B_ldimxn for negative stepwise follow-up of the correction value in summing unit SU is delayed in a delay element VZ2 by a fixed debounce time TLDIAN and sent to a switch S5 and an OR gate OR1. If the condition B_ldimnx for negative stepwise follow up of the limit value exists, switch S5 connects input 4 of summing unit SU to a read only memory SP1 in which the increment LDDIAN for negative follow-up of the limit value is stored. If the condition B_ldimxn is not met corresponding to a logical 0 at the output of AND gate AN2, switch S5 then switches over to a memory SP2 in which the increment LDDIAP for positive follow-up of the limit value is strored. The following three conditions must be met for stepwise positive follow-up of the limit value; As was already the case for negative stepwise follow-up, a logical 1 must be present, as described above, at the output of AND gate AN1. In addition, the system deviation lde must be greater than zero; even a tiny deviation from zero is sufficent. A threshold value decider SE4 generates a logical 1 at it's output if this condition is met. Lastley, the current integral component lditv of the manipulated variable must be greater than the current limit value ldimx. As already described previously, this condition is checked with comparator K2. The output of K2 as well as the output of threshold value decider SE4 and the output of AND gate AN1 are applied to an AND gate AN3. A logical 1 is present at it's output if the three conditions cited above are met. The output signal of AN3 the condition B_ldimxp for stepwise positive follow-up of the correction value is sent through a delay element VZ3 whose delay time is equal to a debounce time that is taken from characteristic curve TLDIAPN dependent on the engine speed nmot. The condition B_ldimxn for negative stepwise follow-up of the limit value, and the condition B_ldimxp for positive stepwise follow-up, are both applied to the inputs of OR gate OR1. It's output signal, which is present at input 2 of summing unit SU, signals to summing unit SU whether a positive or negative stepwise follow-up of the limit value present at its input 1 is to be performed. The correction value ldimxak present at the output of summing unit SU is also sent to an input 5 of a functional block AS, in which an adaption of the correction value is accomplished. This adaptation is performed only if on the one hand full-load operation of the machine is present, and on the other hand if the condition for a positive or negative stepwise follow-up of the correction value is met. A datum regarding full-load operation can be picked off at the output of threshold value decider SE3. The info as to whether positive or negative stepwise follow-up of the correction value is being accomplished can be derived from the output signal of OR gate OR1. Both the output signal of threshold value decider SE3 and the output signal of OR gate OR1 are conveyed to the inputs of an AND gate AN4. Of these two conditions are met, the output signal B_ldimxa of AND gate AN4 is a logical 1. The condition B_ldimxa for adaption of the correction value is present at input 6 of functional block AS. Whenever the condition B_ldimxa = 1 is met, the current value of the summing unit SU is transferred into a corresponding memory cell of functional block AS, in which a plurality of values reproducing an adaption characteristic curve are stored. The interpolation points stldea for adaption of the correction value in functional block AS are supplied by functional block R3. R3 also delivers a datum B_stldw concerning interpolation point changes. Either the adapted correction value ldimxa from the output of functional block AS, or an adapted correction value ldimxaa in which sudden changes occuring in the negative direction have been limited to a minimum value, is conveyed to an input 1 of summing unit SU to for the correction value ldimxak. The decision between the adapted correction value ldimxa and the limited adapted correction value ldimxa is made via a switch S6. Switch S6 switches to a non-limited adapted correction value ldimxa as activation of the boost pressure control begins, i.e. immediately after the apperance of a rising edge of condition B_ldr for boost pressure control. The rising edge of the signal B_ldr is detected by a flip-flop AF. Otherwise switch S6 is located in the other position, and sends limited adapted correction value ldimxaa to input 1 of summing unit SU. An input 3 of summing unit SU receives from the output of an OR gate OR2, the information as to whether a rising edge of the boost pressure activation signal B_ldr is present, or whether the signal B_stldw is signalling an interpolation point change in funct block R3. The limited adapted correction value ldimxaa is formed as follows. In a node V10, the current correction value ldimxak output by summing unit SU is subtracted from the adapted correction value ldimxa present at the output of functional block AS. The difference signal ldimxad is conveyed to a limiting stage BG2. Limiting stage BG2 limits sudden negative changes in the difference signal ldimxad to a predefined limit value LDMXNN. The limited difference signal ldimxab at the output of limiting stage BG2 is added in node V11, back to the current correction value ldimxak, so that from that, the limited adapted correction value ldimxaa is ultimately created. A control characteristic curve would therefore show the dependence of the controlled variable, the boost pressure pvdk, on the manipulated variable ldtv. The curve would normally have a nonlinear profile caused mainly by the actuator, having an electropneumatic timing value, a spring loaded pressure capsule activated thereby, and the bypass valve actuated by the latter. Because of its nonlinearity, the curve has different slopes at working points A1 and A2 located far apart from one another. If the controller were set, for example, to point A1, then a change in the manipulated variable by a value of /\ldtv would cause a boost pressure change /\pvdkl of 40 millibars. If a working point shift to A2 then occurs, that same change /\ldtv in the manipulated variable would cause a much greater change in the boost pressure by a value of /\pvdk2 of approx. 220 millibars. In the event of this shift, an overboost of approx. 180 millibars would thus occur in the boost pressure regulation system. This undesirable effect can be avoided by transforming the nonlinear characteristic curve a, into a linear characteristic curve b. With curve b, a change in the manipulated variable ldtv by a value /\ldtv would cause the same change in boost pressure. Linearization of the control characteristic curve can be achieved by the following action: The manipulated variable ldtv at the output of node V5 is sent to a characteristic map KFLD. In this characteristic map KFLD, the manipulated variable ascertained by the controller is transformed, for each possible working point, into a value such that a linear correlation ultimately exists between the transformed values of the manipulated variable ldtv and the boost pressure pvdk. The transformation values derived during application of the controller from the known nonlinear characteristic curve a are stored in the characteristics map KFLD so that during normal operation of the controller, each calculated value of the manipulated variable can have a corresponding transformed value assigned to it. Instead of the characteristics map KFLD for transformation of the manipulated variable ldtv, it is also possible to transform the proportional component ldptv resulting in the manipulated variable ldtv in a characteristics map KFPT, and/or the differential component ldrdtv resulting in a characteristics map KFDT, and/or the integral component lditv in a characteristics map KFIT. All characteristics maps KFPT, KFDT, KFIT can also be combined in a single characteristics map. In addition to these maps, the the map KFLD for the resulting manipulated variable ldtv can also be present. A further alternative is to transform the maximum value ldimx for the integral component lditv in a characteristics map KFMX. The maps KFLD, KFPT, KFDT, KFIT, KFMX can be proivided alone or in combination with others; in any event they are to be applied in such a way that ultimately an at least approximately linear correlation exists between the manipulated variable ldtv and the boost pressure pvdk. Hopefully that should at least make the diagrams and info on the net a bit easier to understand!
Thanks for posting The way it is written it too technical for me as a casual reader. lol That`s written by an engineer and someone familar with the terminology, expecting it to be read by someone with the same knowledge.
The above writing looks to have been cut out from a Bosch a strategy guide which can be found on the net. Without the logic and flow diagram behind it and how this function interacts with the rest of the ECU torque coordination, it will be quite confusing. Most of this stuff can be viewed on nefmoto.com, where RBPE as been also seen.
I feel like the above has actually made boost control more confusing than just looking at the picture from which it came from. In much more simple terms... Accelerator pedal request > Torque Request Torque (load) Request > Boost setpoint for given load Boost setpoint > initial commanded wastegate position Boost is measured and then wastegate position adjusted to match measured boost with the boost setpoint.
These are not "cut" from anything on Nefmoto as far as I am aware, on this particular info I spent about three weeks going through the patent offices of the UK, USA and Germany (also the Japanese added patents of a very similar nature at the same time, which is why I presume the patents have never been officially granted to one particular company as far as I am aware as many are using fundamentally similar techniques! If you want to spend weeks looking into it yourselves through the numerous patents and offices out there on the subject, then be my guest!). As per the norm for me, I then simultaneously read and wrote it up at the same time (easiest way for me to learn and I then have the file to reference), omitting the aspects I thought were irrelevant, which being coding and/or references was very little! I did, however, lose a lot of info on a number of occassions due to pc's/laptops playing up and although I have been putting them in the clouds now, I thought it best to get a thread I can reference whenever I want on the net and it may or may not help others! Anyone can learn the FS diagrams as they are simple Physics/Electronics diagrams to some but may be hard to understand to most (which is what the vast majority of you usually say), therefore I have added further information and may continue to do so for me or others to reference in future. Would you prefer I put something simple on instead, like this took me a few minutes (as you can tell), but I did it through the FS ages ago and makes more sense for people to understand rather than picking up the FS and thinking "wtf does that diagram mean!". So, if you have ever picked up an FS and thought "what the hell is that all about" the patent info gives a bit of info on that front, add it to the basic 20vt diagram above plus the electronics/physics/maths info and there is a lot more on one thread you can learn rather than going through lot's of forums, threads, FS pages, patents etc. This may look far more complex but it is actually the fundamental basics and if you learn this, it actually becomes easier to understand things like the FS as that is what said diagrams are based on. You may have different evolutionary aspects of your particular ecu, for example the introduction of homog/strat fuelling on later ecu's but they all share similar principles. That way, you can read your ecu, see straight away where the codewords are grouped together and the flip-flops etc without the need for anything, provided, of course, you know what your SSP says about what systems you have. Anyway, to make sense of the first post, this is some of the information you may or may not like to learn about what is going on. Plenty of other info on the systems out there, but if you ever want to write a turbo file from scratch for your VR6 then understanding these principles is best! http://en.wikipedia.org/wiki/Proportionality_(mathematics)#Direct_proportionality http://en.wikipedia.org/wiki/Flip-flop_(electronics) http://www.electronics-tutorials.ws/combination/comb_7.html http://en.wikipedia.org/wiki/Gray_code http://en.wikipedia.org/wiki/Sequential_logic http://en.wikipedia.org/wiki/Combinational_logic http://en.wikipedia.org/wiki/Phase_response http://en.wikipedia.org/wiki/Sine_wave http://en.wikipedia.org/wiki/Linear_system http://en.wikipedia.org/wiki/Nonlinear_system http://www.controleng.com/single-ar...-tuning/010e177f99bb831fc647e8b8975a073a.html http://en.wikipedia.org/wiki/Stratified_charge_engine If you are interested then I will continue adding links, adding things I have wrote on certain systems/ecu's, screen shots etc
The boost control feature in a Bosch E-Gas system works as m1keh described at a high level. It certainly seems to work that way in most of the 1.8T stuff that has been fettled with. I am afraid the ME7 controlled R32 Engines with ECU (022906032**) or 12v VR6 on cable M3.8.1 (021906256) DO NOT have the "Lader Druck Reglung" (LDR) or turbo pressure control functions active/available in the software as they are naturally aspirated engines. Nether does the ADR or AGN engines from the 20v NASP. Most of the so called turbo maps for NASP 6 cylinders tend to be bodged! If you broke down the LDR feature to describe how each of those stages that m1keh suggested work, then you are on to something. Probably best to keep it simple though and bear in mind most would not have the correct means to read, edit, check-sum and write files to an from a modern ECU. PS neither was I trying to be defamatory by the way. Just saying.
Yeah I have played with 50,000+ functions and maps on the ME 7.1.1.'s over the years, I am well aware that LDR controls are not there in the oem's. In fact you could probably build a decent FI file for your R32 using just my photobucket album! They are mainly screenshots of workings out though, the more valuable and correct stuff I keep elsewhere for obvious reasons. When you say "bodged" boost control functions what do you mean? Do you mean, as an example here - this is an incorrectly defined map and values and many do that? It's probably a bit like the M3.8 turbo file someone did I posted which is not fully correct when I looked through it but some "tuners" are using to map peoples cars here in the UK and other places it seems; (in the VR6 forum if you want the link). Or do you mean something else? My master file has the full 196 functions/maps solely for LDR for BDE/BFH tunes, I am doing another one and there may be a touch more than 200 but I'll see. I'm actually playing with the Porsche stuff at the moment and their quad ignition maps With the on/off switches etc, these tend to be grouped together in the ME systems at least so you can spot them if you know what to look for, it's this kind of thing I was aiming to get at rather than a particular system, there's loads of info on that but not much on the electronics aspects which I thought might help some who think "what does FF mean?". I'll put my little boost control modded drawing I did up when I find it and place it in the bucket if you want some easy to read references to the 20vt/ME7 stuff then.
Nice to find some 'controls' background reading on this site. Although probably quite niche reading, please keep it coming. Jon
