Micro controlled VNT project

[regcheeseman]

I am currently working on a VNT controller based around a micro processor.
At the moment I have the potential to monitor boost via a piezo cell, implement a multitude of rev dependent boost limits with a servo controlled VNT.
I also have a crank pickup to drive a standard revcounter and provide an engine speed signal.
I also have a throttle position sensor.

The problem I’m having right now concerns the algorithm/map/whatever, I’m measuring actual boost and then setting vane position dependant on revs and throttle position.

For example the engine is sat at 2000 rpm (the boost limit for this range is 4psi), the throttle is 75% open. Therefore I require 75% of the maximum boost and the servo should adjust to give a maximum 3psi.

I’m concerned that my original setup idea will overcomplicate the system and the power will suffer.

The more I look at this setup it seems wrong, I’m thinking that the throttle input is a redundant signal to measure and will in effect control boost by it’s original method i.e.  adding fuel

So by loosing the throttle position signal… following the same example the engine is sat at 2000 rpm (the boost limit for this range is 4psi), the throttle is 75% open. But the system doesn’t care where the throttle is and adjusts the turbo to deliver a max of 4 psi

However, I’m cruising along now, engine still sat at 200 and throttle barely open at 10% but the system will have shut the vanes right down to try and get 4psi. Is this a bad thing?

If I went back to my original idea the servo vanes would be wide open as the boost limit would be 10% of 4 psi i.e 0.4psi

I’m not sure which way to go and I also notice other people working on similar systems utilise a throttle position sensor, but I have been unable to discover how and why?

So my question is this, how should the control algorithm work?

[booneylander]

I would think,

Assume you are aiming for a max boost around 20psi.

(Values in table expressed as vanes % closed):

Boost      0    2    4    6    8    10    12    14    16    18    20
TPS     
0%         0    0    0    0    0     0      0      0      0      0      0
20%       20  20  20   20   20   20    20    20     20    10     5     
40%       40  40  40   40   40   40    40    30     25    20    15
60%       60  60  60   60   60   60    60    50     40    30    20
80%       80  80  80   80   80   80    80    70     60    50    40
100%    100 100 100 100 100 100   100   90     80    70    50

That way:

-Closed throttle, vanes open for less backpressure (coast for FE).
-As you apply more throttle, the vanes close down to spool the turbo.
-As boost approaches max, you open up the vanes to get as little backpressure as possible for the boost level.

[regcheeseman]

The idle position would be controlled by a small area of the code which keeps the vanes full open no matter what when the revs are below a given threshold.

I would see the accelerator position as the most often used signal that would govern the position of the vanes, with the maximum closed position being based on RPM and with the boost pressure overriding everything to open the vanes all the way regardless of the other positions.

So to confirm, vane position will be the inverse of throttle position, unless the boost limit has been reached in which case open the vanes.

[regcheeseman]

Great help guys, thanks very much.

The term throttle was a regional thing and not meant to be a physical reference to a throttle - The 'throttle' is the name for the foot pedal inside the car. Sorry for any confusion.

My original thinking for having some 'idle' code was a throw back to the first designs for the controller which as soon as the accelerator pedal was lift off idle, would slam the vanes full shut, then open them to regulate boost. Obviously not an ideal situation.

Re the boost spikes, surely my current setup is similar to a ball/spring type boost controller where the opening pressure is considerably larger than the pressure required to keep it open. It is prone to a large amount of hysteresis in it's operation.
Thanks for the heads-up on this issue, I'll try to add a degree or proportional control to the servo response pattern.

Booneylander, thanks for the table. It sure gives my project so extra new ideas. I was aiming for a max boost level depending on revs but you have gone for a TPS controlled max vane position. I haven't really got a feedback for actual vane position in the servo, but it could be implemented.

[regcheeseman]

Been juggling a few ideas about and I'm going to start the vnt control routine by checking a rev/boost limit, providing this isn't exceeded then the vane will be set with a map similar to the one suggested.

I will have a program port on the box to allow the map to be altered in situ

Thanks for your help - I'll keep the forum posted of any developments

[regcheeseman]

Thanks for the input, I notice you are referencing throttle (pedal) position versus revs whereas the map is throttle position versus boost, can I safely assume the boost is proportional to revs?
Will the vanes being too closed produce excessive gas temps or any other obvious signs
For the original table, would you prefer the 100% pedal position to read something like.....

Boost      0    2    4    6    8    10    12    14    16    18    20 - max limit
TPS     
100%    100  90   80  70   60   50   40    30    20    10     0


As for the 20psi limit, no thanks - I'm thinking more 30+ psi limit (1.6 GTD block, AAZ head, 10mm pump, 1Y inlet mani, VNT17)

[booneylander]

You'd have to experiment and see, I really don't think it's appropriate for us to say anything at all about how closed or not the vanes should be at some *point* (rpm/tps/boost) because the setup is so unique. The map I posted was just an "idea" of how I think you would want it to control.

Ideally, you'd need to sit on an eddy current dyno at max rpm @WOT with the vanes fully open, close them down until you hit 30psi, back them out until you just barely start to fall off 30psi, then go a tiny bit the other way again (so the wastegate is just starting to open) and record the servo position . Then 3900rpm do the same thing, 3800rpm and so on until you fall off boost (turbo can't spool even with vanes fully closed). Then fill out the rest of the map as "fractions" of that position, and depending on your taste (ie do you want to cruise at 30%TPS at 10lbs boost so that if you mash the go pedal you hit 30psi really fast, or do you want to cruise at 0psi boost at 50% TPS for fuel economy and only get on boost at higher TPS etc).

That would give you the best map for your setup. Anything else is just guesses in the dark. But I think the "logic" in the map I proposed is solid, rpm and boost are linked so not sure which is best to control for (might be rpm), you may want a 3D matrix?

Also, for the case of being at high rpm at little or no throttle and then mashing the pedal, because the vanes are already open, and need to close down, boost will come on long before the vanes get fully closed, and you'll get into a part of the map that keeps them open appropriately, so boost shouldn't spike, as the software is way faster to respond than the servo ever could be. Not sure if I'm stating that clearly, but hopefully you'll know what I mean.

[55eta]

Correct me if i am wrong but i was always under the impression that the whole purpose of the vnt was to obtain your max boost at low rpm by closing the vanes and then as exhaust flow increases with the rise in rpm the vanes open to maintain that level of boost without over boosting , by increasing the flow through the turbo by opening the vanes dose away with the need for a waste-gate .The biggest advantage off the O/E control system is that they will close the vanes {regardless of pedal position } if the boost drops due to extra load put on the engine i.e going up hill with your foot to the floor and rpm dropping .To sum it up the vnt is all about producing max boost at lowest rpm possible and maintaining it right through the rev range {without over boosting} regardless of load, its not about getting X amount psi at 20% throttle and X amount at 50% its all about creating your boost at low rpm and maintaining it without over boosting and the need for a waste-gate and that would lead me to believe that boost would be the key signal to the control not throttle position i.e if boost drops below your set parameters the vanes would close and if it exceeds your maximum parameters the vanes would open .I just wish i had the electronic s knowledge to do what your doing but i don't so i will just have to play with stone age diesels hope that is of some help though

[55eta]

Forgot to say on tdis the vnt is controlled by the n75 valve which is controlled by the ecu which takes it signal from the map sensor {manifold absolute pressure sensor}  i don't think throttle position comes in to it for the control of the vnt  so as far as i can see if you can set a max -min boost parameter e.g 16psi min 18psi max and control a n75 bleed valve to control the vane actuator you have cracked it all you need then is a vacuum supply for the n75 valve and if you achieve that i will be your first customer

[regcheeseman]

wow! Excellent ideas being kicked about!

I've been working with the values and rationilised them until I could form a simple arithmetic rule
which I further simplified with the addition of a 'tuning factor' a simple change results in a global map shift
Ignoring the minus figures which would be resolved to 0

Vane position Vp = Tp - Tf x Ba x Tp
                                    

TUNING FACTOR Tf=0.05
Ba = Boost Actual
Tp = Throttle position

   
                     Boost   
TPS   0   2   4   6   8   10   12   14   16   18   20   30
0.05                                       
0   0   0   0   0   0   0   0   0   0   0   0   0
10   10   9   8   7   6   5   4   3   2   1   0   -5
20   20   18   16   14   12   10   8   6   4   2   0   -10
30   30   27   24   21   18   15   12   9   6   3   0   -15
40   40   36   32   28   24   20   16   12   8   4   0   -20
50   50   45   40   35   30   25   20   15   10   5   0   -25
60   60   54   48   42   36   30   24   18   12   6   0   -30
70   70   63   56   49   42   35   28   21   14   7   0   -35
80   80   72   64   56   48   40   32   24   16   8   0   -40
90   90   81   72   63   54   45   36   27   18   9   0   -45
100   100   90   80   70   60   50   40   30   20   10   0   -50


What do you think? Fairly sensible figures?

I think if I add in a revs variable I'll produce a 3D map which my head/coding ability cannot cope with even if the processor could!

The revs limit variable will be applied in another section of code, it works by having five boost limits in five different rev bands, these limits are applied before the map above is even considered. Their values will overide any map values.

[regcheeseman]

damn the table looked fine in review, any way to control the way it appears?

[55eta]

Yes i understand tdis take signals from maps -mafs-mats-and throttle-pots to control fueling timing and i also just realized i misread your map/ graph i thought the % was pedal position i now see where you are coming from and i realize part of what i wrote sounded like it was written in stone but it was meant  more as a question than a statement and i cant get my head round the pedal/throttle potentiometer without a mass air flow meter which u obviously cant use with a mechanical pump or am i barking up the wrong tree with that

[55eta]

I would have thought boost pressure sensor{maps} gives you the p.s.i of a volume of air and not the actual VOLUME, a mass air flow sensor gives you the volume of air{cubic meters} before it is pressurized and a m3 of air requires the same amount of fuel added {air fuel ratio}weather at atmospheric pressure or 2 bar, to burn efficiently  and yes they  all play a part of producing /controlling boost /and fueling all combined with a ecu but manifold absolute pressure sensor{maps} is the only one that monitors actual boost P.S.I regardless of volume so you could end up having 10 m3 at 1500rpm and 10psi or 40 m3 at4000rpm and 10psi so the pressure is the same but volume is diffrent so the fuel requirements would be totally diffrent but the pedal positon could be the same {depending on load} with out a mafs and throttle pot and one with out the other is the bit i cant get my head round it would be like having a guard dog with no teeth?

[tomnik]

Sorry to jump in so late.
Since weeks I observe the actuator control pressure (vac) and boost in my TDI to tell my electronic buddy how it works. He might build the VTG controller for mine and his Benz.
It is clear for me that the throttle position (power demand) sets a certain boost level (not max. boost, until full pedal). Also lifting the foot sets the boost to a certain position and I can see the vac(= vane position) is going up (vanes are more and more closing) when speed/rpm go down. I also think that zero pedal is equal to a not zero set boost. The vac tries to keep boost low first then when actual boost is below that min boost the vane close to achieve that min boost (but rpm is not able to reach it). With this configuration the vane position is waiting for the next acceleration in the closed position.
Correct me if I am wrong or even better tell me which situation you need to "see" on my TDI to get further.

Tom

[tomnik]

waiting in the fully open position is the difference to the OE?
In my TDI I see full vac (vanes closed) at idle.
Has anybody a working mechanical solution? I would like to get ahead with the car while waiting for my buddy.

Tom