Co2 Boost control

[Ryan.g]

Been asked a few times before if we can control boost like the AMS1000 Boost controller where you target a set capsule pressure.

Well a friends car had it plumbed this way today and asked me to create some custom maps to control it, as well as link into the Wastegate control strategy.

Posting here in case useful to anyone else

See Output Functions

p.s Make sure you have Wastegate Control set as Run Strategy without Pin

[MReilly]

Ryan,

Thanks for sharing, always wondered how you would control a setup like that!

I don't suppose you have any datalogs you could share, or even screen grabs of interesting snippets if it's secret?

I'd be interested to see the pressure in the top of the gate during a full load run, and how you have to manipulate it to keep boost constant?

[stevieturbo]

Often they will set a "boost" ( dome pressure ) ramp and as said, target dome pressure and not even reference to manifold pressure as such, which is odd. But seemingly more consistent.
But of course, these are purely 1/4 or 1/8th guys mostly using it, with autos, a small rpm range etc etc

Also used with a very weak or no spring, so offering a very low boost pressure as well as very high.

I'd imagine still referencing any ramps to manifold pressure would still make sense though

[crice]

Ryan, is this just closed loop control of a capsule pressure target or does the PI control of capsule pressure work in addition to the primary (or primary and secondary in the case of dual MAP & PRP) closed loop boost control targeting MAP, turbo speed or PRP?

I have a rough idea how it works on an MQ12 but have never seen it implemented on another ECU. Like Stevie suggests, they have 2d table called wastegate duty to pressure transfer function which (as far as I can tell) converts the resultant WG duty total from the primary open loom control to a capsule pressure target. So to my mind the capsule pressure PI is applied *after* the primary closed loop?
But I've also hear it described that the capsule pressure becomes the 'base' map for the main MAP or turbo speed control?! so not entirely sure.

[stevieturbo]

Because often it is drag use only, they use a very simple either speed or time based boost ramp.

And when I say boost ramp....it's a dome pressure ramp. As said, they often seem to ignore manifold pressure ( although obviously the two go hand in hand, but they don't specifically target manifold pressure ) and target a dome pressure ramp instead.

They claim this is more consistent, and whether speed/time/distance based, they can plot out the entire run to manage boost over that distance right from launch which may want lower boost, ramping up as grip allows.

Everyone would need to work out what works best or easiest for them for their application.

[crice]

Hi Stevie, it's not only drag racing, rallying too.
It could be useful anywhere you need a big range of boost pressures, very accurate control and to overcome high EMAP.
Only targeting dome pressure to me is a waste of time, it's essentially a glorified open loop strategy at that point. If the base WG duty is just being adjusted (closed loop) to follow a dome pressure target you may have slightly better 'positional' control of the poppet for any given duty but as you point out this bears no relationship to what is going on in terms of MAP, turbo speed or PRP.
If you look at the MQ12 strategy, I'm quite sure it combines the capsule pressure PI with the usual closed loop PID options. I've not looked at if/how Bosch do it.
The way I interpret the Cosworth strategy is take the WG duty total (basically your base WG duty + all of your compensations including closed loop PID) then convert that to a dome pressure target and apply a PI closed loop on top of it to keep the dome pressure matching the transfer function table.
I do struggle to get my head around exactly how this helps over just having the PWM duty adjusted in closed loop to meet your primary MAP, turbo speed or PRP target, but apparently is does? If anyone can suggest how/why I'd love to know.

This is of great interest to me at the moment along with having DBW control MAP - it would be awesome if you could do it with Syvecs.

Can Syvecs do primary and secondary closed loop boost control? i.e. a target offset to the primary control loop if you go outside the secondary limits? basically dual switching (eg MAP & PRP or turbo rpm and PRP)?

Will there be any future firmware updates for S6plus range?

[RICE RACING]

I do struggle to get my head around exactly how this helps over just having the PWM duty adjusted in closed loop to meet your primary MAP, turbo speed or PRP target, but apparently is does? If anyone can suggest how/why I'd love to know.

How it helps is the antiphase with increasing MAP can have a run away boost effect for the same level of duty so the control is not as accurate, especially on cars where TIP or more correctly MAP/TIP is not increasing in linear fashion with RPM after MAP peaks (stays flat for example). So the idea of an antiphase pressure sensor is to aid in making a programmable force air spring which then can help the control be more consistent gives more stable mechanical control and the software feedback is more effective.

Yes you can do this in LR/Syvecs

However keep in mind for say a car with road orientation you do actually get very high TIP and TEP and instead of needing to run this or a idealized flat duty on the phase side and doing say bleed on control on the anti phase (depends on how you valves are set up) there is not much need to use this strategy complexity, cause you actually need to have the run away effect to counter the adverse conditions when TIP itself runs away.

End result is without this you can have very good control small errors and huge range, I personally only tested up to just under 4000mB MAP but these are 'road cars' using common fuel 100bhp to 120bhp/lt/bar anywhere from 2lt to 3.2lt capacity, you can work out the power but in cars from 900kg to 1500kg its stupidly fast! Every car we run anitphase and phase, and most end up with maximum possible valve duty settings at 9000rpm on either side of the capsule so you reach the limits of what you can do with manifold pressure alone say on a 1 bar mechanical spring. Which is why you need to then resort to a different method Ryan posted about here, compressed gas and no more limitations since you don't rely on MAP. Other way would be to make a better wastegate with a much larger area on the capsule which then can counter the issues of excessive TIP acting against the valve seat pushing it open.....

little side note: 25 years ago I designed some springs for a customer, wire diameter, number of coils to make his gates do the pressure he wanted. None of this is that hard, if I had to just rely on MAP I would make my own capsule, balance the area required to work with the available MAP range, and also select the WG valve area to work with the TIP maximum and put the wastegate near the end of the scroll on the turbine housing, locate the valves near the capsule and use right size ones along with lines of appropriate internal diameter and you would be shocked how much control range you would have, no need for bullshit compressed gas bottle in the car!

[crice]

How it helps is the antiphase with increasing MAP can have a run away boost effect for the same level of duty so the control is not as accurate, especially on cars where TIP or more correctly MAP/TIP is not increasing in linear fashion with RPM after MAP peaks (stays flat for example). So the idea of an antiphase pressure sensor is to aid in making a programmable force air spring which then can help the control be more consistent gives more stable mechanical control and the software feedback is more effective.

So you're referring to normal phase/antiphase here or fill/empty? By antiphase pressure sensor do you mean capsule (top port) pressure? Confused because isn't antiphase usually bottom port?
Does having a regulated pressure source (co2 or compressor/accumulator) not combat the effect you speak of in itself?
In normal phase/antiphase you'd adjust your phase linearization to help combat the runaway effect.
I also struggle to see what a fill/empty setup does that a single 3 port doesn't? A single 3 port will apply pressure to the diaphragm or vent it.
The fill empty installation pictured is an apparent contradiction of http://www.syvecs.com/forum/viewtopic.php?f=14&t=292
With the fill solenoid de-energised, the NC port is holding back the boost or co2 pressure. With the fill solenoid energised the NC port of the empty solenoid is trying to contain the pressure in the top port.

However keep in mind for say a car with road orientation you do actually get very high TIP and TEP and instead of needing to run this or a idealized flat duty on the phase side and doing say bleed on control on the anti phase (depends on how you valves are set up) there is not much need to use this strategy complexity, cause you actually need to have the run away effect to counter the adverse conditions when TIP itself runs away.

TIP = turbine inlet pressure
TEP = ???
In a phase/antiphase setup are both solenoids not bleed on (I struggle to understand bleed on or off because surely whether the solenoid is applying pressure or bleeding it depends on the duty at the time)?
I'm talking about a restricted rally car with a load of boost low down and about 0.5 bar gauge by the end.

little side note: 25 years ago I designed some springs for a customer, wire diameter, number of coils to make his gates do the pressure he wanted. None of this is that hard, if I had to just rely on MAP I would make my own capsule, balance the area required to work with the available MAP range, and also select the WG valve area to work with the TIP maximum and put the wastegate near the end of the scroll on the turbine housing, locate the valves near the capsule and use right size ones along with lines of appropriate internal diameter and you would be shocked how much control range you would have, no need for bullshit compressed gas bottle in the car!

There must be something in it for WRC teams to do it. The weight and complexity of accumulator and/or compressor must be worth it as 'boxes were hydraulic. If you have compressor already for paddleshift then it makes little oddds.

[stevieturbo]

I don't think you could use a normal phase/antiphase setup here where the two are directly linked, otherwise it would go through huge amounts of co2.

AFAIK, the fill/vent valves need independent control, even if both are still related to each other.

With phase/antiphase setup, there would likely be many scenarios where both valves would be open at the same time, which might seem pointless and wasteful of co2 ?

For me, if I was using it with normal boosted air, normal "Phase" would be the bottom port to blow the gate open, "antiphase" would be in the top port blowing the gate shut.
Wouldnt really matter either way as long as the rest of the control setup is good though.

[crice]

Yeah, understand you wouldn't use phase antiphase with CO2.

I'd never use CO2 anyway, I'd use a compressor and/or accumulator.

As I understand the fill empty valves still have to have a linear phase and antiphase?

For me, if I was using it with normal boosted air, normal "Phase" would be the bottom port to blow the gate open, "antiphase" would be in the top port blowing the gate shut.
Wouldn't really matter either way as long as the rest of the control setup is good though.

Ah OK, I was going off this
But actually I think that image is wrong as you want a duty of 100 to give max boost and a duty of 0 to give low boost (presumably with the phase being the master) like you say, it probably doesn't make a lot of odds but could cock up your closed loop function if the wrong way round.
It also says plenum as pressure source which is 'unconventional'.