Am I right in saying that nearly all electronic boost controllers are not closed loop? From my understanding, the only aftermarket electronic boost control that is closed loop is the HKS EVC. By closed loop I mean you input a boost pressure and the unit does the work for you. Open loop is where you input a duty cycle to achieve an unknown boost pressure.

It's also true that all ECU based boost control systems are closed loop, correct?

Just wanted to clarify this point. It always seemed odd to me that many electronic boost controllers were not smart enough to work in closed loop. They all have map sensors so I never understood why they couldn't do it.

 

incorrect.

all electronic boost controllers are closed loop. if you have a MAP signal input, that is your feedback.

the gain you adjust is simply the gain for a PID controller (most likely anyway, which requires feedback). on a greddy profec B you have "load" and "gain" where load would be your P of the PID, and the "gain" would be the D of the PID. the profec B doesn't have much of a I because it tapers still, and doesn't seem to correct much at all.

others use more of a lookup table, like the apexi AVCR, which has a gain and a target boost for each rpm setting. but again, it acts like a PID, with the target boost compared against the feedback of the actual boost, giving you a delta boost. if you watch the AVCR it will self adjust the gain to hit the target boost. this is done by attempting to zero out the delta (indicitive of the integral gain being adjusted on the fly). i couldn't tell you exactly how they are doing this, but i know they are definately closed loop controllers.

 

one way to really notice the closed loop element of the PID controller on the profec B is to set the gain way high (very high Derivitive gain). the boost will overshoot, and oscillate bad because it is overcorrecting over and over (i.e. an underdamped system). if you have the gain set very low, it will be extremely slow to spool (an overdamped system). then when you find the magic gain setting where it spools quickly, and doesn't oscillate, you have a critically damped system and your gains are set correctly

 

Boy was I wrong then!

I've just seen so much setting of duty cycles in boost controllers like AVC-R, profec-b, blitz, etc... none of them allow you to just put in a boost target, which made me think they must not be closed loop. But I guess they just need some help?

And with the profec, I had the old version a long time ago, when I would turn the dial for a low or high boost setting, you mean internally that dial was sweeping through a range of 0-30 psi or so? The unit I'm referring to is this one:

http://www.advancedjapautoimports.co...controller.jpg

With the new profec, do you actually just input a boost target or a duty cycle?

And the AVC-R, correct me if I'm wrong, will auto-learn, but once it has learnt, it creates a map of duty cycle for rpm. Is it the case that it then follows that duty cycle and make corrections while following that baseline?

 

i've never seen a boost controller where you can simply set the target boost value and go (seems like that would be the most ideal boost controller though ).

reason they don't make one like that is because some cars might use a 16lb wastegate spring, and if you dial in 25psi, it may need very high proportional gain values to reach the 25. then you upgrade the wastegate spring to a 24psi, and those same proportional gains would send your boost soaring above 25psi. this is why they almost always have some type of gain (or duty) to adjust.

if your adjusting duty (AVCR), you ball park it, and it will use the feed back to keep it in that range. its still not a straight rpm lookup table spitting out duty cycle though, it still is closed loop in that it uses the MAP reference to keep the actual boost near the target boost. try setting the desired boost to what you want and set the duty as low as it will go, and then one pull at a time, bring the duty up and above what it should be to see what happens. youll see the internal controller at work trying to keep the boost at your target. i'm sure i've got some logs demonstrating this, but it would take me a while to find them (and even then, i don't record the EBC settings, just the boost curves)


our stock ecu controlling a 3 port once modified is more open loop then those EBC's. the stock ecu has nothing but a duty cycle lookup table based off RPM. it does have the boost error correction which would be your feedback (i.e. closed loop), but often times you simply zero it out to make boost hold better (making it open loop). additionally, it has the load limit which kills boost if exceeded, temperature "turn on" for the boost, and a target load (Which is what the error correction tries to close the loop on)