Please help me fix my boost spike! Pics and video included

[CyFi6]

Here, I am just going to upload the excel spreadsheet of the logs, I already inserted the graphs at the bottom of the data, but I figured the more info you have the better (if you need it).
Keep in mind I have the wastegate shimmed with 2 washers on each bolt (3-4mm IIRC)
I appreciate all the assistance IJ it is a lot of help!

View attachment Base Boost Logs.zip

 

[IJ.]

Be really cool if you could overlay them

Gumby method of flicking between the 2 shows an interesting trend that No controller is always climbing with controller is decreasing Vs RPM, that would make me believe the Controller is overshooting then over correcting trying to catch it.

I'd try a different frequency on the Valve first and log the results, as I said earlier PID tuning is a bit "black art" but having logging available make it 1000x easier!

 

[CyFi6]

Been messing around with excell il try overlaying them next time, just looks a bit cluttered.

That is just the problem though, I don't really know what else to change. When you see boost start to fall and rise up again, it is probably the high RPM correction factor working (it is enabled). With sensitivity at 0 or less than 5, and gate pressure at 0 or very low, no matter my setpoint duty cycle i should have the least amount of overshoot right? IE Sen: 0 GP: 0PSI it should build boost slow and level off at a max, but even with these at 0 when i start raising my setpoint it spikes regardless. Just doesn't make sense that the solenoid begins cycling at 0PSI of boost and still overshoots when boost comes on.

I tried to get some help from the TurboSmart guys themselves and they recommended i try a manual controller and see if the overshoot still happened, which it did, and that would tell me that "it is likely that the actuator cannot act on the exhaust flow acurately enough when transitioning from off boost to boost." Thoughts?

---------- Post added at 01:14 AM ---------- Previous post was at 01:07 AM ----------

Here is what i have found as far as changing GP, SEN, and SP

Factory SEN is 20, though i get large spikes with SEN set this high, I have to reduce it below about 6 or 7 to lessen the spike, the higher the SEN the worse the spike

GP doesn't have any affect on the spike (though it does affect spool time) until it is set too high. I found about 13PSI to be the max before the spike starts to get worse than it already is.

SP - As I raise SP (DC%) boost increases, but anything above about 10PSI boost it begins to overshoot/spike regardless of any of the other GP or SEN settings.

I have not found any combination of settings that will not induce a spike or overshoot (above about 10PSI).

 

[IJ.]

Tried it without the preload washers on the WG?

 

[CyFi6]

Yeah, in fact I only put the washers there in effort to combat this issue and get a more stable boost curve, though it really didn't seem to help or hurt anything.

 

[IJ.]

Just thinking back to my 3540r, tried lowering the gate pressure?

 

[CyFi6]

Cant get much lower than 0 :icon_bigg
The unit is super easy to adjust on the fly so I have tried an array of settings, gate pressure will affect spool time until I raise gate pressure up to about 13psi, any more than that and it begins spiking worse. Sensitivity too high results in an initial spike and an oscillation of boost after.
That's really why I am beginning to think its less of a controller issue and more a mechanical issue somewhere, I just don't know what could cause the issue, or if it is just inevitable with my setup.

 

[IJ.]

0 is bad as it delays boost/spool, the WG will start to open from .5 psi (depending on preload)

With the 3540r it was always a battle to get it preloaded enough to stop this happening and not too much as that caused creep up top due to the small internal gate it had.

Try googling PID settings there are some good tutorials out there that will help you understand the basics.

I have a MoTeC one here somewhere I could dig out if you have an e-mail addy?

 

[CyFi6]

As far as I know, I feel that I understand the settings as well as what they do pretty well, but the problem is they are not doing what they are supposed to do.

I know that 0 PSI gate pressure is bad for spool, but my immediate goal is to eliminate the initial spike. As of right now I could care less if boost drops off at high rpm, all I want to figure out first is how to eliminate the spike at the beginning of boost, I can adjust the other stuff later. Like I said i can't find any settings at all that will prevent me from spiking. Are you saying that more preload on the wastegate can help reduce a spike? If so I can try preloading the WG some more with either shims or a helper spring.

Il do some more research on this and check out how some other people tuned their controllers like you suggested, maybe someone has experienced something similar.

 

[IJ.]

It's all a bit of a juggling act due to how the actuators are configured, have a read through that PDF a couple of times, some of the settings are counterintuitive (at least they seemed that way to me)

More preload usually results in more creep up top, I found with internal gate turbo's I could have one or the other but never perfect control over all the gears so I always settled for perfect on the top end in the higher gears as I couldn't hook it up in the low gears anyway.

 

[CyFi6]

Thanks for the read, it was informative. One thing I am still confused about, never really understood- How does the control unit correct for changes above and below my "set point" if no actual target boost pressure is input? The only parameters the control box has are SetPoint, Sensitivity (gain), and GatePressure. Without knowing what actual pressure I am aiming for, how does the computer know how much to correct and in which direction? Considering every turbo/wastegate setup will react differently to the same SetPoint (one turbo might produce 15 psi at a setpoint of 50, and another produce 25 psi at a setpoint of 50) how does it know what my target is?

Unfortunately, only being able to adjust those 3 parameters, I am sort of limited as far as how much I can adjust it for my application. Somehow I need to alter sensitivity (gain or derivative component from the motec article) and set point to get the boost to come up solid without spiking or fluctuating (over correcting and oscillating) and stay level. The problem I am having though is that regardless of setpoint, too much sensitivity or gain results in a higher and higher initial spike (won't oscillate until my sensitivity is much too high- over about 35). Lowering the setpoint simply lowers my boost pressure away from my target of 16 psi, so I am really at a loss of how else to manipulate these two settings to avoid a spike.

 

[IJ.]

Never used an E-Boost but I thought you dialled up a number on the readout?

 

[CyFi6]

dialled up a number on the readout?

What do you mean? I have never input an actual desired PSI. Only setting to increase or decrease PSI is SetPoint. I have the Eboost Street, but even the high dollar TurboSmart boost controllers use the same inputs to set boost.

Heres all inputs as far as setting boost goes.
SetPoint
GatePressure
Sensitivity
RP1- RPM to commence correction
RP2- RPM to terminate correction (redline)
FAC- Correction Factor- % of correction

The last three data entries are for the high RPM dropoff option which is really just an extra feature that is not 100% necessary and can be added once I get the other things figured out. Post #17 had it active, but I have since turned it back off to make things less confusing.

If you dont mind, could you take a look at these quick videos? These are simple guidelines to setting up the boost from TurboSmart.
http://www.youtube.com/watch?v=iA4zpX5MzKc&feature=mfu_in_order&list=UL
http://www.youtube.com/watch?v=We3n92yKauM&feature=mfu_in_order&list=UL

Of course these are simplified, but that's how its supposed to work. Only problem is lowering GatePressure doesn't eliminate my spike like they say.

 

[IJ.]

LOL

What size restrictor for port 3?

Saw the 1st episode of this here.....

Ohhh my it's even uglier when it's moving/talking.. :nono:

 

[mecevans]

Have you ported your wastegate?

 

[CyFi6]

Saw the 1st episode of this here.....

Ohhh my it's even uglier when it's moving/talking.. :nono:

Lol, it is quite a horrific sight isn't it

Have you ported your wastegate?

No, waste gate is completely stock besides being shimmed.

 

[IJ.]

Lol, it is quite a horrific sight isn't it


No, waste gate is completely stock besides being shimmed.

Just waiting for a Japanese ship to sail by and Harpoon it

Man you guys have some bizarre laws over there watching this show, if some fat ugly slag grabbed me on my own property like that she'd end up with a steelcap boot hysterectomy...

 

[speedfreak]

i have the same problem, finally went to a mbc which helped but did not fix the problem. i used to use a greddy profec b spec 2 it helped to turn the gain higher than the setpoint and leave the sensitivity all the way down dont really need a solenoid flopping around in there complicating things, at least not yet. i believe the problem is really the wastegate being forced open by the exhaust pressure i also have my w/g shimmed with 5mm worth of washers and that helped hold pressure to redline alot, it also smoothed out the entire boost curve, i think im going to try 7mm worth of washers next time. Anyways im no IJ but it seems to me that the boost controller is trying to correct for the premature wastegate opening, it holds the pressure well until some of it escapes then trys to compensate but overdoes it and it makes it spike. just my two cents. would love to see a driftmotion ct26 adjustable wastegate actuator someday!