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

[CyFi6]

I have been having trouble controlling boost from the beginning, but after getting a quality boost controller and trying everything I still cant get rid of my boost spike. my mods are as follows:
57 trim CT26
2.5" intercooler/piping
full 3" exhaust
Apexi intake
Supporting fuel mods
Wastegate shimmed with 2 washers- Base boost of about 8PSI
Turbosmart Eboost Street electronic boost controller

I am aiming for a solid 16 PSI but cannot run 16 psi because spiking to 18 psi is putting me into a danger zone. If i MUST, I could probably adjust my tune to allow me to spike and not hurt anything, but I would rather have solid boost control.

As it sits now, boost will spike 1-2 psi and quickly come back down to a steady PSI. I have altered with the settings in the EBC as much as possible to where i should have no spike, but i still have a spike. I tried my Hallman pro manual boost controller and experience a spike as well as drop off with this setup.

My boost only line comes from the compressor housing, and the actuator line goes directly to the actuator. Hoses are tight and neither is over 12 inches, no leaks.

Running on base boost there are no spikes, but as I start raising my PSI the spike becomes more apparent. Here is a graph of base boost. X is sample # and Y is KPA absolute.

Here is a video of my boost gauge in 2nd gear where you can clearly see the spike. This is with the electronic boost controller. With the manual, the spike is still there but doesn't seem to drop quite as fast, but more of a gradual drop off.
http://smg.photobucket.com/albums/v313/cyfi/?action=view&current=MVI_2134.mp4

Here are some graphs of the spike while running the EBC (multiple boost settings)

Here are some spikes when using the Hallman pro manual boost controller (set around 15 psi)

Things I have checked:
Wastegate actuator moves smoothly
No leaks in WG diaphragm
1 or 2 small cracks around wastegate flapper opening
Vacuum lines tight and sealed 100%
No apparent boost leaks (have not done a boost leak test though)

I cant really think of anything else to check, please help me if I am missing something obvious. Any help is appreciated.

 

[IJ.]

Try a 2mm restrictor in the Boost line to the EBC.

 

[CyFi6]

They claim that there is already a restrictor built into it for quicker response, but I have no idea what size it is. Should i still give it a shot? Where do you think i can get one, you think a bead from a craft store for necklaces would do the trick?

 

[IJ.]

I make my own on the lathe, problem with a bead is it might turn in the hose and cause issues, depending on the hose size you use a Holly carb jet may do it.

The idea behind all this is to slow the response of the WG so the controller doesn't overshoot trying to correct, in the AVCR you could adjust the PID response curve to tune the overshoot to a managable level, not sure on the e-boost as I've never used one.

 

[CyFi6]

Oh ok, I thought a restriction was used to increase response though, as the air velocity will increase? Im probably over thinking it. In the Eboost there are 3 major settings i can adjust. They are
Set Point- Duty cycle % of the solenoid
Gate Pressure- PSI at which the solenoid begins to open and cycle
Sensitivity- How quick the controller will try to react to changes in boost

I have tried setting both gate pressure to 0 (will build boost slow however) and sensitivity to 0 (sensitivity to 0 will not allow the eboost to respond to changes in boost) and there is still a spike. After trying the manual controller too it seems its something unrelated to the actual controllers, though I cant be sure.

 

[IJ.]

Nah the restriction slows everything, The settings you have will be the normal P I and D just named differently, pid setting can be a bit of a black art and the restrictor just makes it easier as it dampens the swings, ALL pid devices need an overshoot/error so they work and they are self correcting, the settings just determine how far they overshoot and how fast they try to correct, you can get into a loop though and it will chase itself, yours sounds quite close with only a 2psi overshoot.

Each "system/engine" will be different as you need to match the engines characteristics to the controller, I've always used a restrictor even with the MoTeC Boost control as I said above it makes life a little easier, an 80 thou Holly Jet will do the trick if you don't have access to a lathe.

 

[CyFi6]

Ok thanks for the good info. I was looking through my garage and found this fitting in the assortment of things included with the boost controller. There is no mention of what it is for in the instructions but I measured it and the orifice is about 70 thousandths. I threw it in there and took it for a run, there was little to no change in the spike. I can try fiddling around with the settings more tomorrow but at this point I saw no change. Does it matter where in the line I install it? I put it right next to the rear fitting in the manifold which is dedicated to my boost controller and gauge right now (cruise control and heater aren't hooked up right now).

 

[IJ.]

That's the restrictor!

Now you have to change the settings, do you have any sort of fuel or overboost cut?

I usually put mine in the line close to the Solenoid.

I had one monumental 30+psi spike when I first played with the AVCR lucky for me the engine was built to run that hard.... :nono:

 

[CyFi6]

EEK that could be a recipe for disaster!

I am tuning with the MAFTPRO and the way my tune is now, it is relatively close to the stock fuel cut. The Turbosmart has a built in overboost protection, so that in the event of an overboost (whatever setting I input) it will cut off power to the solenoid (of course, that can't save me from a stuck solenoid or leak in my actuator lines).

When you say you put yours in the line close to the solenoid, I'm a little confused... Is this restrictor going inline with the control unit of the boost controller or in one of the wastegate actuator lines that are inline with the actual solenoid?

 

[IJ.]

Sorry gave you a bum steer on that coupler, it's just for connecting the actual gauge part of the controller, you'll need another restrictor in the line to port #3.

I have to ask WTF is that Gargoyle in your avatar???
(came damn close to infracting you for that thing )

 

[CyFi6]

I have to ask WTF is that Gargoyle in your avatar???
(came damn close to infracting you for that thing )

LOL

Sorry gave you a bum steer on that coupler, it's just for connecting the actual gauge part of the controller, you'll need another restrictor in the line to port #3.

What size restrictor for port 3?

 

[IJ.]

Damn that's FUGLY (I'm guessing "I Bought" would still hit it though.....)

Try a 2mm/80 thou and see how it goes.

Might have to go up/down to hit the sweet spot but you should be able to get it to hold +- .5 psi when set right.

Only time I'd expect you to have issues would be trying to run really low boost as the WG hole might not flow enough to limit the engine and it'll creep.

 

[CyFi6]

OK I'll try to get a hold of another restrictor somewhere or try the carb jet you mentioned. Common sense would tell me that a restrictor in the boost only line to the solenoid would even further delay the actuator from seeing boost pressure, creating a larger spike, is there something I'm missing here (not to mention a 2mm restrictor is extremely small compared to the size of the line)? I have no problem with holding boost once the spike is gone, the eboost street has a pretty cool feature that lets me correct for high RPM dropoff (has an RPM input) and it worked flawlessly, I have this turned off in the graphs I posted so there is some dropoff in those.

 

[IJ.]

What happens is You tell it 20 psi, the boost controller opens at your gate pressure so snaps it open due to the large ID hose, the controller then tries to maintain 20 psi but as I said earlier it has to have an error to chase (it's a "difference engine" meaning if xx is different to yy do zz, xx being target boost, yy actual boost and zz correction + or -) so depending on your settings it'll start to pulse the solenoid, if the frequency is too low it'll over boost for a few seconds till it catches it if too high it'll under boost and the controller will need to ramp up the duty some.

The idea of the restrictor is to slow ALL of this down so the adjustments aren't as critical, without the restrictor I've found boost control to be ultra fussy, the slightest change in anything will mean either over or under boost i.e. Altitude, ambient temp, and so on.

 

[CyFi6]

OK the restrictor seems to have helped some, though it still spikes about 1 psi which I can handle. I will give it some more fine tuning when its not 95 degrees outside...:-/ Turns out the size of the fitting in the back of the control box is the same as the restrictor size, so I just put the restrictor I had in the boost only line to the solenoid. Boost doesn't seem to quite hit 16 psi in second and drops off a bit, and 5th gear almost seems to overboost by about 1 to 1.5 psi, any idea why this is? I actually noticed this same issue without the restrictor as well (overboosting in high gear). I have a W58 with a 4.3 rear end, I can hit full boost in second with the right settings, but then it overboosts in other/higher gears. Is this something pretty normal?

 

[IJ.]

You always set max boost in 4th/5th, a lot of times the combination won't allow for full boost in lower gears due to not enough load on the turbo, I know the AVCR did gear sensing so you could tune each gear/boost ditto on the MoTeC, not sure with the TS though.

 

[CyFi6]

Theres the latest log, thats 1st through 5th in PSI and the bottom is RPM

Still can't seem to get a steady boost line through the rev range, Im not sure if its me being picky, if it actually can be better, or if its just the nature of a crappy internal gate. Any tips?
Also notice how only in 5th it overshoots pretty much through the whole gear compared to lower gears.

 

[IJ.]

Looks like it's trying to hold, Can you do a run without the controller connected and log it as a baseline?
(just a hose direct from the Turbo to the WG)

 

[CyFi6]

Ya I can do that, is it alright if I leave the restrictor in or would you prefer I remove it?

 

[IJ.]

Be better without it if you have a piece of hose handy.