aaronc7

iTrader: (8)

Hey guys, I have been tinkering with CL boost the past week, it's 80-90% there, but there are some situations I'm still running into overboost issues. Rundown of details of my setup:

DIYPNP on firmware v3.2.5
99 Mazda Miata, GTX2867R with Tial MVS 38m ext. wastegate (5psi rated spring)
DIYautotune solenoid, plumbed to both port on the wastegate, like this (bottom diagram): http://www.diyautotune.com/images/pr...te_control.png

With boost control off, it holds a steady ~5.5psi to redline. Open loop boost control all checked out good, 100 = spring pressure, I can't remember the specifics, but lowering the values in the OL table increased boost little by little. I needed less than 50% duty cycle to hit my 225kpa target.

Moved onto CL boost, using the MS3 docs as my guideline, which I believe still totally applies to MS2 3.2.5, with 100 duty meaning 0 actual duty cycle = maximum wastegate opened. I did the tuning at around 4k rpm, per the guide.... I used 4000-4500 rpm as my starting point and found P and I values that worked well. At that point I thought I was done and good to do, but I ran into an odd issue when starting the "gear pull" from a low RPM- it would overshoot really badly. Below are screenshots of my setup and a datalog. In the datalog, I do a low RPM start, and you can see the boost duty does not increase quick enough, it overshoots badly and I hit fuel cut ~245kpa. Then right after, I show you how I went about tuning per the guide and how it works relatively well, hits target quickly and holds. I don't do it in the datalog, but it will hold steady till redline just fine.

Is this an indication that I need larger I values? I also have a problem of overboosting if I go from say 70% throttle in boost, then 100% real quick, it just does not react quick enough. Open to any and all suggestions. Looking to get it squared away before an event tomorrow. If I can't get it sorted, I'll probably just end up doing open loop, which is fine, but I do miss out on some low end/mid range power.