compuw22c

Yes, I'm going to start another one of those threads... because all other information is so scattered and rarely mentions firmware version number so you're left guessing.

As most of us know, EBC changed a few ways on 1.4. They added lower limit delta (awesome), made this setup-basic-advanced mode thing, and also as I understand it changed the type of PID control (raising/lowering P got reversed?). So I'm trying to get my head around it. I'll list how I understand you're supposed to tune it:

1. Set setup mode, tune as though it's open loop with lower limit delta running using the bias table
2. Switch to Basic mode, move the slider around high as you can but where it doesn't oscillate, and you should be done. (Supposedly the slider ignores values in the advanced pid boxes, but I have my doubts)
3. (Optional) Tune PID the following way (starting with 100-0-0)
3a. Set P at 100 and slowly lower until you overshoot by a little bit (not sure what a "little bit" is, 1psi, 2psi, 1kpa, 2kpa?) on an initial stomp in 2nd (not possible without tire spin) or 3rd (my choice)
3b. Raise I slowly 1 number at a time until you maintain your target boost to redline
3c. Add D to dampen oscillations which I shouldn't have with a 2560r

So now I'll explain my setup. Stock 1.8L swap engine in a 93 (no forged internals or special compression pistons), FMII no electronics kit (2560r), internal wastegate that's known for its boost creep with my, 3" exhaust (FM).

Running wastegate gives me an intended 7psi with boost creep as RPMs rise up to 10psi (which I'm fine with).

Now here's how this process has actually gone:
1. Set setup mode, tune as though it's open loop with lower limit delta running using the bias table

I can get to 10psi with about 20 duty cycle, about 21 seems to push me into overboost, but when CL is running it will go higher to 22-24 so who the hell knows... I'm guessing it may need to toggle back and forth between maybe 20-22, which is what CL should do if it needs to to maintain the target

2. Switch to Basic mode, move the slider around high as you can but where it doesn't oscillate, and you should be done. (Supposedly the slider ignores values in the advanced pid boxes, but I have my doubts)

Here's where things get interesting. I'd LIKE to use my bias table the way it is (tops is 20% duty cycle) where I'm getting near my target at around 10psi without hitting overboost, see what it does, and use those values to populate my table. However, it doesn't really work out. It seems when I simply tried to use the slider in basic mode it's still taking into account my values in the PID boxes below. My results before changing the values in these boxes were that it would essentially do nothing. Moving the slider had no change in reaction as far as I could tell. Almost like it was still in setup mode.

What should I set these to to tune the slider correctly?

3. (Optional) Tune PID the following way (starting with 100-0-0)
3a. Set P at 100 and slowly lower until you overshoot by a little bit (not sure what a "little bit" is, 1psi, 2psi, 1kpa, 2kpa?) on an initial stomp in 2nd (not possible without tire spin) or 3rd (my choice)
3b. Raise I slowly 1 number at a time until you maintain your target boost to redline
3c. Add D to dampen oscillations which I shouldn't have with a 2560r

This **** is damn near impossible to do without a helper and/or a dyno, at least without being at least close. I started at 100-0-0 I'm now at 75-15-0 and it at least hits (passes) my target... but the fact that I tried to do this in the prescribed order and it didn't work in that order leaves me with an annoyance that a) it wasn't done right and b) that it still isn't working correctly.

A few more notes, for a while now I've been running my lower limit delta at 200 because I've been fearing overshoots (I thought that the thing spooled up too damn fast that it was already headed for overshoot before PID even kicked in) and now I've lowered it to 30 and it does seem to be better.

I feel like I need help getting the slider set correctly and then the rest I should be able to figure out. Also if someone with a similar setup wants to share their settings I'd be extremely grateful. I know it would be no excuse for real tuning, but getting me close could be very helpful. Also attached my last log and MSQ (I may have been playing with settings while driving so pay attention to the last few full throttle pulls)

Would also be nice for us to have an updated write up for ebc tuning on 1.4 with all the changes that have gone on in the last few years...

Thanks!

LOG file
https://onedrive.live.com/redir?resi...int=file%2cmsl

MSQ Attached



EDIT: Adding log jpg and also to ask, how the hell can I have -1 cats? No cats I can understand, but -1 cats? Do I need to go visit my crazy neighbor down the street and steal one?

aidandj

iTrader: (5)

There is a big long thread on it on msextra, and I have a couple long posts about it around here. I will try and write something up tomorrow. But you seem to have the gist of it.

Try things out in open loop. Also if you haven't already set your frequency to 19hz. It's easier to tune.

It's a lot of trial an error. Because each setup is different.

Good luck.

compuw22c

Ok, so I've made some progress after finally seeing this in the help file (that didn't seem to be in any of the online manuals):



So while I would THINK that being in "Basic" mode would simply force these values, it doesn't seem to. Basic mode continues to use the PID values as far as I can tell. So, I set the numbers in the boxes to 100-100-100 and started using the slider. It works! I also changed my frequency to 19.5 (thanks for that) and it does seem better (although I would think if the solenoid supports the speed the higher setting would be better, it's the DIY autotune solenoid...)

I've so far upped my sensitivity to the level it is now. I seem to be overshooting my target but I can live with that. I'm guessing I could start playing with the PID values in the boxes after setting the sensitivity and maybe tune that out but as I don't know which direction to go for exactly what result I want, I'm guessing I'll have to play by trial and error.

I know most people have much lower numbers than 100 and so I'm wondering if when I start changing things I'll need to readjust sensitivity? But everything I read says don't touch the slider after setting with the slider...

Pics of my log:




Not sure what to do to avoid the overshoot to 13.5. Not saying I'll never want 13.5, but I'd like the algorithm to be tuned so it only gives it to me if I ask for it. Seems like it needs to reduce the DC as it's approaching its target a bit and then drive it back up a bit after it passes it. Maybe increase P so it reacts quicker to move towards the target (and back off once past it?)

aidandj

iTrader: (5)

Basic shouldn't take into account the PID numbers. If it does, file a bug report on msextra.

Also, set your min and max duty cycles. It will help the loop by having a smaller range to work in. Start at a low duty cycle and increment by 5. Then you start getting earlier spool that is your min. Your max is the opposite. Start at 100 and lower by 5 until your spool slows down.

compuw22c

Ok, I'm bringing this thread back from the dead because I'm wondering if something may be up with my solenoid. The issue I'm seeing is that a duty cycle of only 19 is needed to get me to 10psi, but 20 seems to overshoot by more than 2psi (target of 12, so I hit 14, and I don't want to). Isn't this odd? Aren't most people seeing a much wider range of solenoid duty cycle (and thus more resolution)? I've got the standard FM kit with a 7psi wastegate, so I would think I'd have more "resolution" to play with in terms of duty cycle. I see others here running much higher, like 50/60% duty cycle on setups not much different than mine. It works correctly, increasing duty increases boost, and I don't get super crazy boost surges as if it was hooked up backwards. Anything I should look for? I feel like this may be the root of my tuning problems, but I'm not sure what would cause it or how to fix it. Unless it's just a bad solenoid, but it's always been this way and I'd hate to throw away $50 on something that's already working ok...