kenwo91

iTrader: (1)

Practically speaking I need lots more than "pointers"

I've read the tuning Ebc with 18psi and tried it myself and made small adjustments during the course, but since even making data logs is new for me (wtf ami doin) I have tried to scan for changes but I never get any real "AHA!" feeling.

Running MS3 basic with 1.4.0fw and OEM MAC valve equivalent to the DIY autotune one
I do not know what to look for since it feels like I'm chasing my tail trying to change the duty bias table when maybe I should change the settings in closed loop.

I got a lot to do on my tune, I know, my thought is to nail this somewhat ok, then its time for flexfuel sensor and then i would leave it all to a real tuner who knows all the ins and outs.

So.... where do I go from here, also thought about trying out a later fw since fiddling with engine states seems to be the new fab and maybe getting better cl ebc tuning whilst at it.

DNMakinson

iTrader: (3)

Check to find if 1.5.1 has the SD card logging of PID parameters, or only 1.5.1+ beta. That could help you, as I think you are alluding to. I'm not sure the settings are any different, only the diagnostic abilities.

I bought a used MAC valve and it seemed to not have adequate passages to allow it to vent my boost source and give me fast enough response. I put my old (the old style DIY... I don't remember the brand) valve back on with better results. A single data point.

If you read 18's write-up then you already know my pointers.... very aggressive settings to get good performance for EBC.

Mudflap

You need to log runs at WOT from start to finish in say 3rd gear. You are tipping in and cautiously heading towards 100% throttle.

What I found is, like all control systems, life is WAY easier if you are nearly there and just need small adjustments to keep you in line. So you need to have a very good understanding of what boost you get at what EBC %. So you need to log 10, 15, 20, 25, 30% open loop EBC WOT pulls. Now you know exactly what to expect and then can start in on your bias table.

Then you need to log a WOT pull with basically 100% EBC. Which means you WILL hit your overboost limit. Don't do this more than once or twice...

But that will give you a graph of what is possible. From there you can start bumping your bias table towards perfection.

Do open loop tuning on Control Boost Duty table and adjust it until you get that WOT 100% EBC response WITHOUT shooting past your intended boost level. This is hard to do and takes time. Think about what your actuator is doing during and what it should do to achieve your goal. Namely, it should be allowing max spool and tapering off as you approach your goal. Tapering is key to avoid overshoot.

Then take that Control boost Duty table and put that in your bias table. Now you are try closed loop mode and see if it behaves. From there you adjust your PID settings. start small on those values. Otherwise it'll just overshoot.

good luck.

kenwo91

iTrader: (1)

Thanks a lot to you both!

Now I know where I'm heading and what to expect, I'm sure I will get back to this thread in a while after logging the different duty cycles.

Thanks again! Xoxo

Chiburbian

iTrader: (1)

Here is some info that may or may not be useful:

kenwo91

iTrader: (1)

Thanks a lot you guys! Made logs running 10/15/20/25/30 % duty and got some numbers to aim for.

Running in setup mode were quite good but turning on closed loop and all hell broke loose.
I would think my duty % is quite reasonable but still getting over boost protection if i blip the throttle at 5-6kRPM.
Also getting some bumps right when boost builds. Is this solely because the closed loop control handles lazy or what?

In the attached log i run from 2-5th gear and 22 and 27s is the throttle blips.

Re: my own thoughts, is it maybe because i run the higher duty in 26-4600 rpm range while asking to much in the target table im getting duty spikes at 195kpa?

Mudflap

First of all. you need more data points. Go change your communications data rate to max. .

The key to getting your PID to prevent overboost is to not ask too much of your system. The delta between your Boost Target and your Actual Boost multiplied by your P is what is used to calculate the duty cycle of the EBC.

So go plot Boost Target 1 on top of MAP. Plot boost duty on top of that and you can watch the behaviour of the system.

Boost control response is SLLLOOOW. Take a look at the first few gear pulls. You are hitting 185-190kPa at 5000 rpm but are requesting 178kPa. Therefore the duty cycle is dropped way off. So then you get that dip in boost. You can change your Boost Target to 185-190kPa to avoid that dip.

What I see happening on that 5k RPM pull is that your delta between Boost Target and Actual is 50kPa. On the other pulls it is more like 30kPa. So your requested EBC duty cycle is higher than normal. So it runs away on you. You need to use Advanced PID (not basic) and start adjusting P and I. Don't do D right now (keep it at zero). You can pull back P in order to reduce duty cycle.

Again, you have to consider that the boost response to EBC changes is very slow. You have to have reasonable expectations. If you apply too much EBC duty cycle early on, then you'll hit overboost. The boost builds up and the system can't immediately respond.