Rick02R

Thanks in advance for the help. I'm in the initial stages of tuning Closed Loop with my EBC. I have the new DIY valve, 3 month old Rev MS3. Car is 2002 w/ a T25 turbo, 7.5lb waste gate.

Based on everything I've read, the valve is connected correctly (it does actuate). It's closed at idle (if I invert the polarity in Tuner Studio it is open at idle). Wiring is right, vacuum tubing is right.

Regardless of the P value (100, 200 or 20) I get initial boost close to my target boost 175 kpa and immediately get oscillation of 1-3 PSI increasing until I hit boost cut. I and D are set to 0. My understanding is that with P = 100 I should get waste gate pressure to redline. (I have read and re-read the mega manual process and tried some MT.net recommended values for I and D but I think I've got something wrong well before I get to tuning PID).

What's obvious that I'm missing here?

(EDIT: Happy to post logs but I'm not even sure what would be useful at this point)

Braineack

iTrader: (62)

is it on the correct output? sounds like it's not wired correctly and the software is never closing the valve.

mlev

iTrader: (5)

log MAP vs RPM vs boost duty cycle and post that up, see if the ECU is even attempting to modulate the DC to compensate for the rising boost.

FWIW, I was having very similar issues tuning CL EBC PID the week before an autox, spent $20 on a cheap MBC from amazon, and haven't looked back since.

That said, my boost varies depending on weather a bit, and I would LOVE to get my CL EBC working proper... but I was getting PISSED trying to tune it.

Rick02R

Brain,

I thought so too, but I graphed the boost psi vs boost duty cycle % and the oscillation lines up exactly with the changes in duty cycle. It's currently set to use the "boost" output (and I triple checked the wiring against Rev's diagram vs my car).

mlev,

I understand why. I read your thread... the car is definitely trying to modulate the duty cycle. It's at 100% (closed) when the car is at idle and you can see it oscillate along with the PSI when in boost.

What I don't understand is why I'm not seeing 0% duty cycle and waste gate pressure with any value of P.

ScottyP3821

iTrader: (2)

Go back and start with the basics. First check your wiring and set the solenoid in your settings to open loop @39hz then go to you table and set everything to 50 if you hear clicking move on to CL PID. If not then its wired or the settings are wrong.

Once you get to the CL stage set P to 150 and see where your at. Then slowly lower it till you overshoot your target by a couple psi then add I to take out oscillations. Heres a easy way to think of it

P=Power
I= accuracy/anti oscillation

You will also need to adjust your boost target table before tuning PID to your desired kpa

Braineack

iTrader: (62)

have you confirmed with a multimeter, and boost set to open-loop, that you can cycle the valve using the MS? To me it still sounds like the MS is not connected to the valve. The MS might want to be trying to ground the valve, but I don't think it's actually happening. Set your open-loop table to 50%, get it to cycle boost at idle, and test to ground and see if it's rapidly going to ground or not.

mlev

iTrader: (5)

Also works to set it to 50% OL and get out your mechanics stethoscope/long screwdriver and listen to the solenoid. It will make a quite audible "tikatikatikatikatika" if working as intended.

Rick02R

Tested with 50% duty cycle open loop. Definitely cycling the valve based on sound, vibration and volt meter.

Scotty, when I set P=150 do I stab the throttle, roll into it, hold hold boost (for how long?), what RPM range? When you say over shoot target do you mean on the first spike or subsequent oscillation?

Perhaps I'm wrong but I expected when I find the right P I would see a spike to just over my target and then it would trail off to waste gate as RPM increased. I'll go out and get a log after dinner.

Rick02R

Ok, logs attached. First time I've logged anything so I'm not sure exactly what you're looking for. If something's missing let me know.

From my perspective I saw little change between the 4 runs. Each contains a roll in, a stab and whatever else I got on my short test track. All runs were on the same piece of road.

Target Table is 175kpa across the board if throttle is above 20%.

First Run: P=150 I=0 D=0
Second Run: P=150 I=50 D=0
Third Run: P=150 I=50 D=50
Fourth Run (generally accepted settings per Brain in another thread): P=28 I=28 D=100

ScottyP3821

iTrader: (2)

Ill look at your logs after our staff meeting but for the meantime to answer your question what I usually do is set the boost duty target table first. Usually I set it to 0 for 0kpa to rev limit then Xkpa X being your boost target in all the other cells. Now tune you P by stabbing the throttle at around 3.5k and set your P till you hit just over your target. Next add I to to take out oscillation. this will also cause you to drop below your target so you may have to add more P. Tuning PID is a lot of rinse and repeat. Once you have it tuned pretty well go back to your target table and play with your targets to help maximize traction or how you car responds to boost by throttle. this part will be a preference type thing.

Chiburbian

iTrader: (1)

Everyone says "Go to a gear that builds boost quickly and start at like 1500-2k rpm and roll on the throttle quickly."
Or at least that's how I understand it. Edited with the following stuff from Scotty: What I want to know is if they are talking about 2nd gear, 3rd gear?

As a test, try and tune steady boost with open loop. I created an open loop table that gets me about 10psi pretty steady to redline and now that I know that can be done, I have worked out my other gremlins. Just the other day after adding NB alternator control via the latest firmware beta I went back to closed loop. Knowing that I can switch back at any time to open loop and have decent performance (and knowing what different boost levels mean HP and torque wise) has freed me to be less anxious in messing with closed loop.

BTW, you say that no matter what your P is, you hit boost target? I am not sure if it's an option on your version of firmware, but on my firmware there is a setting that is "keep valve closed until within X of target". Perhaps your X is too low? Zero maybe? So what is happening is your PID is never able to do it's thing.

Also, On my tune I have it set to keep the valve at 0% below 100kpa. When I drive around it is basically in "Wastegate" until I get on it. At that point the EBC goes to 100% until 10kpa of target.

Just some thoughts...

Braineack

iTrader: (62)

I would expect no difference between runs 1-2-3, the CL code is essentially turned off.

run p0 i0 d0 and try again; that should overboost.


is your wastegate already just at ~165kpa? seems like it.


when first tuning P, you're just tuning spool. so these should be 2nd gear pulls to only about 4-5K.

ScottyP3821

iTrader: (2)

I do it in 4th from 3k to 5k. With my shitty tires I loose traction in 2nd quite often. Its easy to tune in 4th on the highway.

ScottyP3821

iTrader: (2)

I only use open loop to test if the EBC is wired correctly. **** all that open loop noise. To me CL is easier to tune as you dont have to retune all the time the weather changes. You just Set it and Forget it.

Chiburbian

iTrader: (1)

Well, I set open loop a little below what is my "safe" level and just let the performance rise and fall as weather changes. My open loop isn't for performance more just for a failsafe that works reasonably well.

That said, I really want to get my closed loop working perfectly now.

Rick02R

Thanks for the thoughts all.

I'll run a P=0 test tonight for Brain and try Scotty's instructions.

With the OL Duty Cycle table at about 30% +/- I've got a nice steady 10 PSI pull to read line in 2nd. CL isn't critical but I can feel the power advantage of keeping the WG shut until target boost. I'm also very much interested in having this working so I don't have to tweak the OL table when the temperature changes.

My waste gate is definitely around 7.5 PSI (~150kpa). I pulled the EBC out of the system and verified.

"Boost Control Lower Limit Delta" is set to 100. I believe this basically says when I hit boost the CL code should start running. Very similar to Chiburbian's setup.

ScottyP3821

iTrader: (2)

Dude CL is seriously the freakin *******. Ask anyone thats using it. It made me fall back in love with my turbo set up. Not to mention the cool sound of your turbo spooling fast as ****. I get this crazy high pitched spool and the ****** launches hard. Damn I get excited just talking bout it. I cant wait to get off work now and drive.

Chiburbian

iTrader: (1)

Once you get CL working of course... Until then it's frustrating as heck. The problem I think is that there are a bunch of variables that can cause you problems. If you just happened to not have those issues in your system everything goes together easy peasy. If not well...

I have yet to get a truly reliable closed loop system working. That said, I am knocking on the door now, I hope.

Braineack

iTrader: (62)

no need to do 4th gears pulls to tune P. P is only tuning your ramp to target. When you tune I and D you need to do long pulls.

If you cant spool to 10psi from a roll in second gear without breaking traction you need new tires/suspension.

but yes. CL is the cat's titty pajamas.

plus the newest firmware makes it even easier where you really dont even have to tune PID and you can just dial in the sensitivity IIRC and even use the OL map for reference before PID even kicks in. I haven't played with the latest firmware yet.

ScottyP3821

iTrader: (2)

Im on the latest firmware. I real like the sensitivity adjustment slider. Made PID tuning a very short process.