Megasquirt and EBC, I hate it
 

I hate Megasquirt. It is the source of never ending sorrow for me.

So I have an entirely new setup on the car and everything is going good. I decided to finally install my EBC. Everything is wired and plumbed up correctly. So first issue. I don't know which Port it I need to set it too. So everytime I turn boost control on I get a system error message. Doing some googling I find a thread on this

https://www.miataturbo.net/megasquir...n-error-62986/

I disable PA0. Power cycle and no error. "Thank Jesus" I think, but this is where it really gets fun. So after a couple seconds it generates a difference report. For some reason the controller settings are all FUBAR'd. So I hit, "use tuner studio settings" then it starts flashing some error about something being messed up. I try to cancel the upload but its no good. Then it gets better, go to start the car, fires once then quits. I reload an old tune, same thing. Now I'm really pissed I go to the oldest restore point I have (Just two days ago) No dice. Fires up then quits immediately.

So I'm rather upset right now and about to use my chopsaw on this thing unless I can get some help. Last time I had issues this bad my ignition system ended up frying itself so I'd like to avoid that if at all possible.

Attached is a copy of the tune I'm trying to work with.

Any help would be greatly appreciated.

Quick update. Now turning the boost control on no longer gives me an error message no matter what I do. Still wont start though, fires up then dies.

Why do you do this to yourself? I gave my advice, which was, if running ms2, don't even mess with EBC. Keep your shit simple and you should have less failure.

It shouldn't be this hard though. And it definitely shouldn't have screwed me this bad.

Here's some more interesting information. I noticed the fuel load with just the key on was 50% not 99% as it should be. I was able to track down an old tune from awhile ago. Loaded that and its back to 99% but still wont stay running.

It will stay running slightly longer this morning. If I crank the ASE way up it'll idle for a couple seconds.

Here's a composite log of trying to start this morning.

Is your fuel pump staying energized? What's the fuel pressure doing?

I believe so. I don't have a fuel pressure gauge but the fuel load stays up at 99% then when it fires up slow drops down. Its like everything just shuts off after it fires up. Playing with the throttle does nothing.

Yeah so I updated the firmware that made it worse. Still doesn't start and as a bonus rpm's aren't synched to the megasquirt anymore.

Is this one of my old units based on the DIYPNP? Did you flash the code for the MS2 or for the Microsquirt Module?

Yes, flashed for microsquirt module. Finally got it start. Rpm's synced. Stayed running only if I kept the rpm's above 1000. Letting off the gas makes it die.

bonus problems galore. Starts and sort of runs. If I use the throttle to get it running it will run at like 500 rpm. Not sure if its not getting spark or not. I will have a datalog in a second.

Datalog. I think maybe some settings got lost in translation or something. Also included current tune.

Ok so I can get it to idle (sort of) and I drove it around and it feels fine aside from the fact that my fuel map is apparently fubar'd and what not. I checked my idle control settings and it says none. Can't remember what it was before. Idk it definitely doesn't sound right at idle but at least it "sort of" runs. Can I throw money at someone and make this go away?

Send me your current map. Also remind me for the Nth time the unit's serial number.

Email sent. You are a god among men. The amount of times you saved my behind is amazing. Thank you so much for the help.

Replied.

Thank you ever so much! Car works great now although I still get a system error with the boost control turned on.

Car idled fine. Made it all of a mile before it shut off and now won't restart

Today sucked serious horse clocks but I think I'm all better now. Rev kindly fixed my issues and told me my serial to usb cable was broke. So I got a new one. No more issues. About to go test out the ebc. Also after I jiggled some wires I was able to get the car started again so I took out the last few vamp taps I had left. Hopefully no more issues.

That's good news on the positive progress.

Yep good progress except I still cant get the EBC to work, Open loop inverted or normal nothing seems to happen.

Amusing, I'm having a similar issue with my DIYPNP. I get a config error when I try and connect, and toggling the EBC settings does jack. Oddly enough if I disable EBC entirely, I immediately hit overboost protection, but the thing is wired up correctly and worked just fine last year. No physical changes.

If you get a config error when you enable boost control with the port set to PA0, then something else is also set to use PA0. Please post a screenshot of the message.

Here is the error message.

https://www.miataturbo.net/attachmen...ine=1402275177ebc ms error by Gorillazfan, on Flickr

Again, I found that other thread and disabled the PA0 port and after realizing I didnt splice to a switched 12v and fixing that it seems to be working.

My only other question now is I'm operating in open loop mode and I can hear the valve clicking away furiously (when into a non-zero duty range). Is this normal?

Yeah the valve should make noise, it will be louder if its mounted solidly to the body.

Having said that i have a few MAC valves that have become EXTREMELY noisy.

Dann

Of course I mounted it to the cowl too. Looks like its getting some rubber insulators.

As long as its not hurting anything I guess it will be ok. Still alittle confused as to how to set it up. Do I want like a smooth curve? Or basically everything the same duty except 0% at 0.00 throttle?

Still hate this thing. My tune always sucks, or it doesn't, depends on the mood of the ecu. Now it randomly drops offline when I'm datalogging. I thought I had open loop holding a nice 10 psi. Go out the next day and it immediately hits 13psi. Scale my stuff back. Hitting 10-11psi. Fine. Go out later on the highway now its hitting 12psi. Whatever, I have the fuel for it I just really wanted to get a solid tune at 10psi.

When I turn on closed loop the car runs like ass. Idles at 2000rpm, which randomly jumps around. Then the best part is that it will shut off and then not want to start up.

Megasquirt. The joy.

Tune your shit. The enjoy...

So what does you not knowing how to tune/megasuirt have to do with EBC?

Not everyone is cut out for DIY.

You should probably buy a new NC.

I've had this car turbo'd for 3 years. I've not used a single kit part. I can DIY just fine. I do suck at tuning though. That's why I'm here.

What does not EBC and MS have to do with each other? How about the fact that my tune was fine I throw the EBC in and it all goes to shit. Again this is why I post here because it's literally the only place I can get help.

I can handle tuning the fuel. The issue is that it never seems to be consistent. I just drove home from my buddies. It started fine, drove fine, fuel was fine. I park the car shut it off but then I have to move it. Won't start. Power cycle and it starts but it's idling at 2000rpm again. I don't know settings to look at to adjust for things like this.

Well you have a conflict error on your MS.

you have two things trying to use PA0.

That's an issue. this is simple MS 101 stuff here.

Only one thing can use one output. You need to determine what two things are trying to use PA0, we know one, what's the other? Did you even follow the directions of your screen shot and open mini terminal and see what the conflict was?

Have you deteremined that boost out is actually on PA0 and not something else?

How do you possible expect boost control to work when you have a tune and settings that barely run the car and have conflicts that cause errors?

Yes I did follow the error message, which yielded no results, which led to me googling the subject and finding the thread I linked in my first post. In said thread Reverant says that the PA0 port needs to be disabled in the port settings, which I have done. Doing this made the error go away and car run. And yes I visually checked to make sure PA0 was correct

Disabling the EBC from using PA0, also disabled the EBC. So yeah, you'll never be able to use it. Sure that cleared the error, but what Rev was suggesting is that you need to turn off whatever ELSE happens to be set to PA0 so then you can actually use the EBC output.

Until you do this, you'll never have a working EBC.

That would make sense but the EBC does work. I make more boost thank wastegate. Changing the settings changed the boost levels etc. Unless I really messed up.

So are you suggesting you turned off the conflicting output using PA0 and not Boost Control?

Maybe? I'll go hook my laptop up right now and get the error again and try the mini link again

With PA0 enabled I get "Conflict with boost pins"

youre not answering the question.

With PA0 enabled WHERE?!

when you get the conflict it's enabled in TWO places. You are switching it on/off FUCKING WHERE?DSAFSFDSAJFKDASJVKSD: Jsadkfjda
sfdjafsfj
dasj
ajcv
xccj
zcCZXC

jesus fuck.


something else has to "conflict with boost pins."

I don't know, you should understand by now that I don't have a firm grasp on the megasquirt. When I go to on/off ports this is what shows.

https://www.miataturbo.net/attachmen...ine=1402497429Capture by Gorillazfan, on Flickr

Turning it off and having EBC on it seems to start up the same as if EBC is off and PA0 enabled.

Yeah, you definitely need to disable PA0 from the programmable outputs menu.

Thank you for indulging my ignorance. So after that I just have to figure out how to tune to fix the other issues?

yes. that's your programmable outputs page. keep it disabled.

if you notice on the boost control settings page, pa0 should be selected as the output as well.

Here you can see the unusual idle thing going on. I shut the car off at a light and restarted it and it stopped.

https://www.miataturbo.net/attachmen...ine=1402511416idle oscillation by Gorillazfan, on Flickr

Here are my current boost control settings

https://www.miataturbo.net/attachmen...ine=1402511416boost control settings by Gorillazfan, on Flickr

My biggest issue, I guess, is the unstable idle. For example yesterday I had been driving for about 15 minutes and was stopped at a light. When I went to pull away the car stalled and I had to power cycle the ecu for it to start again.

looks like you need to tune. this would involve tuning things like your fuel table and idle control settings.

I have been tuning for the past couple days. What I dont understand is why, even though my idle settings have always been crap, it would suddenly change with the addition of ebc. For instance with cl ebc on it does the idle oscillation as seen but will not stop, then when I shut the car down it doesn't want to restart.

Found an old, un-corrupted tune with my idle control settings. Somehow they got way messed up. Changed my current settings to the old settings and turned on Closed loop ebc seems to be working just fine. Does seem to idle droop more then it used to. However I can deal with that for working boost control.

Ok slightly confused. Tuning PID with polarity set on normal. P=100 should yield wastegate pressure? well I hit my target with lots of oscillation.

I then changed P=120 and it hit the target without oscillation and basically held it. Which confuses me because I thought P had to decrease.

So I changed control interval to 51ms went back to P=100 and it hit and held target with minimal oscillation. I read that you want to overshoot target by about 10kpa or so before adding any I. At P=70 I overshoot by about 10kpa. But again it doesn't really oscillate and it seems to hold it. Granted I haven't done any pulls to redline yet.

Is the reason I only have to lower P slightly because I'm only going about 3psi more then wastegate pressure? Or am I missing something in the set up instructions. I found somewhere in my searchings that P typically ends up around 25-30 range.

what firmware?

The most recent 3.3.2. I know they changed the boost duty from previously being 100% as wastegate to 0% but I wasn't sure if the PID settings changed for closed loop.

Still not really sure whats going on. I've switched back to open loop for the time being but I'd like to get the closed loop done correctly. If I put P=0 I overshoot the target but also if I put P=100 I essentially hit the target with minimal overshoot and some oscillation. I has the dumbs.

I've also had the idle oscillation issue under the same circumstances. It's not related to boost control in my case. It's like the MS2e loses sync or something. In my case it ran the injectors and ignition but couldn't control the alternator so went high alternator output voltage and then went into some weird idle rpm oscillation. I had the battery light (1999 miata) on constantly in these circumstances. Then sometimes blipping the throttle would kill the light and sort the MS2e sync out.

I think Reverant is aware of a possible mod that may help. But I'm not 100% sure.

In my case once the MS2e is happy then when restarting key on ignition to position 2 for 10 seconds before cranking. This seems to reduce the chance of it starting into the oscillation thing.

Lloyd

Ok so I'm making progress but still not sure whats happening. So I got annoyed went and changed some settings including setting P=20. This resulted in 14.5psi. So I worked my way back. P=50 nets my target of 10psi and basically holds it to redline with no oscillation. P=40 hits ~11.5-12psi and holds it to redline as well. So I started adding I and it didn't seem to make a difference. So right now I have P=50 and everything seems ok. So other then that I'm not really sure where to go from here? Add a lot of I? it doesn't really overshoot and oscillate though. It already feels better to drive then open loop though.

Also whats the deal with MS and skipping like 2 seconds in my datalog? You can see in the screen shot how it jumps from like 3800rpm to 5000rpm.


https://www.miataturbo.net/attachmen...ine=1403795833boost control settings2 by Gorillazfan, on Flickr

https://www.miataturbo.net/attachmen...ine=1403795833cl ebc p50 by Gorillazfan, on Flickr

you must be barely adding boost over wastegate...

how well does it track with throttle changes? I think you'll find you'll need some I to maintain targets in less than ideal situations.

i was about to suggest it's because you're using the intial tables, but you're not.


when tuning P your primary focus should be spoolup, this should be done in 1st and 2nd gear. If you're doing it in a high gear, you might not hit the target in a low one. Once you tune it so when you mash it in 2nd gear and you're just overboosting over your target, pull to redline and see if you need any I or D to help long term tracking or damping.

wastegate is 6.5psi target is 10psi so not a huge jump. Ok but I'm on the right track now? I'll have to do more datalogs all I've done so far are 3rd gear from 3500rpm

I don't know how I missed this part. Yes I didn't use initial tables because I was confused and thought it would make it simpler off the bat. And yes today I did a pull in second with P=50 and I=20, I overshoot to 12.4psi then it dropped back to 10psi as rpms increased. So I will continue to work on this. Thank you Brain for holding my hand.

It's getting there I suppose. So I did what Brain said to do, got P=40. Then I started adding I and doing pulls in 3rd gear. so adding I little by little just wound up giving me an initial overshoot and then tapering down. So I just went to I=100 and the results were initial overshoot then quick taper down. As seen in the image below.

https://www.miataturbo.net/attachmen...ine=1403905732cl ebc p40 i100 by Gorillazfan, on Flickr

So I figure now it's time to add D to dampen the overshoot correct? So I set D=50 and I get basically the exact same thing as before. See picture below to compare.

https://www.miataturbo.net/attachmen...ine=1403905732cl ebc p40 i100 d50 by Gorillazfan, on Flickr

I then set D=75 but the datalog is garbage and I don't believe it because it pulled immensely hard but the log only shows 10.9psi.

So am I being foolish? Should I only add a little bit of I and then add D? With I=10 I overshoot then slowly tapered back down to 10psi by redline. So I assumed (perhaps foolishly) that I needed to add more I.