73 replies to this topic

[Jimbo]

I don't want to sound like a naysayer but perhaps you need to start again from scratch? Get the stock injectors back in there and have the standard tune loaded. At least you can rule a few things out. It sounds like the mob who did the work for you had NFI what they were doing and have made changes in the operation. 

It sounds like everything is working mechanically but the timing of it all is way out. 

[neggles]

I really, really don't want to put the stock injectors back in... Getting to them is about the least fun thing I've ever done

 

But at this point they're one of the few changed things left... Maybe it's related to the 3 bar MAP? There might be a MAP voltage threshold table in the ECU that I don't have access to?

 

Got It Rex had issues getting it to behave as well, and they began from the stock map file with no heed paid to the work done by the idiots who shall not be named.

 

E: Thing is, while I agree with you in principle, the idiots who installed the injectors didn't have access to anything I don't have access to, and everything except fueling/ignition is set to "stock" - I'm still not convinced this had stock data in it when I sent it to ECULabs.

[El_Freddo]

THIS

 

THREAD

IS

AWESOME.

 

Just saying - I'm new to boost (single turbo), but I'm finding the TT setup fascinating!

All the best with sorting it out neggles

 

Cheers

 

Bennie

[neggles]

It's a really cool bit of technology, and when it works properly, it's pretty damn great. Torque curve is really fun, it feels like you've hit a "turbo" button like you're inside a movie when it stages.

 

Unfortunately you can achieve similar power figures and area under the curve with a modern engine management system and an appropriately sized twin-scroll turbo, without the multi-page threads on owner forums asking for help... Subaru were determined to make the system work, made it work perfectly (or as perfectly as it was going to) with the DE revisions just to prove they could, then promptly threw it away and went single turbo on the 4gen GTs because that was the smart thing to do.

 

But where's the fun in doing the smart thing? Anyone can wind 30lbs of boost into a forged internals 2.2L (stroker) or 2.5L block on E85 and make some absurd power figure. It's not special. I like having something that's not so common, and once I get to the bottom of this nonsense I'll be able to clean it up and show it off and have something pretty close to unique. Definitely not something you see every day.

 

My daily driver is an MY15 Fiesta ST with some minor modifications. I don't need another car with a nice flat torque curve and a turbo that spools in the blink of an eye, heh I will admit, I didn't expect to spend as much money as I've spent on it. But that's always how that goes...

 

Anyway, ECV rod was already the correct length. It's gotta be programming, or something else's dicky in the ECU. I might buy nickknack's ECUlabs'd revD ECU (same PN as mine) just in case the chucklefucks broke it.

 

What's so hard about changing injectors? I thought it was one of the more painless jobs.
Changing turbos, now that's something I don't want to do again in a hurry. It's forked.
And I say.... NAME AND SHAME AWAY!

 

There's a whole bunch of accessories and lines and a straight-up pair of armor plates in the way - the actual injectors themselves are OK-ish to get at, but the bolts holding the steel cover plates on are not especially convenient to access.

 

With regards to name and shaming, the place I took it to first was ETuner in Campbellfield. ETuner appears to be pretty much a two or three man shop, and while I would probably trust them to handle tuning of a haltech or anything else that is software-only (and not ECULabs, as I don't think they ever wrapped their heads around the ECULabs software) personally I've no plans to ever do business with them again. Poor communication, long time delays, failure to deliver, and IMO overcharged me for they work they did actually do as the car left less functional than it came. I paid less money to GotItRex for the tuning done thus far than I paid them to install two sensors and four injectors, mess up some of my wiring, and generally make a mess of things.

 

At one point they spent 3 weeks trying to diagnose why an injector wasn't firing only for me to work out within 15 minutes of getting there that they'd plugged the injector for the passenger side front cylinder into the connector for the cam angle sensor, and the cam angle sensor wasn't plugged in at all. I understand how the mistake could have been made in the first place, as the ID1000 adapter cables are not keyed and the cam angle sensor uses the same plug except with a different key, but they were running bypass wires and everything before even confirming that all the plugs were in the right place.

 

So uh yeah not a huge fan. And you can see why I might think that the ECU is damaged given how they forked with the wiring.

[LukeFranky]

What's so hard about changing injectors? I thought it was one of the more painless jobs.
Changing turbos, now that's something I don't want to do again in a hurry. It's forked.
And I say.... NAME AND SHAME AWAY!

I've changed turbos countless times now. You get better at it

[LukeFranky]

You're right. Those metal shields around the injectors are a major pain... That's why I threw them away!


P.S. I have a rev D. Even with the positive side of the ECV blocked off it would open wayyy sooner that 6k RPM and that's with the rod fully tightened shut.

[LukeFranky]

I had an idea in the shower.

Do you have a boost gauge you could tee into the secondary turbo? Measure the boost there?

Maybe you could even use it tee'd into the intake valve and see when it opens, likewise for the ECV lines? Might unearth something.

[Jimbo]

 

 

My daily driver is an MY15 Fiesta ST with some minor modifications. I don't need another car with a nice flat torque curve and a turbo that spools in the blink of an eye, heh I will admit, I didn't expect to spend as much money as I've spent on it. But that's always how that goes...

 

 

If you don't make a build thread for the Fez.....ban!

I sold my LV XR5 a few years ago, fast bum dragging euro hot hatches hold a strong place in my heart.

[natho]

i haven't read all of this yet, but it sounds like you haven't looked at the map that actually triggers the change to TT mode yet. In my tuning i was able to control the changeover point pretty easily. i'm not sure what it's called in the ECULabs software, but the definition i have actually refers to the map as engaging the SPRV with solenoid 2. There were some other maps to do with changeover that do nothing at all.

 

as for the secondary duty cycle, this is badly named. it's the ECV duty cycle when it's being used for boost control. there are 3 boost control modes for the TT:

 

1) primary turbo only, using the primary turbo wastegate to control boost

2) primary turbo only, using the primary turbo wastegate AND the ECV duty cycle solenoid with ECV positive pressure enabled (pre-spooling the secondary)

3) both turbos, using the primary turbo wastegate to control boost

 

There is a map to control when #2 is turned on (ECV positive pressure enabled). This is on as low as 2000rpm in the factory tune, but you get better boost on the primary by getting this to come in later.

 

I had my 'secondary wastegate duty cycle' map all set to 100%. The boost control when this is engaged is actually not very precise, and i found it topped out around 15psi even with this set to 100%. Some advice on here has you bypassing this circut entirely - you get better boost on the primary when you do this, but a sightly bigger boost drop in VOD.

 

I hope that makes sense, I spent literally hours figuring all this out when the car was still TT. If you're having trouble finding these maps with the ECULabs software, I'm sure we could work it out with a screen share or something.

[natho]

just thought of something else. the map to engage secondary is load vs revs. to use big injectors, sometimes tuners do a trick where they put a false injector scaling in, and then change the MAF scale to suit. if this was done, then even the factory values in this table will be out.... and i just checked your datalog, i'm pretty certain this is what's been done for you. you're only at 1.36 engine load at 6000rpm when your SPRV #2 solenoid is engaged... normally engine load would be way higher at this point.

 

ask your tuner what they did, and apply the same logic to the SPRV2 engage map. i *suspect* you need to halve the load values in this map.

[natho]

oh, the map to engage ECV pos will need to be changed the same way also.

[LukeFranky]

Can I marry your brain natho?

[natho]

here's some info on the "MAF Hack" - http://forums.nasioc...d.php?t=1363692

[neggles]

That's... a very interesting point.

 

Here is my stock MAF calibration:

 

 

Modified MAF calibration:

 

Load comp stock

 

Load comp gotitrex

 

So... Yeah... that would definitely do it, right?

[El_Freddo]

We need another Sunday for further test runs... I can't wait another week to see if the above "fast and furious" data logs have sorted the issue.

 

So when in twin mode both turbos are spooling and producing boost at the same time?  I didn't know that, I thought they switched between the primary and the secondary and only ran on one turbo at a time depending on revs etc...

 

Cheers

Bennie

[neggles]

Nope, both turbos are almost exactly the same size and it runs both in parallel - much better for packaging, and two small turbos are cheaper than a small and a large

 

Oh, I should add, the MAF to Load factor is the same in this gotitrex ROM as the stock ROM, 23040.

Edited by neggles, 17 July 2016 - 11:22 PM.

[neggles]

Okay, so I've combed through the Got It Rex file and move some of the tables over to a fresh copy of the "stock" ROM.

 

I've noticed a couple of possibly weird things, and there's some things I haven't copied:

 

Possibly weird thing: Closed loop max RPM was set to 6800/7000 respectively. Stock ROM is 3400/3500 - possible ECU won't sequence in closed loop mode?

2nd one: Injector scaling is set to 1000, when IIRC it should be 1030 for this particular set of injectors. So i've set it to 1030.

 

Not copied:

 

Left closed loop max RPM on stock settings

MAF calibration and airflow load compensation tables (posted above) not copied.

Secondary initial and max duty cycle tables

 

I suspect the fueling might be a bit out of wack but we'll see. I'm not convinced the differences in tables I posted above are enough to be the reason behind the potential load mismeasurement issue?

 

Edit: Here's a log from when the car properly staged and rolled 1.5 bar of boost; https://www.dropbox....7-trim.csv?dl=0

 

Engine load was around the 3.3ish mark. Still too low?

[natho]

OK, doesn't look like the MAF hack has been done. Just checked the data log again, the reason the load is so low is you're only running 2psi at that point. I've read most of your posts in this thread, and i'd suggest the following:

 

go back to the 'came-with-car secondary duty map it makes 1.5bar of boost on the primary all the way up to around 4500'. using 100% for the secondary map is what worked for me to maximise boost on the primary. i'm fairly sure this isn't the source of your problems.

 

find the SPRV #2 opening map in your tune and check the values. Here's a stock ADM one for comparison: http://imgur.com/vzR4N4A

 

If you don't have this map, contact eculabs and get them to add it to their definition. it's probably there, just called something else. 

[natho]

leave the injector settings the tuner put in there... it's quite normal for them not to match what you think they should be. if you look at various stock subaru tunes they use slightly different values in cars with the same injectors. Changing them will also change your MAF scaling. Looks like yours aren't perfect, but unless you're planning on re-scaling it yourself, you're better off not changing it. Also, when trying to solve a problem like this, avoid making random changes because things don't "look right". change one thing at a time and verify each step of the way, or you'll never work out what solved the problem.