12 replies to this topic

[Paul]

I've got a small issue thats been bugging me, and i'm trying to get to the bottom of it, so i'm here to pick your brains.

The Evo II 63t has been running in the Satria since November and whilst run in, boost has been controlled straight off the wastegate. The car is using all standard E2 ECU, intercooler and piping, standard intake/airbox etc.
Yesterday i hooked up the stock boost control solenoid and relevent vacuum lines hoping to run the standard 12-13psi.

There's one line off the nipple on the cooler piping, to the top port on the solenoid, then from the bottom port it goes to the wastegate.

Took the car for a spin, and it freeboosted past 1.2bar and i backed off. Drove it back into the garage and started troubleshooting.

First was to make sure all vacuum lines were still attached and plumbed correctly. Check.

Next i removed the solenoid from the car and tested it. I tested it by putting 12v across the terminals and blowing air through it. Remove the 12v and air couldn't be blown through it, hence solenoid was opening and closing as it should.

I started to think why the solenoid wasn't opening and thats when i realised the ECU should have stepped in at roughly 15psi and delivered boost cut.

So it looks like the ECU isn't completely reading the air going into the motor as theres no solenoid trigger and no boost cut. But where does the ECU get its source?

I've spoken to a couple of trusted guys about it, with both mentioning the ECU gets its signal from the volume of air measured through the MAF. This makes complete sense, though i'm still not convinced this is the only source for the ECU as the AFR's appear to be all normal.

In my travels, upon removal of the turbo intake piping (MAF-turbo) i noticed theres a small hole just below the j-pipe. Its definately supposed to be there, and looks like factory intended. With nothing in it, its obviously sucking in unmetered air, so what goes in here?

I've swapped between 609 and 055 MAFs and an Evo III Ecu with same results.

Thoughts?

Sure i can, and i will be putting a boost controller on it and making the standard BCS redundant, but its bugging the shit out of me not be able to track down the problem. Plus, something else may be reliant on said source, and i just haven't come across it yet.

Edited by Paul, 22 February 2010 - 04:32 AM.

[JSTYLE]

modified ecu with no boost cut?

Jon

[Paul]

Nope. Should have mentioned, standard soldered chip in 2468 ECU.

[JSTYLE]

have u opened up the case? whats the eeprom number? EB94A?

Jon

[evopwr]

i thought stock boost is 8psi? sorry for being slightly off topic.

[Paul]

Yeah Jon, EB94A. The caps are all good etc etc..



Alex, the stock wastegate spring is 8-9psi.

[ENGINR]

Internal gate on the turbo sticking and not opening when the actuator tells it to? If it was something mechanical, then the system would be thinking all is OK, as the signals etc would going through and the ECU reading it as fine - but the actual mechanics of the system are failing.

[evo-gsr]

The bottom hole on the intake pipe should be blocked off... It usually has a plastic nipple sticking out of it to which you can attatch a small piece of hose to, with a screw siliconed into it it to block it off...

[VR-4Squid]

stock boost control solenoid
There's one line off the nipple on the cooler piping, to the top port on the solenoid, then from the bottom port it goes to the wastegate.

Took the car for a spin, and it freeboosted past 1.2bar and i backed off. Drove it back into the garage and started troubleshooting.

Suprised nobody else picked up on the fact it's hooked up incorrectly.

Run a line from the nipple on the cooler piping to the wastegate with a T-piece halfway, run a line from the leg of the 'T' to one port on the solenoid, run a line from the other port on the solenoid to the inlet pipe (bewteen afm and turbo).

The stock boost solenoid does not work like an aftermarket one, it simply lets the ecu choose to bleed 2psi out of the wastegate line, or not.

With a handful of T-pieces and some hose you can daisy chain several factory solenoids and with separate switches have multiple boost levels (in 2psi increments).

Also, the factory 'boost' cut isn't actually based on boost pressure at all, it's based on the air flow through the air flow meter, when the ecu see's a certain Hz (Karmann vortex afm's produce a frequency not a voltage) it triggers the 'cut'.

Edited by VR-4Squid, 23 February 2010 - 02:43 AM.

[evopwr]

Damn vr4 squid you are one smart cookie!

Paul, refer to one of the schematics found in the evo1-3 service manuals regarding factory solenoid vacuum line hookup points, from the sounds of it you've mixed them up.
What vr4-squid described is exactly whats drawn on the schematic.

-alex

[Paul]

I assumed it would be something to do with that nipple on the intake. After reading previous threads on here, i convinced myself into plumbing it in the way i had it.

I'll hook it up as mentioned, and test it out.

Big thanks.

+ rep.

[vr401]

Are you using SAFC to tune bigger injectors?

[Paul]

From the legendary AndyF on evo123.net

Right then....
As far as I know the boost solenoid doesn't feedback anything to the ECU at all. It is run purely on an output channel from the ECU and so it just tells it when to open the solenoid and nothing else.

The boost control on the ECU is done by measuring air volume (calculated using the MAF, barometric pressure and intake air temp sensors).

It has a table for the 'base wastegate duty cycle' that it uses but then has on top of that modifications to the duty cycle based on where it is in comparison to it's target volume. If it is less than target then it begins to open the solenoid more to increase the boost and likewise, if it is above target then it reduces it.

It's a proportional type of control as well so it you're well below target it opens the solenoid more than if you're close etc. so it's not just a simple level of control.

It also does it based on vehicle speed and revs so it will adjust it per gear to a certain extent as well.

It's got a lot more going for it than people will probably realise.

It also drops boost when detecting levels of knock as well IIRC but I'd need to recheck that in the code.