New Turbo for project GSi

[PrecisionBoost]

Oh one side note.... I think I killed the BOV.... it stopped working after the first couple passes for some reason.... hopefully it's just a loose connection or something.


Sound wise....very nice.... I also have a screamer pipe that makes a nice high pitch sound as the wastegate starts opening around 4500 RPM.

The boost does come on a bit later.... but it's a very progressive strong pull up to redline..... unlike before.

How best to put it..... before it was:

nothing, nothing, full blast, that's it?

now it's nothing... oh...oh ya.. ya baby.......yayayayayayayayayyayayayayayayaayaayyayayaayayayayay.... I wet my pants

It gives the impression that the engine is making continuously more power as you approach redline.... as opposed to leveling off or dropping at the top end.


Really makes the next gear smooth and powerfull..... you don't get that delay in power between shifts.... it's like one continuous powerfest.


I love my Garrett Ball Bearing turbo..... you certainly get what you pay for.

[exist3nce]

Hey Chris

I understand what you are saying with the factory ECU giving only 45% duty cycle and as you try to increase that by playing with the MAP, timing is not what you want.

Why not do the setup like mine where the Emanage is tapped "on top" of the injector signal directly, and reads pressure off its own dedicated greddy MAP sensor? As boost rises, it will add further duty cycle as needed based on how you tune it. This still leaves the issue of too much fuel at idle, as you cannot decrease duty cycle this way. But you can go back to the factory injectors and have 100% perfect fueling in all off boost conditions. Then use the large injectors in an "additional injectors" setup directly controlled by emanage on a separate fuel map and weld them into the charge pipe or manifold.


Garrett GT series FTW!!

[PrecisionBoost]

I allready have the Greedy with it's own MAP but the issue is that it can only modify the duty cycle by a percentage of original.

So if your stock ECU only puts out 40% duty cycle your limited as to how much you can increase the duty cycle with the Greddy unit.

That is to say I need the factory ECU duty cycle closer to 60% to get where I need to be with the tuning of the Greddy.

During initial tuning we had the Greddy maps maxed out at like 3500 RPM and it still wasn't enough fuel.

We ended up throwing in my adjustable 10:1 rising rate fuel pressure regulator and restting the Greddy Maps to 0% additional Duty cycle.

Then we tweeked the RRFPR and managed an almost perfect 12:1 A/F at the tail pipe ( my wideband said it's 11.5:1 a few feet down from the turbo )

In the end we didn't even have to tune it with the Greddy.... it was amazing how perfect it matched.

The downside was that it resulted in quite a bit of fuel pressure.... which was fine for running 10psi.... but I'm wanting to up the boost now so the RRFPR has to be adjusted downwards.

When I do that I can again use the Greddy to help compensate.... but I'm pretty sure at this point I won't have enough adjustability to tune it correctly.

That is to say I need the factory ECU to put out more fuel ( duty cycle ) to allow more adjustability for the Greddy.... which can only be done via manipulation of the MAP/IAT.... which I don't really want to do because it will throw off the timing.

Basicly I'm screwed for a high HP engine with the current setup..... I'm sure I can make it work up to 15psi perhaps 18psi but after that I'm going to run out of tuning ability.

If the Nubira had used 19lb injectors instead of 21lb injectors then I'd probably be able to bump that up significantly and run 25psi. ( since the factory ECU duty cycle at max load would be higher)

It's possible that the Optra ECU is a better tuning platform.... but at this point I'm not going to spend a billion hours converting the entire harness over to use the Optra ECU.

I'd be far better off throwing in a standalone.... or running 8 injectors.... or 4 injectors and 4 jets.

Jets are perfectly fine for "rough" tuning.... it's what you use in wet Nitrous system.

Jets are also what you use in all carbureted vehicles..... it's just a simple matter of calculating jet size versus fuel pressure.

If I go with Jets then I can throw in as much fuel as I feel like.... and tuning is simple.... you just replace one set of jets with another larger or smaller set to change the tuning curve.

Last advantage of jets is that they are very small ( usually 1/4" wide and 1/2" tall ) so they are very easy to mount into an intake system.

I have been looking at going to 8 injectors but the best way to do it is to put them directly into the manifold runners.... putting them in before the throttlebody is a bad idea.... the better solution is a Throttle body spacer plate with injectors in there, so that they injecting behind the TB instead of infront.

remember that I'm looking for 30psi of boost.... each cylinder must have the exactly perfect A/R or that engine will only last a few minutes.

[PrecisionBoost]

Ok..... so the car has had an oil change and is all together now.... and I captured a few 3rd gear runs with it set to 10psi to see what kind of spool up time I can get from the GT3071R

I have to say.... I am loving this turbo.... the spool up feels great.... the car feels ultra responsive and strong.

I dropped the car down to about 3000 RPM and hit it......within a fraction of a second I felt the boost come on... no hesitation at all.

At 3080RPM I was at exactly 0psi of boost... 100RPM later (3180 RPM) I was allready making 3 psi....another 100RPM (3280 RPM ) and I was up to 5.4 psi.... another 100 RPM (3380 RPM) and I was at 7.5 psi

So in a matter of just 300 RPM I went from 0 psi to 7.5 psi..... and all down low before 3500 RPM !

By 3450RPM the engine hit 8psi..... at which time the boost controller settings kicked in to slow things down and ensure I don't have a boost spike.


So... in a word.... I am pleased with the turbo upgrade.... well worth the money to go with a real Garrett/ATP ball bearing turbo.


Interestingly enough I was concerned about the duty cycle...... but after plotting this out it seems I'm in better shape than expected..... I managed nearly 53% Duty Cycle up around 5110 RPM.

Here is a graph of the data I plotted..... you can see the boost rise very quickly then the curve slows down as it pushes in towards 10psi ( I averaged 9.7 psi in the upper RPM band with a max boost of 9.9psi )

[PrecisionBoost]

I did however find myself a small problem.

If I go from WOT and let off really quickly the engine misfires once. (set's off an OBD Code )

I feel the issue stems from the fact that the Nubira ECU is cutting fuel as soon as the TPS signal changes.

That is to say the ECU cuts the fuel when it sees the butterfly close..... which is a bit of a problem because the manifold is still pressurized

I noticed 7.8 PSI of boost still in the manifold just a fraction of a second after let off and an Air Fuel ratio close to 14:1


I'm sure some of you are probably thinking "what about the blow off valve? " and the answer is that it's working just fine.

This little spike is only a fraction of a second...... there is a very slight delay ( perhaps a few hundred milliseconds ) between when I let off the throttle and the BOV releases pressure.

Unfortunatly the BOV is on the other side of butterfly..... so it will only realease the air between the turbo compressor outlet and the throttle body.... it can't release the pressure inside the manifold after the butterfly has shut.... the only way the pressure decreases in the manifold is from the cylinders gobbling up the left over air during the next firing sequence.

Keep in mind two cylinders suck in air per revolution... so at 5000 RPM the first cylinder sucks in air just 6ms after let off.... and then the next another 6ms later.

That is to say the 10psi inside the manifold get's gobbled up 6ms to 12ms after I let off the gas..... which is crazy fast.

Unfortunatly however electrical signals travel at the speed of light.... so it's pretty much instant.


Somehow I have to delay the fuel cut to the injectors after a let off..... probably just for a fraction of a second..... even 50ms would probably fix the problem

The problem doesn't occur if I let off slowly.... but obviously that's not an option when your rowing through four gears in just a few seconds with a let off in between each shift.

So.... more and more I'm thinking about going standalone..... since you can configure things with regards to let off situations.

I can't really blame the Nubira ECU....after all....that type of fuel cut isn't going to hurt the naturally aspirated motor that only makes a bit over 100whp.


I also suppose if I am adding other injectors or jets in the manifold they will be spaced back from the cylinder head slightly..... so that would in essence create a fuel delivery delay (as the fuel being injected would take a few milliseconds to travel down the runners and into the cylinders.

Basicly the air that is causing this problem is up at the top of the manifold by the butterfly..... by the time it gets down to the cylinders the ECU has allready stopped injecting fuel.

I suppose meth injection may also help...... as I belive this "misfire" is one single knock.

Who knows...... perhaps as usual I am overthinking this problem.... I know it's just on knock and it's not under load.... but it still bugs me.



Hmmm.... I was editing this post and I just had a thought..... I was having problems with the throttle being too soft.... when I hit bumps I would push down on the throttle slightly causing the car to jump forward creating another bump..... which in lower gears caused an evil neck breaking resonance ( like riding a pissed off bull )

So.... while I was doing this upgrade ( with Tom's help ) I installed a secondary spring to tighten up the throttle..... now I don't get that evil jerking in lower gears when I hit a bump.... but perhaps it may also be partly responsible for this problem.

After all..... the spring is causing the butterfly to shut far quicker than it did stock ( spring is fairly stiff )

Perhaps the Nubira's ECU delay in cutting the fuel directly corresponds to the time it normally takes for the butterfly to shut on a factory car.

Perhaps with the heavy spring this throws off the timing for the last squirts of fuel.

I think I will have to try that and see if it helps.

[daewoomofo]

im wondering what would happen if you had a hose that bypassed the throttle body, with a 1 way chech valve in it? would that possably help lower the pressure in the intake manifold when you let off the throttle, or would it just basicly be a waist of time?

[PrecisionBoost]

Hmmm... that's not a bad idea.... if the pressure in the manifold was higher than the intake it would push air out of the manifold to equalize pressure.

In theory the manifold pressure should allways be slightly lower than the intake... even under boost.... so the check valve would stay closed.


I just finished testing and rewiring the injector harness on the Greddy E-manage..... and now things are a bit better.

There was a bit of a problem with grounding.... the reference for the E-Manage was different than the injector ground.... which seems to be part of the problem.

I literally just fixed it ( changed ground reference ) and it seems to work great.... and I was wrong about the injectors before.... I had it in my head I could only get a 50% increase in duty cycle but the real factor is 100%.

As such I only need 40% duty cycle to manage 80% total duty cycle.... which I have at 4000 RPM... so all is good.

It's hard to explain but basicly the issue was the Greddy was not adding duty cycle as it should due to the problem with the ground reference.

The voltage put out by the ECU was not the same voltage the Greddy was reading.... so it was not getting the correct information fed into it.... which ended up creating a major issue with the output from the Greddy.

It's hard to explain ground references.... most people simply think everything is grounded 100% to chassis but that is a wrong assumption.

In some cases ground references are connected to chassis ground through a diode or very low resistance ( for example 10 ohms )

As such the ground reference can be as much as 1V higher ( or lower if it's a "virtual ground" ) than the chassis ground.... most of the time this is use to filter out unwanted noise created from other parts of the Engine ( such as the ignition coils which backfeed EMF into the chassis ground.... as well as the Alternator which can create an AC voltage on the chassis ground )


So now I'm just going to fool around with some settings and adjust the fuel pressure so that I can run more boost.

[PrecisionBoost]

Well... good news bad news.... mostly good.

Tom showed up and we did some work on the fuel system and as we were just getting ready to pull the car from the shop and do some tuning when he noticed this giant puddle ( 3ft x 2ft ) under the car.... he touched it with his finger and said "Oil ! "

My heart sank.... I thought the engine had something catostrophic go wrong ( since it was on the end near the gearbox)

I took one smell of the oil and cheerfully said "it's not engine oil.... it smells like gear oil"

And sure enough when we look under the car it was dripping out of the front bolt on the oil pan of the F20 gearbox

The front center bolt was severely damaged as well.... which may have happened the other night when the suspension bottomed out under boost when I hit a major bump ( launched me a good foot in the air )

So.... the good news is that the car is ok.... bad news is I had to do some work to get the gearbox sorted out.

I had a spare F16 which has a steel pan instead of the nice aluminum pan that came with the F20 ( came from Sweden or Finland... can't remeber )

So after some work removing rust it looks good as new.... here is the steel pan... and the aluminum pan with three nice cracks in it.

[exist3nce]

Glad things are getting sorted out...... I can't wait to see what numbers this monster will put down.

[PrecisionBoost]

There are still a few kinks here and there....I would have like to have had the methanol injection system already on the car.

We haven't had snow yet... despite single digit temperatures... so I've been scrambling to get the car on the road to sort out any of the small bugs.

Hopefully tomorrow morning I will have a chance to give it a quick street tune with the new setup.

In the end the fuel pressure was too high.... ended up getting rid of the variable ratio rising rate fuel pressure regulator that I had in there.

So... the Astra VXR stock fuel pressure regulator is in the car.

Right now I figure I must be making something in the range of 250whp.... right now traction is an issue in the first 4 gears.

That is to say on a good launch I can't actually stop the wheels from spinning until I hit 5th gear.... at which point it grabs.

Needless to say I shift as fast as I can and it's still not fast enough.

So I think I will limit the boost down to 8psi for the first two gears.... then perhaps bring it up to 15psi for the moment.

Who knows.... if all goes well I might punch in 20psi of boost, adjust the tuning and give it a 4th gear pull to see how it responds.

My biggest problem is finding a place to tune.... I have found a number of nice deserted highways.... but 100km/h comes up in what seems to be 3 or 4 seconds just as I shift into 3rd gear.

I don't mind speeding a bit.... but to get a decent pull I need to exceed the highway speed limits by a very significant margin.

3rd is the first gear I can manage to keep traction with on a rolling 3000RPM start ( as long as it's perfectly smooth pavement )

Even on a rolling start 3rd will spin if I hit a bump during boost.

The car is just too light, the suspension is too stiff and the tires are not sticky enough... especially in this cold weather.

So.... Tom called up a friend of ours while I was working on the gearbox and I should hopefully be able to get a set of Koni Sport Adjustables.

I really liked the idea of a coilover suspension but that would require some modifications and I want to get the car "streetable" for now.

I was thinking of putting on drag radials.... but then again our streets might be too crappy for that..... so perhaps a set of nice R compound tires.

[PrecisionBoost]

Well.... I spend a couple hours tuning this morning and I'm still not happy with the setup.

The Astra VXR regulator seems to be a 1.5:1 ... but I could be wrong... it might be a 1:1.... it's hard to take readings from an analog gauge under full acceleration.... it seemed to hit about 60 PSI.... on 10psi of boost.... so that would be a gain of 15psi of fuel pressure.

But then again a 1:1 would have given me a fuel pressure of 55psi..... so it's not impossible that what I thought was 60 was actually 55.


So that's good news... nice safe fuel pressure levels.


I managed to push the A/F down into the 10:1 range for most of the pull.... because I was being very conservative to save the engine.

At 10psi of boost and 4000RPM the duty cycle ( with adjustment by E-Manage ) was 72%

I know I'm running very rich and should be shooting for 12:1 A/F but still.... 72% DC @ 10psi doesn't make me overly happy.

The fuel quality up here is crap.... so that might have something to do with it....I might have to buy a 55 Gallon drum of high octane fuel for the GSi.

Of course adding methanol injection would also help things.

I'm still having problems with let off.... each and every time I let off the throttle after boosting the engine spikes to 17:1 and the ECU detects a misfire.

I tried adjusting the Greddy Map but in the end that hurts drivability when not under boost ( A/F too rich and it stumbles )

Overall I'm not really that happy with the current engine management strategy.... and I don't think it will get any better at this point.

It's simply not advanced enough to give me control over certain aspects of the management system.

For example.... on let off I am decelerating... which has a different MAP than it would be if I was cruising at the same RPM level.

With the E-Manage you have no control over this.... so if you add or delete fuel at a particular RPM and pressure your stuck with it under acceleration, cruising and deceleration.


I was hoping to push it to 15psi on pump gas today.... but I decided not to risk it because it may cause even worse misfires on let off.

The other issue that is bugging me is that the engine was designed for 8000 RPM and the factory Nubira ECU cuts out at stock RPM levels.

I don't know if I'd push it all the way to 8000 RPM.... but I'd like to see 7500


So I've definitely decided as of today that the GSi will get a complete overhaul this winter with respect to wiring and engine management.

This engine is too nice to be stuck with a crappy management system.

I have many options.... but I'm pretty sure I will go standalone.

I may even decide to get off my ass and design a new management system

[Big Jeff]

So I've definitely decided as of today that the GSi will get a complete overhaul this winter with respect to wiring and engine management.

This engine is too nice to be stuck with a crappy management system.

I have many options.... but I'm pretty sure I will go standalone.

I may even decide to get off my ass and design a new management system

if you do design your own, tell as how
I would be instresed in making my own but would know why to start

[PrecisionBoost]

Well to be honest I'd end up selling it as a high end standalone... so I probably wouldn't want to tell you exactly how to make it.

Do you know much about electronics? ( microprocessors, ADC's, DAC's, software ? )

I don't mind talking about the subject... as long as your going to be able to understand what I have to say.

[Big Jeff]

i know a little about it, i know a comany that sell all type of electronics in the uk (rs)
but i do learn a bit more day by day and i love soldering

[PrecisionBoost]

I will start up a little post in the electronics section about this stuff.