Fuel-cut cannot be defeated

[Guest Ash]

 

 

Did some data-logging today with a VVT-I MKIV.

 

Connected a P/Betts FCD exactly as per instructions. Immediately ran an A/F of 25.4 at 5000 RPM. Thankfully, I had said to my customer beforehand, "If I shout STOP then get off the gas and don't even think of asking why."

 

Removed fuel-cut device and attached peak-hold DVM to signal wire of the PB/FCD.

 

Set boost controller to maximum and attempted to run to maximum rpm, uphill in 2nd gear. Hit fuel-cut at 4.84 volts at 10.0 A/F.

 

Set PB FCD to clamp at 4.40 volts.

 

Tried the same uphill run. This time measured A/F of 11.1 A/F at 1 Bar boost pressure. Whereas a solid A/F of 10.0 had been registered on the previous run at a boost-pressure, fractionally under fuel-cut, of 1.1 Bar.

 

Also, the engine hit fuel-cut (even though it was making less boost and with the PB FCD connected) and the motor felt very sluggish thereafter.  

 

The car was returned to base, the ECU reset and the PB/FCD removed.

 

Blitz boost controller was set to 37 at a gain of 50, settings which had previously been determined to keep boost fractionally below stock fuel-cut

 

Car runs fine.

 

Has anyone actually fitted the PB FCD to the VVT-I car?

 

Because it seems to me the unit doesn't work.

 

Yours,

J

 

 

 

 

 

 

 

[Nick]

I've tried both the TRL and the HKS units on my car.

 

Using the HKS unit and following the instructions that come with it (or rather the translated version) fuel cut was removed.  However I STILL, even with a brand new set of HKS S40i plugs get a high rpm miss fire in forth gear upwards at WOT and max boost, around 1.4bar with the cats removed.

 

I didn't try the TRL unit using the wire stated in the HKS instructions because by that time I was thoroughly fed up with buggering around with the car.  Pete fitted the original unit but had to find the correct wire to use as the VVT-i ECU wiring is different.  It didn't work on this wire, which was a different one to the wire in the HKS instructions.

 

I'm not happy with the way the HKS unit works (dividing the signal) and I'm even more unhappy with the fact that there isn't anyone I know of that fully understands the VVT-i.  As far as I'm concerned the fuel cut device is yet another cludge, a proper replacement ECU for the VVT-i is what I yearn for.

 

If you want to know what wire Pete used and what wire HKS say to use let me know and I'll have a look at the ECU.

 

 

[Guest Ash]

 

 

Nick, I now know the true cause of the high-boost misfire that affects cars with stock fuelling and running boost pressures upwards of 1.2 Bar (or thereabouts).

 

The reason is, the fuel/air mixture entering the cylinder is too lean. Possibly as lean as A/F 20+.

 

When I was toying around with Justin's EVO I confirmed for myself what a tiny number of people had been saying about the HKS FCD in the way it leans out the fuel/air ratio.

 

I reported this to the list and a representitive of a (so-called) leading tuning company came back saying, "Ah, but that's because the EVO doesn't have an excess of fuel at top-end like the MKIV".

 

Immediately, I thought, what excess?

 

The MKIV has a 2-Bar map sensor. Therefore, the maximum possible boost pressure it can detect is 14.5 psig. However, that doesn't take acount of fuel-cut which comes in at around 13-14 psig.

 

But what of these people who were running their cars (including myself at first) at boost pressures of circa 1.4 Bar with stock fuelling?

 

I initially theorised that the turbos were being pushed so far out of their maximum efficiency island, the subsequent heating of the intake air was causing the air to become progressively less dense. This lessening density, I figured, was having the effect of richening the fuel/air mixture. Though I did say that it was all a bit of a mystery as I had not had the chance to take a look at a car.

 

For days following my experience with Justin's EVO, I would study the MKIV engine in my mind. I'd think of every sensor and its operation. But no matter which way I looked at it, it all boiled down to the same two conclusions:

 

1) there can be no extra fuel entering the cylinders after the fuel-cut threshhold has been surpassed.

 

2) fuel-cut is an inherrent aspect of the ECU that cannot be defeated by fitting either the HKS or the PB FCD.

 

Just recently I had the chance to take a look at a MKIV and I now know, for a absolute fact, that there is NO excess fuelling with the MKIV. The moment the fuel-cut threshhold is surpassed, the motor quickly begins to run lean (in exactly the same way as the EVO).

 

So what stops the motor destroying itself?

 

Quite simply, the engine has a *very* efficient knock-detection system. In fact, it is SO effective I take my hat off to the engineers that designed it. Anyone who can design a turbo motor that can hold together at fuel/air ratios of 20+ at WOT gains my absolute respect.

 

From what I have looked into so far, I can already conclude that fitting a fuel-cut defencer without any other provision for feeding in extra fuel beyond the fuel-cut threshhold is sheer folly.

 

Yours,

J    

[mark brown]

Ash

 

You have clarified something for me with your last post. When I initially started modding my car I used the standard ECU with an FCD and I subsequently changed to a Blitz ECU without an FCD (not needed) and the performance improvement was quite big. People had been suggesting that this was due to more ignition advance and no knock detection. Well I have confirmed there is still knock detection and since fitting a separate wide band lambda and AFR meter the mixture remains rich at boost levels of 1.4+. As an experiment I have just tried the standard ECU with FCD and my AFR meter shows the mixture going lean at 1.1 bar. Obviously I've put the Blitz ECU back in.

 

I know you are not a fan of just throwing extra fuel in and I guess I am lucky in that the Blitz ECU fuel map  is a reasonable match for my mods.

 

From what you are saying, I am assuming that you are recommending that anyone running standard fuelling should not exceed boost levels of 1 bar if they want their engine to survive and those running above this level are only still running because they are not pushing the engine very hard. If this is so, a lot of people on this list either need to arrange extra fuelling or limit the boost a.s.a.p. Do you think the same issue applies to UK spec cars?

 

Mark

[Guest Martin F]

 

But Ash, what about the different size injectors fitted between the Jap spec and the UK spec ???

 

Surely the larger injectors must allow more flow at WOT when running higher boost. If so that would lead me to suspect there is more head room on the UK spec car.

[GavinL]

Quote: from Martin F on 8:10 pm on Nov. 3, 2001[br]

But Ash, what about the different size injectors fitted between the Jap spec and the UK spec ???

 

Surely the larger injectors must allow more flow at WOT when running higher boost. If so that would lead me to suspect there is more head room on the UK spec car.

 

and that the UK spec cars are set up for 95 RON fuel.

[Adam W]

So what you're saying is that fuel-cut CAN be defeated (ie the car will continue to operate at high boost pressures) , but that the car will immediately begin running lean at this point.

Although this is obviously a Bad Thing, surely the action of the knock detection system, as you have described it, will provide some measure of safety for the engine?

What I'm trying to get at is, should all those people running 1.2 bar+ on standard fuelling be panicking, or should they be aware that the engine is operating outside of normal parameters and struggling (though apparently succeeding) to keep itself intact?

[Adam W]

Oh, hang on, I just read your initial post again Ash and it says that fuel cut still happened on the VVTi car . . . but do my above points apply for non-VVTi cars?

[THOR Racing]

It just so happens I've built my own AFR and injector duty cycle monitor.

 

I fitted it at the weekend and observed the following (please note that I'm using the stock O2 sensor and that I'm fully aware that it cannot take accurate reading above or below stochiometric AFR or 14.7 or 1 lambda. But it does indicate a general RICH or LEAN car.) Here goes..... this is on a NON VVTi car so I'm shocked that Ash is saying the VVTi is so different.

anyway...

 

 

The top row of LED's is injector duty cycle. I connected one of the injector control lines from the ECU to it.

The bottom row of LED's is the AFR.

 

If you start the car from cold the indicator will show LEAN. This is

normal, the sensor is not warmed up yet. Gradually as the engine warms

up the indicator will rise to RICH and then gradually start to pulse

(Knight Rider style) up and down between RICH and LEAN. This is the

closed loop response from the ECU. It's cycling the fueling to adjust

the AFR to 14.7 which is 1 Lambda or Stochiometric. This is the ideal

AFR.

 

You'll also notice the injector LED is at the lowest position, one or

maybe two LED's lit.

 

If you now drive the car you'll see the pulsing get faster or if you

accelerate faster you'll see when the ECU switches between CLOSED loop

to OPEN loop. In OPEN loop mode the ECU does not use the O2 sensor for

fueling as it's sensing the MAP pressur sensor. However the O2 sensor

still senses changes in AFR. You'll see the AFR shoot towards RICH.

You'll also see the injector duty rise as it delivers more fuel to your

engine.

 

Now under the worst conditions of hard acceleration the following will

happen.

 

You stick you foot down, the boost builds and you start to fly. You'll

also notice that the AFR is reading steady (ish) around RICH and the Inj

duty is showing probably about 80->90%.

 

In the very very worst conditions where you are getting into trouble the

inj duty cycle will rise to over 95%, ALL LED's on the top row are lit.

 

The AFR should still read RICH! BUT if you've pushed the boost too high

or the engine is NOT getting enough fuel the AFR will fall to LEAN.

This is VERY BAD and detonation is sure to occur v.v.soon.

 

At the end of the day so long as the AFR never falls to LEAN or below

STOCH you'll be O.K. I think, but Ash is saying on the VVTi this happens just above Fuel Cut. Wow! I was running 1.2Kg/cm2 and the AFR never budged from full on RICH.

 

If you notice that the Inj Duty is running hard at 95+% then maybe

it's about time to invest in some bigger injectors which a few people have done.

 

Fitting a Wideband O2 sensor close to the existing O2 sensor in the

downpipe will mean the display will now read something more sensible.

Essentially the stock sensors are really only any good around 14.7 (1

Lambda), taking any meaningful measurements above and below this mid

point only really serves to indicate RICH or LEAN. A wideband sensor

will indicate a larger range correctly.

 

This would certainly explain a lot regarding the VVTi fuel cut scenario and why the non-VVTi NEVER has a problem!

 

Nice one Ash. What to do about it though? Did you measure the injector duty cycle at all?

 

Regards

 

Pete

 

[GavinL]

Quote: from TRL Performance on 8:24 am on Nov. 5, 2001[br]If you notice that the Inj Duty is running hard at 95+% then maybe

it's about time to invest in some bigger injectors which a few people have done.

 

 

 

I read an article, I can dig it out if you wish, which states that Pintle type fuel injectors which I understand are fitted to the Supra, are prone to failures at 86-88%  

 

 

 

regards................

 

 

(Edited by GavinL at 9:22 am on Nov. 5, 2001)

[THOR Racing]

Yes it would be nice to read the article.

 

I know most injectors only really like to handle 80% duty cycle and above that they are a bit stressed.

 

Fitting a rising rate fuel pressure regulator or bigger injectors would be the key here.

 

Pete

 

>I read an article, I can dig it out if you wish, which states that Pintle >type fuel injectors which I understand are fitted to the Supra, are prone >to failures at 86-88%

[GavinL]

Quote: from TRL Performance on 9:30 am on Nov. 5, 2001[br]Yes it would be nice to read the article.

 

 

http://www.supras.com/~riemer/sonictech/fuel_injectors/RCtech.html

[Ian C]

You techies can clear up a couple of things for me along these lines:

 

1.  My Apexi unit has an injector duty cycle reading, and I've never seen it above 60%.  How accurate would this be, and how come it doesn't get higher than that?  J-spec injectors, but with an FSE valve giving it some extra wellie.

 

2.  My boost gauge is in kg/cm2 - I thought it read in Bar.  Are the two measurements the same thing, e.g. 1.45Bar is 1.45Kg/cm2?

 

3.  I recently had a 'blip' in the boost pressure (the only time it's ever done it), and it went up to 1.5kg/cm2, and I felt a noticeable power drop - would this be the engine retarding the timing?

 

-Ian

[Guest Ash]

 

 

Mark..........

 

Forgive my delay in replying, what you say makes a LOT of sense and entirely concurs with my figures. I have little data on the UK car as all I ever seem to see are modified J-specs. Gavin L has kindly donated his car for analysis. So when I get the chance, I will be able to come up with some data.

 

Yours,

J

 

[Guest Ash]

 

 

Ian..........

 

Which Apexi unit are you talking about?

 

1 Bar = 14.5 psi. To convert kilograms per square centimeter to bars, multiply by 0.980665. To convert to pounds per square inch, multiply by 14.223343. Therefore, 1 kg/cm = 14.2 psi (to one decimal place).

 

As regards (3) yes, timing, and because the fuel/air mixture would have been way too lean. An overly lean mixture will cause a big power loss. An overly rich mixture likewise.

 

Yours,

J

[Doughie]

OK I am just trying to reconcile the following 2 statements/opinions :

 

Ash : "I now know, for a absolute fact, that there is NO excess fuelling with the MKIV. The moment the fuel-cut threshhold is surpassed, the motor quickly begins to run lean"

 

Pete : "I was running 1.2Kg/cm2 and the AFR never budged from full on RICH."

 

Now, I know that one car was a VVTi and the other wasn't and Pete was using the stock O2 sensor which apparently isn't that accurate, blah blah.

 

BUT the above 2 observations/statements/opinions do seem to directly contradict each other.     If there is no excess fuelling above fuel-cut (14psi ish) on a mkiv, HOW can Pete's car be running RICH ay 1.2kgcm2 ?      seems to me that SOMEONE must be wrong.   (noting the above caveats).

 

since as someone else has already remarked, that almost all of us are running boost levels above fuel-cut, by definition, we are all very interested in hard facts on this fuelling issue.

 

Pls could Ash and Pete suggest why there is clearly a big difference between their 2 experiences.

 

(Edited by Doughie at 9:49 am on Nov. 7, 2001)

[Guest Ash]

 

 

 

Doughie, there is a *big* difference between attaching thousands of USD's worth of calibrated UEGO instrumentation to the exhaust, and hooking up some home-brew gismo to the stock oxy sensor.

 

But don't get me wrong, Pete is entitled to his opinions. I have purposely declined to directly answer the points raised in his post as I do not want this issue to turn into a "me against him" scenario.

 

Yours,

J  

[Doughie]

 

All I want is to find out the truth as to at what boost a UK MKiv TT will start running lean, that's all.

 

- really doesn't matter who's right, just want to know !

[Guest Ash]

 

 

Me too, I just want to know the facts.

 

That is why I've spent around £8,000 on various items of *calibrated* instrumentation that will tell me precisely what the situation is. I already know about the VVTI car and on Friday, if it is dry, Justin and myself will have a play with his. And once I finish Mark Ayling's car, I'll have a little more time to spend with Gavin L who has kindly donated his UK car for fuelling analysis.

 

Also, I'll take Mark A's car out and check it for comparison with JF's car. It's just that, up until now, I had neither the time nor the cars to test.

 

Yours,

J

[Rich J]

Pete : "I was running 1.2Kg/cm2 and the AFR never budged from full on RICH."

 

Mine was exactly the same at 1.2-1.4kg.cm2, I even went so far as to influence the fueling with my S-AFC and had to set it 10% lean to get the last light to turn off (didn't even turn off for the full run)

 

 

 

(Edited by HardHead42 at 7:23 pm on Nov. 7, 2001)

[Doughie]

aha !

[Guest Martin F]

Quote: from Ash on 7:51 pm on Oct. 27, 2001[br]

 

 

Blitz boost controller was set to 37 at a gain of 50, settings which had previously been determined to keep boost fractionally below stock fuel-cut

 

Car runs fine.

 

Has anyone actually fitted the PB FCD to the VVT-I car?

 

Because it seems to me the unit doesn't work.

 

Yours,

J

 

 

 

Just reading this again.............Errrrr gain of 50 !!!!!!!!!!!!!!!!!

 

What are you trying to achieve, scorched earth ??????

 

 

[THOR Racing]

I completely agree that $8000 worth of proper fueling/tuning equipment is a lot better than a £100 device to effectively measure the O2 sensor. However even the stock O2 sensor has the ability to measure accurately air fuel ratios which go above and below the stochiometric value of 14.7:1

 

The transfer function of the stock O2 sensors gives you a voltage vs AFR curve which is very very steep around 14.7. In fact very small changes in voltage can cause large changes in the output of the sensor. This is what you see when you observe the closed loop operation and watch the LED's (or voltage) pulsing up and down between rich and lean.

 

 

That fact is that on my car the voltage never even hints at going towards the lean area as it's permenantly stuck at rich. If it was border line (close to 14.7) then the LED's would rapidly shoot down to LEAN and back up to RICH to warn me something was about to go tits up.

 

If it ever got to show permenantly lean (i.e.

 

I hope I've explained that well.

 

To be more accurate then a wideband O2 sensor can be used, all these give are flatter characteristics and have a linear output for a greater range.

 

I'd like Ash to discuss it. No need to talk about my device or his equipment (mines bigger than yours etc), we just need to discuss WHY the VVTi is going LEAN? in the first place.

 

I've sent one of my units to Paul Whiffin (MkIV) and Terry Steele (MkIII) so hopfully we can have some other info on the stock sensored cars.

 

It maybe my car has a dodgy fuel pressure regulator and that's how I've got more fuel. I know when I did the 12V fuel pump mod (which I don't recommend) that it was MEGGA rich and was using almost twice as much fuel. This wasn't the case for most MkIV's that tried the same mod.

 

I admit that using the stock sensor is only good for showing general LEAN or RICH conditions, in fact you could argue you only need 3 LED's for this.

But I did it this way so I can attach a more accurate Wideband O2 sensor.

 

I don't agree that $8000 worth of equipment is any better at indicating the same general RICH/LEAN condition than the stock O2 sensor or WHY would Toyota have fitted it?  After all that's what is for. The ECU cycles between RICH and LEAN for the SOLE purpose of averaging out the AFR to 14.7 !

 

I agree it's FAR FAR better at giving precise fueling details but no better at the simple binary result of RICH/LEAN.

 

Now that was not a flame or any criticism of anyone just my professional opinion which I am entitled to. The reader is entitled to his own and I respect that.

 

regards

 

Pete

 

 

 

[Guest Ash]

 

 

Just reading this again.............Errrrr gain of 50 !!!!!!!!!!!!!!!!!

 

What are you trying to achieve, scorched earth ??????

 

 

 

 

Not sure what you mean, the car ran great. Have you had problems with the Blitz EVC as regards the gain settings?

 

Yours,

J

[Guest Martin F]

 

Just that most of us with the DSBC or SBC-iD are running gains of no more than about 15, as any more results in very large boost spikes.

 

This also correlates with the settings given in the un-official user manual.

 

What peak reading did you get after a run with the gain set at 50 ??