john banks

Anyone actally experienced the full fuel cut head through the windscreen thing and at what boost. MY99/00 I am interested in.

I have the factory solenoid disconnected as I use a boost controller - the solenoid makes the boost come down if it is looking too high (but can't now it is disconnected ), but I just wonder how high you need to go before the fuel cut. I am running 18PSI held @ 0 degrees C (with up to 0.5PSI peak on the gauge), with no knock on Select Monitor, with good ignition advance. I don't want to run any more, but want to know what happens if a hose comes off. I have seen over 20PSI briefly when setting boost up on one occasion and another when an unclamped hose came off the turbo nipple (stupid thing to do they are all clamped well now) with no cut because I backed off right away.

[Edited by john banks - 12/29/2001 1:52:01 PM]

steve rally

John,

The PPP works by modifying the MAP sensor signal before it enters the OEM ECU.The map is basically a percentage modifier against revs and base MAP voltage.This makes the ECU run higher boost as the closed loop control system thinks it is seeing lower boost than it actually is. This also give some more advance as the OEM map is based on boost.The fuelling is not changed as far as I know cos this is derived from the MAF and the OEM map.There is also an overboost timer in the PPP software and this is probably why you aren't seeing the BIG CUT. I think this is set to 1.4 bar for 50 secs....

Hope this helps

Steve.

john banks

Where do you get your info for this Steve? Reason I ask is that they give you a new ECU (STi branded but altered obviously). Also on the Select monitor, the ECU reports the true boost even though it is the same program as MY99/00 ECU that is used on the monitor - if the signal was modified before the heart of the ECU then the ECU's serial port would not report the correct boost to the Select Monitor unless you are saying the Select Monitor knows it is a PPP ECU and alters the figure accordingly. Also modifying the MAP sensor signal has no direct effect on boost except to move the fuel cut - ie you slap a FCD on - even a simple resistor or voltage clamp a la Superchip and the boost doesn't change until you put the bleed valve on. A few people have suggested that the PPP ECU fuels more accurately at higher boost when they have modded a PPP car. 50 seconds at 1.4 bar seems a bit lax to me.

I am puzzled. Everyone seems to have an idea on what the PPP ECU actually is, and the ideas all seem different, which is why I ask the source of your info, certainly not to be antagonistic. Prodrive obviously do not give much away at all. The most I heard from Mike Wood was that there as still a fuel cut and you may find it if you fit a downpipe. Despite seeing over 20PSI I did not find it although I backed off quickly and the solenoid certainly has a go at controlling boost.

mutant_matt

John,

Being the cynic that I am, it seems to me that perhaps ( ) the reason that Prodrive are very tight lipped about it is that if we knew what the PPP ECU was doing, we would probably not buy the product and could get equally good results using other ( cheaper ) products.

Having said that, you can't blame Prodrive really as they are a business and are in it to make money - not for the love so why should they divulge their secrets, time and effort!! I think a lot of the cost is there to build in the fact that your warranty is not invalidated.

Anyway, Steve, I backup John's question and would too be very interested in your source of info as it is different to most of what I have come to think (guess) was the case about the PPP ECU?

Ta,

Matt

john banks

Agreed Matt. I don't have a warranty as mine is a Euro Import - whilst it may have had years 2 and 3 from Belgium I have invalidated that with PPP. For me when I had it fitted I wouldn't have known where the wastegate was - I could just find the intercooler and turbo (!), so it was a helpful local fit thing that was supposed to need no tinkering or monitoring. Now I dabble a bit more (mainly because when I breathed on it it overboosted and I was trying to sort it out) and would probably choose differently, but knowledge is wonderful in hindsight.

Also I WAS (really ) going to stop with just the PPP.

EvilBevel

John, like we discussed before ... there is no "closed loop" boost control, the ECU uses a duty cycle map for the boost solenoid, and the PPP may use a slightly more agressive one. For the sake of it, let's say their ignition & fuelling is also a lot better than witht he OEM ECU. And since you are seeing the correct boost at the ECU, it does not alter MAP signals (it will just have a different cutoff point)

If you really want to find that point, try using higher boost for more than 2 (yes, two) seconds. That's what trips the MY99 OEM ECU, and I would not imagine Prodrive to change that mechanism.

If you *don't* find it still, well... erm... not so good is it ?

I've tried it on my MY99 with standard ECU, and the boost started to fluctuate at 1.1 bar (it reacts by changing solenoid duty cycle, you obviously wouldn't see that anymore with your setup). Holding it for more than 2 seconds (I deliberatly experimented with this) produced the "desired" face-to-windscreen effect.

My guestimate is that the same *should* happen with the PPP, but say at 1.3 bar or thereabouts.

john banks

Fair point Theo. But we have a Rally Prep Techie and yourself whose approach and help I really appreciate and respect saying completely different things! One says closed loop boost control and MAP modification, the other effectively says it is a remap and there is no closed loop boost control.

"Desired" fuel cut - I just want to know it IS there, and where, so that I do not run very close to it, and so I don't worry what would happen if a ball stuck or a hose came off. Maybe I'll have to try it, but going much above 1.3 bar would freak me out.

EvilBevel

Well, I could be wrong, and I reserve the right to talk boo boo

But ... if there was closed loop, there should be no overboost with a PPP + downpipe (and Mike Wood warned against that happening, and he is from Prodrive last time I checked ) as the ECU would constantly measure MAP & adapt duty cycle accordingly.

At least, that's what logic would suggest ? That's what closed loop means, no ? Still, I could be completely wrong.

john banks

No I don't think you are wrong, I am just trying to understand how the thing works.

Some people imply there is a degree of closed loop control in that no matter what breathing mods you do the HELD boost seems to always be the same which would suggest some degree of closed loop control. The restrictor size and the breathing mods seem to affect the rate of spool up and the peak you see, but no matter what I did to my PPP, it always stabilised at 16.5-17PSI held - as long as it had got that high on spool up - otherwise it held a lower figure with no overshoot. How does it do this in all temperatures with different breathing mods if it is not closed loop? Is it just a function of boost pressure passing through a restrictor and reaching steady state with a duty solenoid (controlled by a static dumb map of RPM and TPS) causing a bleed effect - so how does it hold the same value with different restrictor sizes?

I am keeping you busy here Theo on two threads at once Thanks again for all your help! Much appreciated.

[Edited by john banks - 12/30/2001 10:03:16 AM]

EvilBevel

LOL, maybe we should email this, although I think the reasoning (even if wrong) may be of interest to other people.

>>How does it do this in all temperatures

I knew you were gonna say this . I do believe the ECU reads environment temps, and may alter the duty cycle accordingly (add/remove a percentage ?). Same goes for altitude I *think* (don't know how it would read that though Even then, it's a known fact that you see slightly higher boost on cold damp mornings, even with a standard car.

john banks

We also get threads from people saying that a bleed valve only increases peak not held boost, but then when Superchips fit a FCD and a bleed valve on the same or similar hose their held boost goes up as well.

So are we saying we could make a safe boost controller using the OEM system by using a notch type FCD which only raises the cutout rather than abolishes it, and then reduce restrictor size to give a higher peak which is then held with this (possibly) closed loop system based on a fooled MAP voltage? Presumably there would be a rise of held boost across the range? I would have considered a Superchip IF it did not abolish fuel cut entirely.

The obvious caveat would be that the fuelling and timing are OK, but this seems likely given that most UK/Euro spec cars overfuel and also that the ECU is still getting true readings from the MAF sensor.

I think the Superchips method is quite sensible as an easy mod to an overfuelling turbocharged car - if they had just MOVED the fuel cut by a few PSI (which would stop the end user cranking it up to 25PSI) and used a new restrictor instead of a bleed valve to also reduce fiddling then I think they could have taken a bigger chunk of the Subaru market. I would have bought one anyway.

http://www.mr2.com/ARTICLE/FCD.html

This article shows a suitable circuit. It looks a lot more complicated than it is. I think this is the sort of thing you mentioned in your email to me previously Theo, but this is the only circuit I have found on the web for it. All the others remove the fuel cut or mess up the signal all the way along.

The voltage response is like this (voltages 1 and 2 can be moved with potentiometers):



I should title this one "How to make a better superchip for £10"

Perhaps we do NOT have closed loop boost control of any description as a Superchips car holds a higher level of boost - now if you are holding 20PSI like some do (!) then this is above cutout so the voltage response of the FCD will be a flat line if it is just a voltage clamp giving no real signal for the ECU to use as a closed loop target. So perhaps it is a dumb system after all??

[Edited by john banks - 12/30/2001 11:31:10 AM]

T-uk

john,

how big a job would it be for you to set your PPP/Dawes to 16psi and then refit your standard ECU to see if it boosts the same,you could also see if the lambda shows any difference between the two.you may have to run the car for a few miles before the standard ECU would boost while adjusting to the new changes.

if you try it give me a shout and I'll come through to watch boost so you are watching the road.

also probably stupit question,but with the Dawes have you tried it with the solenoid in place just to see what happened.

john banks

It would boost the same - the only thing controlling the boost is the Dawes - ECU has no role at all. Now I am on a three LED lambda display (which is neatly mounted in one of the plastic blanking plates by the foglight switches with no nasty boxes floating around) I only know if I am at or over 7% CO with the green LED - I don't know how far over as this is my richest LED.

If you want to buy an LM3914N LED driver chip from Maplin (order code WQ41U) on Dalry Rd when you are next in Edinburgh or can mail order www.maplin.co.uk (they are only £3.99) and a 9V PP3 battery and bring your car round I can link up a 10 LED bargraph display to your car just by scotch locking a temporary connector at your ECU and check your fuelling for you. Unfortunately my only chip is soldered onto a board in my car now hence we would need another, but I still have all the other bits required.

The only fuelling data that will be relevant to your car will be taken from your car. I would be wary about doing or not doing something just because the fuelling was OK or not OK on someone else's car. I am also not entirely sure that you would be able to tell much difference between ECUs fuelling unless you use a posh exhaust gas analyser on a RR. The lambda sensor only gives you a rough estimate, useful though it is. Somehow I knew my car was running right before it went on the Select monitor With that amount of richness on boost and at the top end I could not forsee a problem.

Edited to say that the Dawes would work OK with the solenoid in place if set to a low setting (say below 10PSI) and would just reduced wastegate creep. However, if you set the Dawes to a high setting and connected the solenoid at the sort of boost levels I run it would overboost horribly and then the solenoid would pull back its duty cycle and it would undershoot. But I haven't tried it, but something tells me it would be very naff if you had a high boost setting.

[Edited by john banks - 12/30/2001 12:40:21 PM]

T-uk

john,

I have a knocklink/sensor and lambdalink on order,I have to admit I am impressed with the way you have concealed your lambda but I may still go Link ECU one day so just ordered the Link stuff in case.

john banks

This afternoon we tried different sized restrictions in the hose off the turbo nipple of T-uk's car on std ECU. Whatever we did the held boost refused to budge from 14 PSI. The size of the restrictor changed the peak boost and if we went over 16 PSI peak even for a short time the solenoid pulled back the boost to 13-13.5 before stabilising again at 14 PSI.

This convinces me that the held boost is controlled by a closed loop system using info from the MAP sensor. Hence a FCD alone or a bleed valve alone will not increase held boost but the two together do.

Bitten Hero

Very interesting thread - keep it up folks!

The restrictor thing backs up what StanS found. He changed to a 0.8mm restrictor (I think) from a 1.2mm, and was getting 22psi peak boost and around 18psi held. Changing to successively larger restrictors reduced the peak boost but left the held boost unchanged. He had (at the time) a complete Scoobysport exhaust (no cats etc) including downpipe etc which is probably why he was seeing a reasonably high held boost. This was with a PPP ECU and Samco hoses but no other PPP bits.

I agree with John on the closed loop bit too - I can see a peak of up to 19psi on cold days in 4th/5th after some aggressive driving, but it always heads back to a steady 17.5psi (held) after a couple of seconds. (MY00 + PPP + air filter and BladeRunner's intercooler splitter, otherwise standard).

Incidentally, I tried the 0.8mm restrictor in my MY00 (pre-PPP). BAD results - the loop becomes unstable, and the boost fluctuates between 9psi and >25psi Thankfully I was sensible enough to test it with only about 1/4 power in low gears and once I saw what it was doing I immediately swapped it back. Spinning all four wheels in 2nd at 30mph on a perfectly dry road is not normal behaviour!

Cheers
Richard

[Edited by Bitten Hero - 12/30/2001 6:05:34 PM]

steve rally

Hi All

Thanks for that feedback.
John, I think you have come back to my point that the OEM system is closed loop with MAP signal as base data - but you are right to point out that the restrictors do influence response. The system is limited simply by the complexity of the OEM software in that the PID (proportional,integral,differential) parameters used are not that refined - so you can move the target boost by altering "mechanical" parameters.
So that does mean that modifying the MAP input will mod boost.You point about a Superchip is valid - but I have tried a few experiments on a MY99 and if you move the clamp voltage (by turning the 10k pot) you do change the boost.(You also change the advance too!!)

The reason the Select Monitor reads true is simple - it is still taking its data before the implant.

How do I know this??

We are ruuning a V4 Type R GpA rally car with Electronic Centre Diff and have used the PPP module as a means of controlling the diff lock through throttle position.So we take the 0 to 5V TPS signal and modify it using the PPP "map" into 0 to say 3V to control the diff.Obviously we can change the map with the right software. My estimates on overboost were just that based on looking at some test maps that were in the module.

Hope this clears some points

Steve

john banks

Thanks Steve. I had just presumed that the value reported to the Select monitor would be what the boost control circuit was seeing as well since by the time it got to the select monitor it would have been through a substantial part of the ECU to be digitised and converted to serial data for the serial port. However, it sounds like the Select Monitor interface is functionally quite near all the sensor inputs into the ECU.

So by moving MAP by say 2 PSI that may be where we get more sensible fuelling and advance rumours about PPP and a 2 PSI higher cutout and we might be able to do something like a PPP ECU by moving the MAP a bit with even a simple resistor - would there be any implications of shifting the MAP a bit across the range?

steve rally

John,

Moving the MAP signal is what the PPP is doing fundamentally.If you view the software you see a 3D table with RPM as X-axis,MAP raw input Volts as Y-axis,then a grid of percentage modifiers.So you can change the MAP the Std ECU sees up or down.You obviously tend to always reduce the raw input.Also there is an override of extra percent mod which is time adjustable.
There is still overboost protection because you are still seeing the true MAP value so you can set this to overide the table if you want.
The fuelling is not modified at all as the Std OEM MAF derived map is still used.
Advance,as you poiny out,moves by virtue of the ECU seeing a lower MAP value,but again the advance map is Std OEM as far as I know.

Not sure you could move it with something as simple as a resistor.The input impedance the MAP sensor sees is very high to start with and the ECU also has some check functions in there to look for say 2.32V at ignition key ON.(Atmospheric pressure).

Steve

EvilBevel

Hmmm, Steve, it looks like there's humble pie on my NYE menu tonight then eh

Let me chew on this a bit ... but I think you may be right about a few things, and me wrong.

Theo [crunch crunch ]

john banks

It couldn't be too difficult to hook up a PIC and transform the MAP data with a look up table. Hey presto PPP for £10 But to me the solenoid duty cycle map on the PPP is still not as good as it should be - if it held more at the top end I could forgive the slower spool up than an MBC.

The ideal would be a boost controller which also transforms the MAP properly rather than using crude voltage clamps. This could plug into a standard car and use the standard solenoid and absolutely transform it.

Theo - nothing wrong with being wrong - I am all the time and you have corrected me before. It nourishes the soul and keeps the head small

Now what I want to know Theo with your PWM duty cycle map with your Unichip whether the duty cycle is static, or whether there is a higher duty cycle until you are getting near target and then it adjusts closed loop style to home in. I am trying to write the code for my MAP modifier/boost controller as we speak

[Edited by john banks - 12/31/2001 7:46:49 PM]

steve rally

John,

Do you want me to re-map your PPP - and not tell anyone?????

On the rally car we use the std boost control valve with full closed loop control. The duty cycle is set with closed loop swithched off to get as near to target boost,then an error table keeps the boost at target(which is set at 400RPM intervals) with closed loop switched on.It is very accurate!This is on a full GEMS system.So I feel that the OEM valve is OK.

Steve

john banks

Steve it is great to get this sort of input on the BBS.

Reprogramming the ECU is certainly an enticing possibility. You are saying this ECU can be flashed or something? Basically we could choose duty cycles and boost targets?

The std duty cycle is 14Hz I gather. Can the solenoid run at 100% duty cycle to minimise wastegate creep or would it not be rated for this?

I am still wondering - if you use a static duty cycle that is too aggressive you would get too high boost but quick spool up. That is why I am asking if the duty cycle is modified DURING spool up? Am I making sense here? If the duty cycle is static it would take longer to reach equilibrium (ie target boost) than if the duty cycle was varied as target boost was approached? Is this why the Dawes is so quick? How does the std ECU do it? (or Link for that matter Sam?)

steve rally

John,

I have the software!

On the full GEMS system we run the valve at circa 45 Hz.To get 1.6 Bar peak boost at 3000 to 4800RPM the duty cycle is typically 75%.We start with 100% duty cycle from 1200RPM and ramp it back at 2800RPM to prevent overshoot.

The std ECU would have PID loops in the software to give good control.So it does look at the rate of change of boost and adjusts the duty cycle accordingly.The GEMS is not as sophisticated as this - you have a base duty cycle to get you "near" then a table that you put in duty cycle offsets (in percent) for a given error from target.E.g if boost is 0.1 Bar away from desired target we put in +10% offset and so on.If you make the offsets too large the response time is fast,but you can get "hunting".If too small then you may not hit target boost.

Steve

john banks

What does the software run on and how is it connected to the ECU? Does it use flash memory or EEPROM?

Does it use some expensive proprietry system?

steve rally

John,

It uses an EEPROM.

If you want to discuss feel free to ring me on 07836 635001.Not going out tonight!!!

EvilBevel

John I'm way out of my knowledge here (and sad for posting this at this day/hour ) but...

The unichip boost controller runs at 15 hz...

I think I sent you a link before to the map, but it goes from approx 95 % from 1600 rpm to 3200 rpm, 90 % at 4000 rpm, 88 at 4800, and then gradually goes down to 70 % at 6400 rpm.

Mind you, this is after we changed the restrictor. Also, this is for a VF23 which is a different thing altogether. But it might give you an idea.

HTH,

Theo

john banks

Steve thanks - I'll give you a ring - maybe a bit late at this moment - I was up from midnight to 8am for all but 1/2 hour earlier today so I am not staying up. Too busy doctoring when I should be building boost controllers

Does anyone know how the standard ECU or the aftermarket ones control part throttle boost - what sort of adjustment to duty cycle or is it done some other way? Certainly I would have thought there would need to be compensation for this - with the Dawes the part throttle response can be snappy although it is improved with an additional bleed. Using a duty cycle would be slower spooling up than the Dawes I reckon but the PPP spool up is fine it is just the targets I don't like.

A duty cycle map based on RPM on my turbo would be about 100% until about 2300rpm, and 100% again after 5000rpm. In between I would want it flat for 18 PSI. If I add compensation for TPS voltage and a correction factor it sounds as though it would work quite well. A resistor equivalent to the solenoid would be put across the ECU's solenoid output. PIC 16F84 seems an ideal chip to use - I already have an RS232 programmer. If I use a power transistor and a diode to drive the solenoid, then I can use the PIC to do look up tables and duty cycles in software. I would also put in a 19-20 PSI 0.5 second cutout by dropping the duty cycle on the solenoid. I may need to look at a PIC with onboard A/D and D/A conversion if things get fancy, but I don't want to be soldering 40pin DILs if I can avoid it.

If this does all get off the ground and work I would put the circuit diagram on here and the PIC code.

[Edited by john banks - 12/31/2001 11:47:28 PM]

mutant_matt

John,

There's so many Dawes related posts at the moment I don't know which thread to post this in so here is my question:

Can you post the URL of the thread which discusses how to make an AFR gauge? (or email me the URL)

Cheers,

Matt

mutant_matt

John,

Ignore that, I just found it in my Favorites - duh!!

Matt