john banks

As title.

I notice that RC Developments tuning packages for most Impreza applications are based around Apexi AVC-R and S-AFC with a few other bits and pieces.

Notably there is no device for independently remapping timing.

So the questions:

1. Are there massive gains to be had by modifying timing on a car on original turbo and intercooler? How close does the knock sensor take you to "optimum" advance? Or does it advance up to the limits of the timing map subject to no knock?

2. How does altering the MAF signal with a S-AFC or similar (I think you know what I am getting at ) affect timing?

3. How does the MY99/00 work with timing? IE is it able to retard so far and can you tell from watching a Select Monitor how much retard is still available before it would not retard any more? How much of a buffer do you want in terms of degrees for safety?

4. Under what circumstances does the ECU reduce boost if it detects Knock?

I have some ideas for a project to remap the MAF sensor, from a simple two channel design with variable RPM switch point, to thoughts on a uC design with multidimensional mapping.

I would certainly be a lot happier to lean out a mixture from 10 to 8% CO at the top end whilst measuring knock than I would actually playing with the crank and cam position sensor signals.

Before anyone jumps in and says I will blow up my car, chill please. Please answer my questions if you can, and trust me to use the knowledge wisely on my own engine which I am responsible for. Thanks.

Particularly I want to know the effect of MAF alterations on timing if anyone can help on that.

steve rally

John,

From dyno work I reckon that 1 degree of extra advance gives about 4% more torque so if you could separate the timing from the fuelling from a MAF derived map then optimising would be useful.
However not sure how you can achieve this without a mappable ECU!!

The Air Flow corection devices by increasing the AF signal and hence tricking the ECU into supplying more fuel are at the same time retarding the ignition as the std map will reduce advance as boost(AKA air flow )increases.

How near is the map to det?
Not sure on UK cars - but STi V4 runs circa 21 -23 deg at 1 bar boost (checked this with select monitor) and did a GEMS on a Sti 5 yesterday and was getting det at 21 deg!! We run 101.5 RON fuel in rally car and this is worth 7 Deg more advance than SUL so so could say that 100 ron is 5 deg "better" therefore det point would be 26 deg on 100 RON so std ECU has approx 3 to 5 deg safety???

Hope this helps

Steve S.

john banks

Thanks Steve!

The UK cars are safe to run on 95 RON and I run on 97 + booster, so I presume they are a lot further from det than the STis?

I was in fact hoping for a slight advancing of ignition with a lean out over say 4000 RPM since my car runs over 9% CO. It seems a rule of thumb would be about 10-11 lbft/degree of advance then. Others report increase of 10-15 BHP/AFR point leaned out at the top end.

It is just how much advance I would get from leaning out by 1 AFR point using a MAF modifier.

I guess I'll need to see on a select monitor.

I presume if more advance is possible it will only do this if not detecting knock. I would hope that running the higher octane will therefore unlock quite a nice slug of torque then.

How rich do you map with the GEMS at the top end?

Anyone setup a S-AFC or similar would care to comment?

[Edited by john banks - 3/13/2002 8:26:10 PM]

steve rally

Std map advance difference between 1 bar and 1.2 bar is 3-4 degrees.i.e a 20% AF difference retards by 3-4 deg.

Steve

steve McCulloch

Interesting

1 degree = 4% torque...

This means that the rolling roads are total bolloxxxx - which I suspect they are - bit of pub talk

I will be advancing my ignition advance by about 5 degrees with W Injection - albeit the fuelling will be lower - but I expect to get about 10-15lbs torque and about 20bhp - so about 2-3lbs for very degree of advance........

Steve -from a start point is this right - perhaps its just a case of dimins=ishing returns futher up the performance scale

Do you know what the break point is on the ignition advance? - i.e the point at which increasing advance actually reduces power.....

Tony - you should speak to Bob Rawle on this one - or Adam M/R19KET/Trout/CraigH - who I'm sure know absolutely shed loads on the subject that you are looking at.......

Would love to know where I could get 101.5 ron fuel from - that would bee nice.......

I run Shell Optimax with Millers 1/2 bottles per fill up. (2 bottles up north, where for some reason the Optimax is very, very suspect as my knocklink lights up like a christmas tree -[ no such problems down south).

I'd love to see an RC developments car on the rollers - head to head - I'd love to see them beat me on the rollers... I run 740cc ones and thats prior to internal work at 1.4 bar, with no det at all on loads of advance. The injectors just allow me to run safer - as the 550's would be well maxed out....

[Edited by steve McCulloch - 3/13/2002 10:58:31 PM]

steve McCulloch

John

The answer to your question is almost certainly yessssssssss

john banks

Thanks both Steves!

Keep the ideas rolling....

Any thoughts specific to a UK car?

Maddog

Who's Tony!!??!!

Cosie Convert

John


I'm not sure that the question of whether ignition timing is influenced by MAP signal has been answered yet ?

Regarding phoning the 'experts' when I asked exactly this question of BRD whilst ordering my Knocklink I was told rather abruptly that "You have no idea of what you're doing, playing with MAP input signals" Uh, perhaps that's why I asked ! Anyway I never did get an answer

Perhaps the 'non profit making' tuners would care to contribute ?

Andy

[Edited by Cosie Convert - 3/14/2002 12:17:46 AM]

Cosie Convert

Steve Mc

How much power were you producing to max out 550s I presume this must have been at a 'safe' (read high) CO ?
I'm still running 380s with surprisingly good results

Andy

Scott.T

iTrader: (2)

Having just fitted a Prodrive Chip in my MY94 I would say yes.
This chip only gives a boost gain of about 1psi, but the car feels one hell of a lot stronger and more responsive due to the optimised fuel and ignition map.

Althouh is is only giving about 13psi it feels better than when I ran 15psi with the Dawes.

Scott.T

iTrader: (2)

P.S Only 13psi as it is an MY94 with TD05. Prodrive would not allow any more than 1 BAR as that's all the MAP sensor can handle, plus warranty concerns with the plastic intake pipe to the intercooler.

[Edited by Scott.T - 3/14/2002 8:24:48 AM]

john banks

All we are looking for are a few little bits of info to help us on our way, but I suppose we can't expect a commercial mapper to tell us.

I think your results speak for themselves CC.

I am thinking then if I lean by 1 AFR point I might also get up to 2 degrees of extra advance from what has been said so far, which given I am running at least 4 RON above the safety limit I hope I could make use of

JamesS

Engine torque output with regard to spark advance is actually quite flat. You go up the slope of more torque for more advance. This trend slows down around +/- 5Deg of MBT spark (Max Best Torque) and then decays at the same rate with too much advance. It is actually quite hard to determine MBT timing withn 3-5Deg (and thats on an engine dyno). On chassis dyno set-ups it will be even harder.

However as spark is advanced past MBT, exhaust gas temperature will climb.

EvoRSX

John,
Just going back to your original post, you say that there is no independant device for setting the timing....

I believe that on the RC Evo6 (not sure about the Impreza), they use the Apex'i ITC to adjust the timing independantly.

Have a look at the following link on RC's site and you'll see they mention the ITC

http://www.rcdevelopments.com/whytune.html

"The APEXi Ignition Timing Converter (ITC) is a product that allows the user to set the spark curve to +/- 15 degrees at 5 different rpm points: 800, 2400, 4000, 5600, and 7200 rpm. The ITC unit draws a straight line between those points to complete the timing curve."

Andy

steve McCulloch

JamesS - do you know what the reasonable advance settings are, before the critical level is reached..

I guess I should have gathered that the torque increase was fairly flat as Trouts lbs is not much higher than mine running WI - but his bhp is much higher....

Cossie C - I dont actually know about the 550's - but my old sti5 ones (which ones are they?) maxxed out just below 6000 rpm - not a major problem but the injectors might have failed

The 740's are running at max of 76/77% in high revs

77%*740= 570

So the 550's would be borderline

This is running at only 350bhp at Well Lane

I dare say the 550's would be ok, but I want the injectors to be safe, and realistically you should not be running them much above 85% loading capacity

The cossie ran much smaller injectors, at 420 bhp, depends on the engine characteristics - the bhp is irrelevant - as you can have a 600bhp Skline running fairly safely on 550's

Just need to wait for someone to hire out a Dyno or rolling road for the day.... Trout/Bob??? to find the max feasible ignition advance

john banks

For all their Subaru packages though the ITC is not included though?

I think CC runs at 100% IDC and controls the mix with the fuel pressure ?

[Edited by john banks - 3/14/2002 10:35:00 AM]

Stuart Knight

iTrader: (2)

Just a quickie on injector sizes.

In my car (MY94 WRX with TD05) the injectors ran out of duty cycle at

380 = 294 bhp (99% duty cycle, ouch)
440 = 336 bhp (95% duty cycle, too close for comfort)
550 = ? don't know, being remapped as we speak.

EvoRSX

The way that RC get over the fuelling issues (I beleive again ) is that they increase the fuel pressure above standard using an uprated pump and adjustable regulator to provide extra fuel right across the map and then use the AFC to lean it back to what they require at the various load points. This saves the need for having to swap injectors etc when they max out at standard fuel pressure.

Not saying this is the best way to do it but it is an option if your injectors max out and you're in need of just a little bit extra.

Andy

[Edited to add that they also do this so that they can use the AFC to fool the ECU into believing there is LESS air going into the engine so that they can defeat the fuel cut limit but still get the correct amount of fuel]

[Edited by EvoRSX - 3/14/2002 11:19:46 AM]

john banks

If you only want to run 5-5.5% CO for max power and control the det with WI and run 100% IDC adjusting fuel trim for your fuelling, and the injectors do not overheat or stick open then it seems that you can get more life out of 380s on a TD05 at 23PSI as CC does. This is considered "unconventional" though but clearly it works for him.

john banks

Thanks Andy. My intended application is a bit different. Fuel cut is MAP mediated not MAF, and I would like to lean it up top and advance the ignition where safe to do so. ( I am running 9 - 10 % CO at the top end - very rich indeed ).

" Ignition Timing: As boost pressure rises, the ECU retards the timing - in fact by 15 degrees between -10 and +15 psi of boost. Originally I had assumed that boost pressure (i.e. output from the MAP sensor) was used by the ECU as a major input to the ignition map. On further investigation, it appears to have a minimal effect. I would guess that the MAF sensor plays a more significant part (along with RPM etc). " http://www.force8.demon.co.uk/impreza/bcontrol.htm

I gather from those who have experimented that on the MY99/00 there is little input to the ignition timing from the MAP sensor.

Cosie Convert

I need to get my car on the rolling road to conclude the 380cc Injector max output thingy. It did however run a 1/4 mile terminal speed of 115mph when a similar Subaru with 300+ bhp & 314ftlb was running 108 mph. On the same day a stripped out 512 BHP Escort Cossie was running 118 mph.

My dyno calculator (USA software package)tells me my car would need 50 more bhp to increase it's terminal from 108 to 115.

A side by side run from 20 > 90 through 2nd 3rd & 4th gears saw it pull 2 car lengths on an EVO7 which had been dyno'd at 372 bhp.

I can only assume that the figures banded around as being the max outputs of injectors are calculated at an AFR of 11:1 (8%+ CO or thereabouts) If you can run safely at 13:1 AFR (5% CO) then the potential output increases accordingly 300x13/11 = 355bhp.

Then you have the option of further increase via a fuel pressure adjustment, the flow increasing proportional to the sq root of the pressure drop across the nozzle (allow for boost increase in this calc as compensation is not linear at higher boosts !)

So, preparing to be shot down in flames, I'll stick my neck out and say 360bhp is possible on 380cc injectors

Or have I missed something ?

Andy

EvoRSX

Andy,
Just done some quick calcs for you on your injector theory....

Using 4 * 380cc/min injectors = 1.216 l/min of fuel max

at 80% duty cycle and AFR of 13:1 you will get approx 250ish bhp
at 95% duty cycle and AFR of 13:1 you will get approx 305ish bhp
at 100% duty cycle and AFR of 13:1 you will get approx 320ish bhp
at 100% duty cycle and AFR of 13.5:1 you will get approx 345ish bhp
at 100% duty cycle, AFR of 13:1 and fuwl pressure of 50psi, you get 350ish bhp

I Think!!!

Hope this helps

Andy

BugEyed

John

Beware the complexities of the Subaru ECU!

The standard MY01+ ECU (and by association, the MY01+ PPP) runs a very optomistic amount of advance as standard - to the extent that knock occurs) - and then retards the ignition setting dynamically to take account of the amount of knock found. In theory this will allow the car to always maximise the benefits of the available fuel, air temperatures, etc etc. In practice however, it pulls the whole map back rather than the individual points at which a problem occurs. This would mean that to alter the timing via a device to change the ignition timing on the standard ECU would be hard to programme - every time you map one point correctly you will pull back the map for all the other points. This experience in Australia and the USA has lead people to abandon using the Apex-i ITC and similar.

However, I'm not sure how this maps to the "classic shape" Impreza, as the ECUs have got progressively more advanced. The MY01+ ECU constantly re-checks the amount to pull the map based upon the knock sensor and adjusts it over a period of less than a minute. The older models don't, and some would require an ECU reset to achieve the same effect.

Hope that this helps!

Duncan

Cosie Convert

Thanks Andy

Did the calc for pressure take account of the boost pressure being 23 psi ? are you saying 50 psi above boost making 73 psi at WOT ?

CC (to avoid confusion !)

StephenDone

Hi John,

I have advanced my ignition significantly at low revs. This has dramatically improved my turbo spool up, and I still have no det. I have an equal length manifold, and this may have something to do with the increase in available margin on my car.

Steve

EvoRSX

CC (that's better)

Right then,
Calcs done using:
122 cu in displacement
8.0:1 compression ratio
23 psi boost
90% volumetric efficiency
72% turbo efficiency
80% i/cooler efficiency
7000 rpm redline

@13:1 gives power output of 350ish bhp

to do this requires 26.95 US gallons worth of fuel per hour

To get this from 4 * 380 cc/min fuel injectors there are various options

1. At 80% duty cycle for the injectors, the base fuel pressure needs to be 6.2 bar!!
2. At 90% duty cycle for the injectors, the base fuel pressure needs to be 5.2 bar
3. At 100% duty cycle for the injectors, the base fuel pressure needs to be 4.14 bar

The fuel pressure figures are the base pressure and not the pressure above the boost pressure.

Andy

john banks

Thanks Duncan. I wouldn't need to bother worrying about changing the timing on a MY01 from the sounds of it as it is supposed to be optimal.

I wonder if the classic shape has a maximum possible ignition advance and aims for that subject to no knock, and that the timing aims are conservative especially on higher octane fuel. OTOH the new shape sounds as though the aim points are more aggressive and uses the knock sensor more aggressively? Certainly I gather the MY99/00 will advance the ignition on higher octane fuel without a reset, whereas earlier models apparently did not.

Perhaps someone could confirm or refute the above and the mechanism on MY99/00 for timing? Stephen? Pretty please?

[Edited by john banks - 3/14/2002 1:55:19 PM]

AndrewC

John,

I believe this would be a good time to invest in/design some sort of Knock monitoring!

Have you attempted to contact Colin Pearce re. his work which you quoted from to earlier?

Andrew...

john banks

Yes I tried to contact him - would have helped a lot, but no reply, and I have not been able to trace him sadly.

Yes, knock monitoring essential isn't it (!) , I will be working with CC to get his K/link to activate his WI, so will consider what I can build then It is only an LM3914N, and an op amp with a few caps and diodes, and I would really quite like to integrate it into my design and allow the boost or AFR controller to act on its signals. Adam suggested I build a whole ECU, but he overestimates my abilities, and the uCs I am using are frighteningly simple at this stage - not so if I go for high pin count packages and larger code sizes for a multitude of reasons.

I have some very high impedance input 10bit ADC inputs on the uC which I could sample the knock signal with - probably just a case of connecting it (as long as it is 0-5V) and writing some code - no components necessary.

[Edited by john banks - 3/14/2002 2:17:54 PM]